fix(environment)!: keep unsupported HAR body encodings opaque

Preserve declared encodings in capture Evidence and distinguish unsupported
response decoding from an absent body in shared replay and state extraction.
Use the production HAR decoder in replay regressions instead of synthesizing
its binary sidecars.

BREAKING CHANGE: replay metadata body_encoding can now be "unsupported".
Clients must treat its null body as unavailable decoded content while
body_recorded remains true; the cited Evidence retains the encoded source.

.agents/skills/rea-analysis-optimization/SKILL.md

@@ -0,0 +1,113 @@
+---
+name: rea-analysis-optimization
+description: Optimize REA providers and shared analysis workflows for agent context, latency, CPU, and memory while preserving correctness and Evidence. Use for analysis bottlenecks and cache or query design; CI/test-suite optimization belongs to rea-ci-tests-audit.
+---
+
+# REA analysis optimization
+
+Find a concrete source of unnecessary work, change it at its owner, and verify
+equivalent completed investigations. A useful outcome can also be a measured
+decision to leave an implementation alone.
+
+Follow the checkout's [performance priorities](../../../AGENTS.md#performance),
+[tool design](../../../docs/tool-design.md), and
+[testing guide](../../../docs/testing.md). This skill supplies an optimization
+workflow, not additional permission to execute targets, publish, or merge.
+Use the current contracts, provider prerequisites, and user authorization.
+
+Read [the Kuna case study](references/kuna-case-study.md) when selecting a
+reuse strategy or interpreting a benchmark. Its pinned observations illustrate
+tradeoffs; they are not current-provider guarantees or universal speedups.
+
+## Pin the question and the correctness oracle
+
+Describe the analyst's completed task, including the facts and relationships
+needed to answer it, requested detail, source access, and meaningful unknowns.
+Choose a representative real artifact and a small or empty case. Record the
+baseline revision, target digest, engine identity, options, host, and entrypoint.
+Use a separate checkout for changes without disturbing other work.
+
+Establish an oracle independent of the changed implementation: source-built
+fixture expectations, independently mapped file bytes, or a native producer
+paired with checks of its interpretation. Comparing two wrappers around the
+same faulty decoder does not establish correctness. Preserve address widths,
+same-address alternatives, unresolved endpoints, coverage, and source linkage.
+An empty result proves absence only within the actually examined scope.
+
+## Locate the earliest unnecessary work
+
+Trace one public CLI or MCP task through shared application code to its producer.
+Measure relevant phases: admission and identity checks, native analysis,
+decoding, lookup, projection, serialization, Evidence retention, and cleanup.
+Include catalog discovery and extra calls when they affect the completed task.
+
+Use the evidence to choose the smallest structural correction:
+
+| Observed cost                     | Candidate correction                                       | Condition to establish                                                               |
+| --------------------------------- | ---------------------------------------------------------- | ------------------------------------------------------------------------------------ |
+| Repeated scans of validated facts | Reuse existing order or lookup structure                   | The owner establishes ordering and identity; every matching alternative survives     |
+| Repeated equivalent engine work   | Share a validated result or in-flight producer             | Inputs, freshness, lifetime, and per-caller outcomes remain explicit                 |
+| Full reports for selected facts   | Push selection or projection to the earliest capable owner | Required facts and evidence remain reachable; examined coverage is reported          |
+| Large temporary transport buffers | Decode incrementally or use bounded source reads           | Validate the whole admitted document; retained facts and copies still consume memory |
+| Many repeated homogeneous reads   | Share admission, decoding, or batch work                   | Preserve item identity, order where required, and individual failures                |
+
+Do not introduce an index, generic cache, batch endpoint, or orchestration layer
+without identifying the work it removes and its maintenance/resource cost.
+Projection after materializing a full report may reduce output while leaving
+analysis time and peak memory unchanged. Fix missing or incorrect facts before
+benchmarking an alternative composition against a better-covered workflow.
+
+## Keep one interpretation and a defined reuse lifetime
+
+Let the producer boundary validate and normalize facts once. Consumers query
+that representation through the shared CLI/MCP implementation; adapters should
+not rebuild relationships or independently reinterpret raw reports. Keep lookup
+identity separate from display formatting. Reuse contracts when their intent
+matches, honoring requested detail and complete Evidence access.
+
+For every cache or shared producer, state its key, owner, invalidation, retained
+size, and lifetime. Account for result-affecting target bytes, engine contents,
+analysis options, architecture/load mapping, and parser/schema versions where
+they affect reuse. Validate live state before calling a cached observation
+current. Cache a refusal only when it is deterministic for that validated key;
+malformed, partial, transient, or cancelled attempts need explicit retry policy.
+
+One cancelled waiter must not invalidate a successful peer. Preserve producer
+failure context and each caller's cancellation reason. Retain ownership and
+capacity charges until shared work and cleanup settle; preserve completed
+observations when cleanup fails. Exercise invalidation, concurrent consumers,
+last-waiter cancellation, close, and cleanup retry when those paths change.
+
+Bound the process family and retained state, not just a parent heap or worker
+count. Account for decoding and serialization expansion before allocation.
+Choose CPU, memory, concurrency, and disk limits from the host and workload;
+stop before OOM and verify owned cleanup. A guard-stopped task is incomplete.
+Do not replace a transport buffer cap with an unbounded retained inventory or
+hide partial results behind a success status.
+
+## Compare completed tasks and limit the claim
+
+Run the affected real public workflow before and after with equivalent inputs,
+results, coverage, and Evidence semantics. Include small inputs so setup costs
+and regressions are visible. Distinguish fresh sessions, warm runtime caches,
+and OS file caches. Alternate or repeat runs when variance affects the decision.
+Choose meaningful boundary regressions for cases the actual engine cannot
+reliably reproduce; a fake engine does not qualify a real provider.
+
+Report task outcome, relevant phase timings, total elapsed and CPU time, peak
+process-family memory and its accounting method, resource stops, and cleanup.
+Check the payload exit status as well as the profiler's. Keep failed attempts
+and identify unmeasured costs; a passing retry does not explain a failure.
+
+For token claims, count schemas/discovery, arguments, replies, retries, and
+continuations needed to finish the task. Label tokenizer estimates separately
+from actual model usage and record the client/catalog behavior. Smaller JSON
+alone establishes neither fewer model tokens nor better investigation quality.
+Do not claim a complex-software advantage without completed representative
+investigations and comparable correctness/coverage.
+
+Finish with the bottleneck, owning change, preserved semantics, executed
+checks, before/after measurements, and remaining limits. Keep any integration
+record current with its tested revision and distinguish merged production work
+from experiment-only code. Stop when the demonstrated bottleneck is resolved
+or further benefit remains speculative.

.agents/skills/rea-analysis-optimization/agents/openai.yaml

@@ -0,0 +1,4 @@
+interface:
+  display_name: "REA analysis optimization"
+  short_description: "Optimize analysis with preserved Evidence and measured costs"
+  default_prompt: "Use $rea-analysis-optimization to find and verify a concrete analysis bottleneck while preserving correctness and Evidence."

.agents/skills/rea-analysis-optimization/references/kuna-case-study.md

@@ -0,0 +1,112 @@
+# Kuna: reuse before adding machinery
+
+Read this reference for concrete examples of the skill's decisions. The records
+below concern static Linux analysis with caller-supplied Kuna. They do not
+establish parity with another provider or a general advantage for Code Mode,
+atomic tools, or any particular orchestration design.
+
+## Exact-address lookup: a measured, narrow improvement
+
+[PR #4695](https://github.com/morluto/rea/pull/4695) reused the validated string
+inventory's numeric address order. Previously, each `list_strings(address)`
+filtered all cached facts. The owner now performs a binary lower-bound lookup
+and returns every encoding at that address: O(log n + matches), without a
+second retained lookup index.
+
+The owning implementation is
+[`KunaStringInventory.ts`](../../../../src/kuna/KunaStringInventory.ts), consumed
+by [`KunaStringQueries.ts`](../../../../src/kuna/KunaStringQueries.ts). Inspect
+their current versions before adapting the example. Numeric u64 addresses,
+same-address encodings, duplicate rejection, source digests, caller selection,
+and the existing session lifetime were part of the contract. Full listings and
+literal/regex searches kept their semantics. An attempted removal of the
+duplicate-detection set was dropped: no stable memory benefit was established,
+and it delayed malformed-input rejection.
+
+### Recorded workload and results
+
+The benchmark used unmodified Kuna 1.833 and LLVM 16 on Linux x86-64, with
+602,544 string facts. Baseline was `be0b2d138`; the final feature was
+`5f564962da0e1b17f00a68a74a312539f0209a01`, subsequently integrated in
+`c6b62754b095a134d6d67877508fa34311a2eda2`. These are historical observations,
+not a benchmark of every subsequent main revision.
+
+The same production stdio MCP task discovered the catalog, opened the binary,
+made one cold selected read and 28 warm exact reads, checked retained Evidence,
+and closed the session. Five selected addresses and two absent selectors were
+read four times. A native filtered scan selected the strings; an independent
+ELF load-header mapping checked the source bytes and terminators. Returned
+facts, source linkage, target digests, and cleanup were checked.
+
+Three alternating fresh sessions per implementation used one CPU, a 4 GiB
+process-family hard limit, a 3 GiB early stop, and no swap. OS file caches were
+warm. Whole-harness costs include the independent native oracle, hashing,
+catalog/client work, and correctness checks.
+
+| Median                                   |     Before |      After |
+| ---------------------------------------- | ---------: | ---------: |
+| 28 warm exact MCP reads                  |   495.2 ms |    90.8 ms |
+| Cold selected MCP read                   |    11.30 s |    11.01 s |
+| Whole harness elapsed                    |    28.24 s |    27.43 s |
+| Whole process-family CPU                 |    27.71 s |    26.89 s |
+| Kernel family peak, including file cache | 2237.6 MiB | 2240.6 MiB |
+
+Warm totals ranged from 491.7–544.3 ms before to 84.8–99.5 ms after: about
+5.5 times faster at this phase. Cold/whole-task timing and memory overlapped;
+no stable improvement in those costs was established. One small SQLite pair
+(7,985 facts) measured 20 warm reads at 88.6 → 80.5 ms and the whole harness at
+13.68 → 13.94 s. The large-inventory lookup benefit does not generalize to a
+whole-task speedup on small inputs.
+
+Catalog hashes, call counts, and normalized facts were unchanged. A
+`cl100k_base` estimate over one pass of raw catalog JSON, arguments, and replies
+was about 846,000 tokens in both lanes, with about 802,562 from the catalog.
+This was transport serialization, not rendered model context or actual usage;
+client discovery, prompts, replay, and protocol framing were unmeasured.
+**No token reduction or improved agent investigation outcome was established.**
+
+Pinned target SHA-256:
+`f62d254b7f2bf42df8c8b07d46ee3bb4c2cafeca436b2e6bc6ccbe4581f58f40`.
+Pinned engine SHA-256:
+`d858918b0766435cd96b999a77f2d5a31a0cc4f07256421c883b5b9dc4dd58e4`.
+All six normalized-result sequence digests were
+`4d2f90269da8568984090b20274163c4a7c2a0538aab63530bd2fe44840a0bba`.
+Selected byte agreement does not prove native discovery completeness; missing
+strings and relationships can remain unknown.
+
+## Other transferable changes and their limits
+
+- [Cached string inventory and search, #4095](https://github.com/morluto/rea/pull/4095):
+  share validated facts across queries in a defined session. Audit key inputs
+  and lifetime before extending reuse across targets or engine revisions.
+- [Shared single-function work, #4045](https://github.com/morluto/rea/pull/4045)
+  and [repeat-request reuse, #4169](https://github.com/morluto/rea/pull/4169):
+  avoid duplicate native work while retaining per-caller cancellation and
+  cleanup ownership. A query count alone cannot reveal duplicate subprocesses.
+- [Streamed strings, #4318](https://github.com/morluto/rea/pull/4318)
+  and [streamed call graph, #4329](https://github.com/morluto/rea/pull/4329):
+  remove fixed aggregate transport-buffer failures through incremental parsing.
+  Streaming transport does not bound every retained object or final reply.
+  These links establish changes; no numerical speedup is asserted here.
+
+Typed reference facts, native string-query projection, and shared native byte
+reads remained experimental at the 2026-10-17 checkpoint
+`65f2f9a22bc892f61d2b00f9a4611124f57d4bff` on
+`experiment/atomic-codemode-20261016`. Some experiments used a patched native
+engine. Check current integration and bring-your-own engine capabilities before
+proposing them; a working experiment is not a stock-provider guarantee.
+
+## Applying the reasoning elsewhere
+
+For another provider or format, inspect its actual representation and native
+capabilities before borrowing the implementation. Existing numeric ordering
+may support range reads; an on-disk native query may avoid materializing an
+inventory; an immutable parse may support several views. Measure the relevant
+alternative rather than requiring every provider to use Kuna's cached array.
+
+For live observations, immutable artifact caching rules are insufficient:
+validate the process/session identity and freshness required by the caller.
+For compact replies, preserve access to source Evidence and report examined
+coverage; projection must not silently turn an unresolved edge into no edge.
+For batching, test mixed successes and failures and measure the entire task,
+including discovery, extra calls, native work, and serialization.

.agents/skills/reverse-engineer-anything/SKILL.md

@@ -2,7 +2,7 @@
 name: reverse-engineer-anything
 description: Reverse engineer native, managed, Electron/JavaScript, packaged, firmware, and browser targets with REA. Use shipped-artifact or requested runtime evidence to explain features, compare versions, decompile code, or guide a reconstruction. Skip REA for ordinary source-repository architecture analysis.
 metadata:
-  version: "42"
+  version: "45"
 ---
 
 # REA
@@ -30,6 +30,17 @@ unavailable, read [connection and recovery](references/connection-and-recovery.m
 Diagnose the observed blocker, continue useful supported work, and resume the
 original task after repair. Client approval, shell PATH, registration and engine
 availability have different remedies; avoid repeating setup for all of them.
+If the host times out without a REA result, do not repeat the unchanged call or
+assume work stopped. Read the recovery guide's deadline distinction before retrying.
+If the MCP connection drops, retry that call once only when it is safe to
+repeat or known not to have started. If it may already have changed the
+target, or it fails again, tell the user and ask them to restart the MCP
+client. If REA could not start its connection, retry once, then tell the user
+and ask them to restart the MCP client. Do not retry a startup failure that
+returns the same unchanged result. Run `rea doctor` when REA could not start
+or tools stay missing after that restart. For `cleanup_incomplete`, a retryable
+capacity limit, or `configuration_invalid`, follow the recovery guide's ladder
+before asking for a restart. Read that guide before a second repair attempt.
 
 1. Diagnose without changing REA or client configuration. `npx` may download
    the package and write to npm's cache. For Codex, run
@@ -89,9 +100,19 @@ explicit path or endpoint and do not need it.
 
 - ASAR or extracted JavaScript/Electron tree:
   `analyze_javascript_application`.
-- Archive/package member inventory (ZIP/APK/IPA/MSIX/AppX or DMG):
-  `open_binary` with the supplied local path; use `inspect_artifact` when its
-  graph and findings help answer the question.
+- Archive/package member inventory (ZIP/APK/IPA/MSIX/AppX, DMG, or a PyInstaller
+  executable): `open_binary` with the supplied local path; use `inspect_artifact`
+  when its graph and findings help answer the question. A PyInstaller CArchive
+  is the application layer; the bootloader is only the native prefix. Start with
+  `detail: "summary"` for large inventories, then use retained inventory views
+  for selected member fields/pages. A complete small inventory can cost less
+  than extra calls and schemas. Runtime options are metadata, not files.
+  Compact delivery still inventories and hashes members. See the
+  [Python packaged applications guide](https://github.com/morluto/rea/blob/main/docs/python-packaged-applications.md).
+- Apple app plist, Interface Builder, asset catalog, or keyed archive resources:
+  `open_bundle_resources`, then the selected resource tool. This does not
+  require `CFBundleExecutable` or a deep provider. Resource sessions have no
+  executable digest or binary snapshot; use `open_binary` for native code.
 - Android APK code, classes, methods, or incoming references:
   `inspect_android_package`, then focused Android tools when advertised.
   Archive member inventory still uses the archive route above.
@@ -114,7 +135,10 @@ explicit path or endpoint and do not need it.
   connect Everywhere's separate MCP server for live traffic.
 - Retained HAR/SAZ/native mitmproxy/aioquic qlog capture: `inspect_web_network_capture`. Select
   original record ordinals when useful; inspect producer fields and byte/number
-  sidecars without fetching recorded URLs or claiming live attribution. See
+  sidecars without fetching recorded URLs or claiming live attribution. HAR/SAZ/qlog
+  old-generation heap is `REA_CAPTURE_DECODER_OLD_GENERATION_MIB` (default 192 MiB,
+  floor 8, maximum 16384), set where REA starts. Host memory and `NODE_OPTIONS` do not
+  raise it. See
   [historical captures](https://github.com/morluto/rea/blob/main/docs/web-network-captures.md) for the exact upstream
   profile and credential exclusions. Qlog support starts with aioquic 1.5.0
   JSON 0.3; stream coverage and endpoint loss declarations do not recover bodies
@@ -154,8 +178,18 @@ explicit path or endpoint and do not need it.
   Select JavaScript glue and candidate paths explicitly and preserve URL/path
   ambiguity. No implicit fetch or target execution; WAT is not original source.
   See the [WASM guide](https://github.com/morluto/rea/blob/main/docs/wasm-artifacts.md).
+- Explicit local Android APK (or ZIP/AAB) signing: `inspect_apk_signature`.
+  Prefer its summary or selected signer/digest/lineage page; retain complete
+  Evidence and use `inspect_analysis_view` kind `apk-signature` for follow-ups.
+  Reports which schemes are present (v1 JAR, v2, v3, v3.1), each signer's
+  certificate SHA-256 fingerprint (v1 included, from its PKCS#7 block), v3 key-rotation lineage, source stamp, and
+  whether each signed digest still matches the bytes (self-consistency). This
+  is identity and integrity, not a trust or authenticity verdict.
+  See the [APK signature guide](https://github.com/morluto/rea/blob/main/docs/apk-signature.md).
 - Explicit local EVM bytecode carrier: `inspect_evm_interface` with caller-selected
-  `raw` or `hex` encoding. Preserve carrier/decoded digests and treat selectors,
+  `raw` or `hex` encoding. The owned worker deadline is `timeout_ms` (default
+  30000 ms, range 1..2147483647, next call). It does not raise the fixed byte,
+  address-space, or CPU allowances. Preserve carrier/decoded digests and treat selectors,
   argument strings and mutability as inferred candidates; this performs no chain
   lookup or target execution. For the complete inferred interface without hex or
   duplicate stdout, discover support and select `detail: "summary"`; complete
@@ -169,14 +203,69 @@ explicit path or endpoint and do not need it.
   absent metadata does not prove a binary is not Go, and unparsed text records
   remain explicit. See the
   [Go binary guide](https://github.com/morluto/rea/blob/main/docs/go-binaries.md).
+- Embedded CUDA/HIP code objects: `inspect_gpu_code_objects`; inventory targets
+  first, then decode selected payloads and query retained GPU Evidence.
+- Bundled GGUF or safetensors model declarations: `inspect_model_artifact`
+  with the explicit local file. Default header hashing leaves full-file identity
+  unknown; `hash: "full"` additionally reads all weights for SHA-256. MCP returns
+  retained summaries; optional `view` selects one page in the initial call. Use
+  `inspect_analysis_view` kind `model-artifact` for
+  batched metadata keys, tensor names or type totals, selected fields and pages.
+  Preserve header-scoped indices and array preview omissions; do not infer model
+  behavior from declared architecture or tensor types. See the
+  [model file guide](https://github.com/morluto/rea/blob/main/docs/model-artifacts.md).
+- Unity IL2CPP names, member owners and string literals from `global-metadata.dat`:
+  `inspect_il2cpp_metadata` with the explicit local file; MCP defaults to compact
+  counts. Query sections, images, types, methods, fields, properties, events,
+  parameters or string_literals through `inspect_analysis_view`
+  with the returned `parent_evidence_id`, using exact filters, batch indices and
+  selected fields. Parameters accept `method_indices`, avoiding a scan of all
+  parameter pages. Missing collections in historical Evidence remain unknown.
+  Versions 24, 27, 29, 31 and 39; no Unity, IL2CPP
+  binary or disassembler is run. Unrecognized magic (possibly encrypted) and
+  other versions are unsupported. Optional matching `binary_path` enables static native addresses and type references; query `method_addresses` with indices or addresses/states, preserving every shared-body method and both artifact identities. Without a linked binary native counts remain unknown. Byte-pattern fallback is inferred; no runtime behavior or same-build pairing is proved. See the
+  [IL2CPP metadata guide](https://github.com/morluto/rea/blob/main/docs/il2cpp-metadata.md).
 - Embedded cargo-auditable Rust dependency metadata: `inspect_rust_binary`
   with the explicit local ELF, PE, or thin Mach-O file. This built-in
   target-free inspection requires no Rust installation or disassembler. Preserve
   package order, dependency indices, and the section file range. A missing
   `.dep-v0` section is null metadata, not proof the binary has no dependencies.
   A rustc commit hash is an inference from a read-only `/rustc/<40-hex>/` path.
+  For selected crates, use `view: {collection: "packages", names: [...],
+  fields: ["name", "version", "source"]}` on the first call; explicitly include
+  `dependencies` for direct index references. Use `detail: "summary"` for counts
+  and ranges, or retained `inspect_analysis_view` Rust package/edge pages for
+  subsequent filters. Keep the complete parent Evidence accessible; CLI compact
+  delivery requires `--evidence_output`.
   See the
   [Rust binary guide](https://github.com/morluto/rea/blob/main/docs/rust-binaries.md).
+- PE Authenticode certificate table: `inspect_pe_certificates` with the explicit
+  local PE32/PE32+ file. This built-in target-free inspection requires no
+  Windows SDK. Preserve WIN_CERTIFICATE ranges, types, payload SHA-256, and
+  PKCS#7 signer/X.509 identity fields, including `subject_key_identifier_hex`
+  when the signer uses that form. A missing security directory is observed
+  absence of that table, not proof the image is unsigned in every catalog sense.
+  Signatures are not verified. See the
+  [PE certificate guide](https://github.com/morluto/rea/blob/main/docs/pe-certificates.md).
+- ELF note tables: `inspect_elf_notes` with the explicit local ELF32/ELF64
+  file. This built-in target-free inspection requires no disassembler. Preserve
+  PT_NOTE and SHT_NOTE ranges, owners, types, descriptor SHA-256, and GNU
+  build-id or ABI-tag fields. A missing table is observed absence of those
+  declarations, not proof the image never carried notes. See the
+  [ELF notes guide](https://github.com/morluto/rea/blob/main/docs/elf-notes.md).
+- CPython 3.7–3.13 `.pyc` header, nested marshal code objects and bytecode
+  identity: `inspect_python_pyc` with the explicit local file. This built-in
+  target-free inspection requires no Python installation. Preserve magic,
+  PEP 552 hash or mtime/size, nested names, and bytecode SHA-256. Bytecode is
+  not decompiled, and a decoded header does not prove the matching `.py` is
+  present. See the
+  [CPython pyc guide](https://github.com/morluto/rea/blob/main/docs/python-pyc.md).
+- Erlang BEAM IFF container: `inspect_erlang_beam` with the explicit local
+  `.beam` file. This built-in target-free inspection requires no Erlang
+  installation. Preserve chunk names, atom order, export/import/lambda tables,
+  and the Code header. Missing optional tables stay null. Bytecode is not
+  executed. See the
+  [Erlang BEAM guide](https://github.com/morluto/rea/blob/main/docs/erlang-beam.md).
 
 - Explicit local SQLite database snapshot: `inspect_sqlite_database`. Inspect
   with Node.js 24.x >=24.15 or 26+ providing native SQLite limits. Read schema
@@ -199,6 +288,8 @@ explicit path or endpoint and do not need it.
 - PE resource type/name/language trees and static icon group references:
   `inspect_pe_resources` with an explicit path. This portable operation requires
   no active target or engine; opaque payloads retain file ranges and SHA-256.
+- PE export directories and ELF dynamic-symbol exports: `inspect_native_exports`
+  with an explicit path; portable, no active target or engine.
 - Native executable, library, or analysis database: `open_binary`, then
   use focused analysis tools directly; call `binary_overview` when metadata or
   inventory context is useful and available from the selected provider.
@@ -217,7 +308,12 @@ facets remain unknown; live IDA results are not replayed from snapshots.
 
 With `REA_KUNA_COMMAND` configured, `provider_id: "kuna"` answers procedure
 lists, lookups, pseudocode, assembly and `analyze_function` from Kuna's cached
-whole-program export; it has no cross-references, callers or CFG. A cold first
+whole-program export. Callers/callees use a separate body-free graph shared by
+the session; `xrefs` returns exact-address incoming sites and filters forwarding
+aliases. These queries and inventory lookups do not trigger a C export. Indirect
+targets remain unknown and can make relationship coverage partial. Reference
+caches expire on close; a fresh CLI process repeats the native reference walk.
+Typed references, CFG, raw-memory tools and annotations remain unsupported. A cold first
 answer can come from a single-function run (its limitations say so); assembly
 waits for the export. For pseudocode and assembly, `raw_result.source` names the
 cached `.c`/`.asm` file and byte span, which you may search directly instead of
@@ -242,6 +338,12 @@ Provider candidate facts describe possible providers, while top-level session
 capabilities describe the current route and `analysis_provider_binding` names the
 selected provider. Do not treat those scopes as interchangeable.
 
+For a Ghidra startup timeout on a large native target, follow
+[deadline recovery](references/native-and-artifacts.md#ghidra-startup-deadline-recovery)
+before abandoning deep analysis or repeating the same cold import. The default
+330 seconds is configurable; changing the server's environment requires a new
+REA process.
+
 ## Work summary-first
 
 Start with the default result and use its inline Evidence and graph context.

.agents/skills/reverse-engineer-anything/references/android-applications.md

@@ -24,7 +24,10 @@ matching bundled native controls and uses an owned Job Object with bidirectional
 stdio. A full JDK, including `jdk.compiler`, is required on every host; Windows
 `JAVA_HOME` selects `bin\java.exe`. Consult the Android guide for actual host
 verification coverage. Calls use one worker and provider deadlines; optional
-`REA_JADX_HEAP_MIB` selects the JVM heap. The APK is never launched,
+`REA_JADX_HEAP_MIB` selects the JVM heap and `REA_JADX_OPERATION_TIMEOUT_MS`
+the per-operation deadline (integer milliseconds from 1 to 2147483647, default
+120000; a change applies on the next process start; the MCP client's tool
+timeout must also allow it). The APK is never launched,
 and no emulator is required.
 
 Manifest entries and decompiled text are static observations, not execution or

.agents/skills/reverse-engineer-anything/references/connection-and-recovery.md

@@ -11,14 +11,15 @@ main and an older release expose the same arguments.
 
 ## Choose the remedy from the observed failure
 
-| Observation                                                     | Meaning and next step                                                                                                                                                                                                                                                                                                 |
-| --------------------------------------------------------------- | --------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- |
-| No REA tools in the active session                              | Inspect registration with `rea doctor --client codex --json`. A skill installation alone does not register MCP.                                                                                                                                                                                                       |
-| Missing, malformed or stale registration                        | Prepare a scoped setup plan; apply only authorized changes with backups.                                                                                                                                                                                                                                              |
-| Aligned registration, but no connected tools                    | Restart/reconnect the client. Doctor checks registration; it cannot establish the current agent's connection. Inspect the client's launch error if reconnection fails.                                                                                                                                                |
-| `MCP tool call requires approval, but approval policy is never` | Codex rejected the call before REA executed it. Do not install an engine or repeat setup to solve this client-policy decision. See the approval section below.                                                                                                                                                        |
-| REA returns `provider_unavailable` or an unavailable tool entry | Read that operation's provider diagnostics or `binary_session.result.tool_availability`. Expand each group's tool names and inherit the shared client requirements unless overridden. Use the operation's availability, reason, and remediation; a complete tool catalog does not imply every operation is available. |
-| `rea` is missing in a child shell, or Node is incompatible      | Inspect the actual shell PATH and launcher/runtime. A prefix installation may be absent from a login shell's PATH; this does not establish a broken MCP registration.                                                                                                                                                 |
+| Observation                                                     | Meaning and next step                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                   |
+| --------------------------------------------------------------- | --------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- |
+| No REA tools in the active session                              | Inspect registration with `rea doctor --client codex --json`. A skill installation alone does not register MCP.                                                                                                                                                                                                                                                                                                                                                                                                                                                                         |
+| Missing, malformed or stale registration                        | Prepare a scoped setup plan; apply only authorized changes with backups.                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                |
+| Aligned registration, but no connected tools                    | Restart/reconnect the client. Doctor checks registration; it cannot establish the current agent's connection. Inspect the client's launch error if reconnection fails.                                                                                                                                                                                                                                                                                                                                                                                                                  |
+| MCP connection dropped or REA could not start                   | If the connection drops, retry that call once only when it is safe to repeat or known not to have started. If it may already have changed the target, or the same failure returns, tell the user REA's connection failed and ask them to restart the MCP client. If REA could not start its connection, retry once, then tell the user and ask them to restart the MCP client. Do not retry a startup failure that returns the same unchanged result. Run `rea doctor` when REA could not start or tools stay missing after restart. Doctor cannot restore a live connection by itself. |
+| `MCP tool call requires approval, but approval policy is never` | Codex rejected the call before REA executed it. Do not install an engine or repeat setup to solve this client-policy decision. See the approval section below.                                                                                                                                                                                                                                                                                                                                                                                                                          |
+| REA returns `provider_unavailable` or an unavailable tool entry | Read that operation's provider diagnostics or `binary_session.result.tool_availability`. Expand each group's tool names and inherit the shared client requirements unless overridden. Use the operation's availability, reason, and remediation; a complete tool catalog does not imply every operation is available.                                                                                                                                                                                                                                                                   |
+| `rea` is missing in a child shell, or Node is incompatible      | Inspect the actual shell PATH and launcher/runtime. A prefix installation may be absent from a login shell's PATH; this does not establish a broken MCP registration.                                                                                                                                                                                                                                                                                                                                                                                                                   |
 
 If the client cannot launch its shell/sandbox helper or read the installed skill,
 inspect its launch diagnostic. For example, Codex can refuse helper aliases when
@@ -34,6 +35,83 @@ answers part of the question, and name the unavailable evidence separately.
 Do not substitute another target. Label observations from ordinary shell tools
 as such; only actual REA output supplies REA Evidence.
 
+## Distinguish client and server deadlines
+
+A host message such as `timed out awaiting tools/call` with no REA result means
+the client's deadline expired before a result arrived. It does not establish
+that REA finished, canceled successfully, or released all resources. Do not
+repeat the unchanged analysis. Inspect available session/provider state and
+retained Evidence only when useful; an idle `binary_session` is not proof that
+every target-free operation stopped. Never invent a missing Evidence reference.
+
+Allow the host tool deadline to cover execution, cancellation/cleanup, and
+response delivery. In Codex, the setting is
+`mcp_servers.<server-id>.tool_timeout_sec` in **seconds**, not
+`startup_timeout_sec`. Select a value for the actual operation budget and
+headroom, not a universal timeout. See the
+[official MCP configuration reference](https://learn.chatgpt.com/docs/extend/mcp?surface=cli).
+Apply configuration changes only within the authorized scope, preserving
+unrelated entries and backups; reconnect/restart as required by the host.
+When configuration is not editable, report the setting needed or use an
+authorized CLI fallback. Do not silently broaden permissions or switch targets.
+
+A returned REA `provider_timeout` instead means the **server/provider** reached
+its deadline. Follow its `remediation.action`, reuse accessible partial Evidence,
+and resolve reported incomplete cleanup before retrying. Change a request
+field or a `REA_*` setting only when the returned remediation explicitly
+recommends that change. Restart REA when that text says the value is read at
+startup. A mention that a control does not raise the deadline is not a
+recommendation to change it. Do not invent a setting the remediation does not
+name. Raising only the client deadline does not change
+this server budget. For static JavaScript,
+`analysis_timeout_ms` (CLI `rea analyze-javascript-application <path>
+--analysis-timeout-ms <milliseconds>`) controls each source's
+startup, analysis and transfer deadline in **milliseconds**. Increase that
+budget or reduce semantic work / select exact inventoried `source_paths` when
+the task permits. Omit `source_paths` for all sources; `[]` is invalid.
+`detail: "summary"` reduces delivery, not analysis work. A per-source budget
+does not bound the elapsed time for an entire multi-source application.
+Partial-Evidence inspection can avoid recomputation; an unavailable reader is
+not itself a reason to block a corrected static-analysis retry.
+The shorthand `rea analyze <path>` does not expose this timeout flag; use the
+explicit JavaScript command when changing that budget.
+
+For invalid arguments, use the connected tool's current input schema, or the
+installed CLI's command help. Refresh the catalog if stale; do not guess fields
+from an older skill or return the entire catalog to repair one field. Correct
+all reported fields before retrying.
+
+## Recovery ladder
+
+Follow the returned remediation. Change a request field or a `REA_*` setting
+only when that text explicitly recommends the change. Restart REA when it says
+the value is read at startup. A mention that a control does not raise a limit
+is not a recommendation to change it. Do not invent a setting. Host memory and
+`NODE_OPTIONS` do not raise a REA admission unless that remediation says so.
+
+- MCP connection dropped: retry that call once only when it is safe to repeat
+  or known not to have started. If it may already have changed the target, or
+  it fails again, tell the user REA's connection failed and ask them to restart
+  the MCP client. Do not run `rea doctor` for the drop itself.
+- REA could not start its connection: retry once, then tell the user and ask
+  them to restart the MCP client. Run `rea doctor` when REA could not start or
+  tools stay missing after that restart. Do not retry a startup failure that
+  returns the same unchanged result.
+- `cleanup_incomplete`, or retained cleanup owners in the error details: retry
+  that exact cleanup owner. Do not start new work that needs those resources
+  until that owner confirms release. Repeating the original call is not that
+  retry.
+- Capacity, when the result is `resource_constraint` and `retryable` is true:
+  wait for the active admission to finish, then retry within the reported
+  limits. Do not restart REA to free the slot.
+- `configuration_invalid`: correct the named setting in the environment that
+  starts REA (the shell or the MCP registration), then restart REA. Run
+  `rea doctor` to see the `runtime-configuration:` check. Re-run `rea setup`
+  for that client when the registration must carry the new setting, then
+  reconnect.
+- `invalid_request`: correct the reported fields from the current schema. Do
+  not repeat the unchanged call.
+
 ## Repair registration when needed
 
 For Codex:

.agents/skills/reverse-engineer-anything/references/evidence-workflows.md

@@ -24,6 +24,17 @@ retain its complete source and receive that view immediately. Without that
 advertised choice, keep the existing complete call. CLI inspection returns
 portable complete Evidence; save it before projecting a selected inline view.
 
+For native procedure/name/string inventories, select `detail: "summary"` when
+the connected tool advertises it and the task needs counts or selected rows.
+Use its parent Evidence ID with
+`view: {kind: "native-inventory", offset: 0, limit: 32, fields: ["value"]}`.
+Search tools retain the existing pattern/mode/case/document selection. Omit
+`fields` for complete records; `record.address` and `source_index` remain on
+selected rows. Selecting `value` keeps its string/decoding qualifications.
+Parent Evidence ID plus source index preserves same-address
+alternatives. Follow `coverage.next_offset` only as far as the question needs;
+exhausting retained rows does not establish complete program discovery.
+
 Use `record_unknown` to track unresolved questions in the user's investigation
 and name the authority or environment still required. Supply supporting and
 contradicting evidence IDs. It updates the session registry directly.
@@ -69,3 +80,27 @@ When the connected schema advertises `include_query_parameter_names`, select
 it in the view for distinct observed names with original record ordinals. Query
 values are omitted; the names establish neither required parameters nor a
 complete server schema. Keep the original parent Evidence for exact URLs.
+
+Use `inspect_analysis_view` with `view: {kind: "environment-manifest"}` on the
+same retained HAR Evidence to project the primary origin's request lines into
+named tools with path templates, parameters, observed query names, value-free
+body shapes and record pointers for an agent environment. Opt into
+`template_rules` only when recorded identifiers need them; every template,
+parameter and merged sibling is an inference, no header, cookie or body value is
+copied, and the three stated unknowns stay unresolved.
+
+Use `view: {kind: "environment-state"}` with the manifest's `template_rules` to
+extract the data those reads observed instead: entity types and identity
+fields, entities with recorded values, documents, and per-tool render plans
+checked against the recordings. The state comes only from reads before the
+first related write; types, identity and list membership are inferences. The
+view also gives each state-changing tool a write plan, served only when a later
+recorded read confirms its effect and none contradicts it.
+`rea environment-serve manifest.json --evidence capture.json` derives that
+state at startup, answers read tools from it, serves the write tools whose plan
+is served, each applying its effect to that state, and serves `data/reset`,
+`data/add` and `data/get` on `/agentenv`; other write tools are not served, and
+each plan states why in its `reason`. It replays the capture instead with
+`--replay-only` or when the state is unavailable (not derivable, or a catalog
+over the discovery budget), and its stopped document's `state` says which
+applied.

.agents/skills/reverse-engineer-anything/references/javascript-applications.md

@@ -18,9 +18,15 @@ only selected files. The result and summary record `source_selection`; files
 outside scope and their relationships remain unknown, so partial application
 coverage does not mean the selected source failed. Use repeated `--source-path`
 on either static application CLI route. Omit the selection when the question
-requires complete application analysis. `detail: "summary"` reduces delivery
+requires complete application analysis; `source_paths: []` is invalid, not an
+unfiltered request. `detail: "summary"` reduces delivery
 and does not by itself reduce analysis work.
 
+ASAR header/index admission defaults to 16 MiB / 16777216 bytes;
+`REA_ASAR_MAX_HEADER_BYTES` (at most 256 MiB, set where REA starts) changes it.
+`source_paths`, `detail`, and `max_heap_mb` do not bypass or raise it. Or
+analyze an extracted directory.
+
 Exact `#alias` specifiers resolve through the importer's
 enclosing package `imports` map to package-local files; nested `node_modules`
 packages use their own map. Wildcard, recursive, and external-package import
@@ -35,7 +41,9 @@ the complete application Evidence inline when it fits.
 When the connected schema advertises `max_heap_mb` and `analysis_timeout_ms`,
 these select the isolated source analyzer's old-space heap and per-source
 deadline. Defaults are 1024 MiB and 300000 ms. They do not change the MCP
-process heap. A Bun-hosted REA does not enforce the heap choice; on Windows it
+process heap or the client's tool deadline. For a host timeout without REA
+output, use the [deadline recovery guide](connection-and-recovery.md#distinguish-client-and-server-deadlines)
+instead of raising the server budget. A Bun-hosted REA does not enforce the heap choice; on Windows it
 instead limits the analyzer's committed memory to that choice plus a 512 MiB
 runtime reserve.
 Resource-limited analyses preserve completed static facts and
@@ -51,6 +59,20 @@ in the same session. A returned `error.details.partial_observation` reference
 can be supplied directly to `inspect_analysis_view`. If the SDK cancels locally, retrieve its recorded ID from
 the Evidence bundle before inspecting the partial analysis.
 
+## Inventory route candidates without returning the application graph
+
+When the advertised schema supports `view`, use `detail: "summary"` with
+`view: {kind: "javascript-routes", collection: "routes", fields: ["path", "confidence"], limit: 20}`.
+The same view works with `analyze_web_bundle` on already selected captured sources.
+The complete source Evidence is retained; use its `parent_evidence_id` with
+`inspect_analysis_view` for `summary`, `routes`, `occurrences`, `sources`, or
+`unknowns`. Filter and batch IDs before requesting more fields, and follow
+`coverage.next_offset`. CLI selections require `--evidence_output` to keep the
+complete record. Static patterns, inferred deployment bases, heuristic system
+classification, and unavailable sources never establish route reachability or
+the complete application route denominator. See the
+[route inventory guide](https://github.com/morluto/rea/blob/main/docs/javascript-route-inventory.md).
+
 ## Read a module without repeating analysis
 
 Use the actual returned parent ID, on the same connection, as `source.evidence_id`.
@@ -94,6 +116,36 @@ recorded observations, identity, hashes, exports and source ranges; it does not
 include every child syntax observation, related edge or original source file.
 An absent item fact is not evidence of absence from the whole application.
 
+For canonical identities and retained source-range counts, use a typed reference
+page instead of reading every full item:
+
+```json
+{
+  "source": {
+    "kind": "retained-evidence",
+    "evidence_id": "ev_ACTUAL_PARENT_ID"
+  },
+  "view": {
+    "kind": "javascript",
+    "collection": "modules",
+    "projection": "references",
+    "where": { "path": "main.js", "kind": "javascript-module" },
+    "offset": 0,
+    "limit": 16
+  }
+}
+```
+
+Filters match exact retained strings; omit a filter to include all values, use
+`path: null` for unavailable paths, and `path: ""` for a recorded empty path.
+Rows preserve canonical `identity` and distinct observation support states in
+`basis`. `retained_source_range_count` counts retained source-range observations,
+not runtime imports or unique reference sites. Use a row's `node_id` with the item
+view above for exact locations and full supporting observations. Follow
+`coverage.next_offset`; page exhaustion describes matching retained rows only.
+Check `source_coverage` for source selection, partial/unknown graph coverage and
+unknown counts before drawing a negative conclusion.
+
 ## Trace only the remaining question
 
 Use `trace_application_feature` on the parent application Evidence for one
@@ -165,7 +217,8 @@ accept complete inline Evidence or
 `{"kind":"retained-evidence","evidence_id":"ev_<64 lowercase hex characters>"}`
 for their application input (`application`, or `left`/`right`). `inspect_analysis_view`
 uses the same retained-reference form in `source`, or portable inline Evidence,
-to project a summary, module page, or one module with its recorded observations. This notation is
+to project a summary, module page, or one module with its recorded observations.
+This notation is
 a template: replace it with the actual returned ID. Versions before 4.1.0 accept only full
 inline Evidence. Use the exact ID
 returned by the producer on the same MCP connection. Resolution does not run
@@ -173,3 +226,15 @@ analysis or select a provider; findings remain inline. Native observation
 arrays still take full Evidence. `close_binary` clears retained references;
 export a bundle before closing, or supply portable inline Evidence in another
 connection. A missing reference includes its ID and recovery guidance.
+
+For exact file/native metadata or processing gaps, use
+`view: {kind: "javascript-facts", where: {paths: ["package.json"]}, property_keys: ["name", "operation", "analysis_status"], limit: 32}`.
+Rows retain canonical node/observation identity and original Evidence. Paths
+match Evidence artifact-path/source-range locations; missing properties remain
+omitted and source coverage stays separate from page exhaustion. Batch paths and
+select only needed property keys; complete Evidence remains accessible by export.
+If names are unknown, opt into `include_available_property_keys: true`; it lists
+the original keys per returned observation without unselected values. Keep each
+row's node/observation identity paired when composing facts.
+These rows select application-graph observations; semantic relationships and
+semantic unknown records remain in the complete parent Evidence and trace tools.

.agents/skills/reverse-engineer-anything/references/native-and-artifacts.md

@@ -1,8 +1,52 @@
 # Native, managed, and packaged artifacts
 
+For release comparisons, export `list_procedures` Evidence from each target before
+closing it. Close clears session records: import both bundles after collecting the
+two targets, or supply portable inline Evidence. `match_native_functions` pairs unique exact names and, when supplied,
+complete observed body bytes. Reuse optional same-context `list_names`,
+`procedure_info`, `read_bytes` and existing `analyze_function` Evidence; avoid
+decompiling every function merely to match it. Its summary retains a parent for
+`inspect_analysis_view` `function-matches` collections `matches` and `functions`.
+Select columns, batch IDs and page individual facts; aliases remain unresolved.
+Unpaired observations do not prove additions/removals. Existing matched dossier
+pairs include inputs for `compare_functions`; otherwise analyze just chosen pairs.
+
+## Apple app resources
+
+Use `open_bundle_resources` with the selected app path, then `inspect_plist`,
+`decode_interface_builder`, `inspect_asset_catalog`, or `inspect_keyed_archive`.
+The resource session reads the app without resolving `CFBundleExecutable` or
+starting a deep provider. Preserve `resource_root` provenance, per-resource
+hashes where reported, and null executable identity; a directory path is not
+an executable digest. Export retained Evidence before closing or switching
+roots. Binary analysis snapshots do not apply. Use `open_binary` to switch to
+executable analysis, and `close_binary` to release either session kind.
+
+The four matching CLI commands accept the app path directly. For an archive
+inside the app, use `inspect-keyed-archive APP_PATH RELATIVE_ARCHIVE_PATH`.
+Host requirements for plist and compiled asset/Interface Builder tooling still
+apply; a resource-only session does not establish provider availability.
+
+## Desktop installation directories
+
+When an installation directory is supplied and its runtime is unknown, use
+`inspect_desktop_application` with its absolute path. Default MCP summary
+delivery retains complete Evidence. Read `files`, `omissions`, `candidates`, `components` or
+`next_probes` with `inspect_analysis_view` kind `desktop-application`, selecting
+exact paths, batched file/omission IDs or families and only needed fields and pages.
+Summary `probe_issues` retains counts and bounded previews. Select files with
+`has_issues: true` and fields `path`, `read_status`, `issues` to retrieve exact
+failures directly; preview text is display data, while `file_id` selects full facts. Preserve source
+coverage, read-range identity and unknowns; directory snapshot identity is not
+a digest of every file's contents. Component path markers do not establish
+runtime presence. For compact CLI output, use
+`inspect-desktop-application PATH --detail summary --evidence_output BUNDLE`.
+See [desktop installations](https://github.com/morluto/rea/blob/main/docs/desktop-applications.md).
+
 ## Go build metadata
 
-Use `inspect_go_binary` for an explicit local ELF, PE or thin Mach-O when the
+Use `inspect_go_binary` for an explicit local ELF, PE or Mach-O (universal
+files need `architecture`) when the
 question concerns the embedded compiler version, module/dependency versions,
 local replacements or build settings. No `open_binary`, Go installation or
 target execution is required. The CLI equivalent is `inspect-go-binary PATH`.
@@ -11,9 +55,38 @@ unparsed module-text line. A missing record is an unknown, not proof of a
 non-Go binary; observed metadata does not authenticate its claims. Function,
 source and type recovery remain a separate native-analysis question.
 
+## Model files
+
+Use `inspect_model_artifact` (`inspect-model-artifact PATH`) for local GGUF v2/v3
+and safetensors declarations without a model runtime. Header-only mode does not
+read weights or establish full-file identity. `hash: "full"` explicitly streams
+all bytes. Retained `model-artifact` views page `metadata`, `tensors` and derived
+`types`, with exact batch filters and selected fields. Arrays retain typed counts
+and eight-element previews; omitted values, unknown types and runtime behavior
+stay explicit. See [model files](https://github.com/morluto/rea/blob/main/docs/model-artifacts.md).
+
+## IL2CPP metadata
+
+Use `inspect_il2cpp_metadata` for an explicit local Unity `global-metadata.dat`
+(versions 24, 27, 29, 31 and 39) when the question concerns assembly images,
+names, member owners, string literals, namespaces or index ranges. No `open_binary`, Unity installation or target
+execution is required. The CLI equivalent is `inspect-il2cpp-metadata PATH`.
+MCP defaults to counts with complete retained Evidence. Query `sections`,
+`images`, `types`, `methods`, `fields`, `properties`, `events`, `parameters` or
+`string_literals` through `inspect_analysis_view`; use exact names,
+batch indices and selected fields. Row indices are scoped by artifact SHA-256
+and collection; page offsets count filtered matches. Parameters support batch
+`method_indices` to read one method's parameters directly. Historical reports
+without member collections retain unknown counts rather than empty inventories. Preserve
+artifact SHA-256 and index widths. Unrecognized magic may mean encryption and
+is unsupported, not decrypted. Linking names to `GameAssembly` or
+`UnityFramework` code addresses remains a separate native-analysis question.
+See [IL2CPP metadata](https://github.com/morluto/rea/blob/main/docs/il2cpp-metadata.md).
+
 ## Rust dependency metadata
 
-Use `inspect_rust_binary` for an explicit local ELF, PE, or thin Mach-O when
+Use `inspect_rust_binary` for an explicit local ELF, PE, or Mach-O (universal
+files need `architecture`) when
 the question concerns the cargo-auditable crate list embedded in `.dep-v0`
 (Mach-O: `__DATA,.dep-v0`). No `open_binary`, Rust installation, or target
 execution is required. The CLI equivalent is `inspect-rust-binary PATH`.
@@ -23,6 +96,51 @@ dependencies. An omitted package kind stays null. A rustc commit hash is
 inferred from a read-only `/rustc/<40-hex>/` path and does not by itself
 identify the compiler. See [Rust binaries](https://github.com/morluto/rea/blob/main/docs/rust-binaries.md).
 
+## PE Authenticode certificate tables
+
+Use `inspect_pe_certificates` for an explicit local PE32/PE32+ image when the
+question concerns the Authenticode certificate table, WIN_CERTIFICATE types, or
+embedded PKCS#7 signer and X.509 identity fields. No `open_binary`, Windows SDK,
+or target execution is required. The CLI equivalent is
+`inspect-pe-certificates PATH`. Preserve artifact SHA-256, table file ranges,
+payload hashes, original validity-time spellings, and
+`subject_key_identifier_hex` when the signer uses that form. A missing
+`IMAGE_DIRECTORY_ENTRY_SECURITY` directory is observed absence of that table,
+not proof the image is unsigned in every catalog or overlay sense. Signatures,
+chains, timestamps, and Authenticode digest coverage remain unobserved. See
+[PE Authenticode certificate tables](https://github.com/morluto/rea/blob/main/docs/pe-certificates.md).
+
+## ELF note tables
+
+Use `inspect_elf_notes` for an explicit local ELF32/ELF64 image when the
+question concerns PT_NOTE/SHT_NOTE owners, GNU build-id, or GNU ABI-tag
+fields. No `open_binary` or target execution is required. The CLI equivalent
+is `inspect-elf-notes PATH`. Preserve artifact SHA-256, table file ranges,
+descriptor hashes, and original owner spellings. A missing table is observed
+absence of those declarations, not proof the image never carried notes in
+another container. See
+[ELF note tables](https://github.com/morluto/rea/blob/main/docs/elf-notes.md).
+
+## CPython pyc files
+
+Use `inspect_python_pyc` for an explicit local CPython 3.7–3.13 `.pyc` when the
+question concerns the header magic, PEP 552 source hash or timestamp, nested
+code-object names, or bytecode identity. No `open_binary`, Python installation,
+or target execution is required. The CLI equivalent is `inspect-python-pyc PATH`.
+Preserve artifact SHA-256, the 16-byte header fields, nested code names, and
+bytecode SHA-256. Bytecode is not decompiled. A decoded header does not prove
+the matching `.py` is present. See [CPython pyc inspection](https://github.com/morluto/rea/blob/main/docs/python-pyc.md).
+
+## Erlang BEAM files
+
+Use `inspect_erlang_beam` for an explicit local `FOR1`/`BEAM` file when the
+question concerns IFF chunks, atom names, exports, imports, lambdas, or the
+Code header. No `open_binary`, Erlang installation, or bytecode execution is
+required. The CLI equivalent is `inspect-erlang-beam PATH`. Preserve artifact
+SHA-256, chunk file ranges, 1-based atom indices, and original atom spellings.
+Missing optional tables stay null. See
+[Erlang BEAM files](https://github.com/morluto/rea/blob/main/docs/erlang-beam.md).
+
 ## SQLite snapshots
 
 Use `inspect_sqlite_database` with an explicit local database snapshot for its
@@ -77,6 +195,30 @@ and tool profile. Text offsets count UTF-16 code units; follow `next_offset`.
 The complete source remains retained after an oversized MCP reply. No target
 execution or network acquisition is implied, and WAT is decoded representation.
 
+## Offline APK signatures
+
+Use `inspect_apk_signature` for a selected local APK/ZIP/AAB. Default MCP summaries
+retain complete Evidence. Select `view.collection: signers/digests/lineage`,
+with scheme or ID batches, fields and paging; use `inspect_analysis_view` kind
+`apk-signature` for retained reads after deletion. IDs retain artifact/scheme
+scope and signer links; `fields: []` keeps identity links. Empty pages preserve
+partial decoding reasons and unknown authenticity. CLI compact delivery needs
+`--evidence_output` to preserve the complete parent.
+It reads the APK
+Signing Block (v2/v3/v3.1) and v1 `META-INF` signing without executing anything,
+and reports each signer's certificate facts (for v1, the certificate the PKCS#7
+block's SignerInfo names), the declared v3 rotation lineage,
+the declared source-stamp certificate and, for SHA-256/SHA-512 digests, whether
+the declared content digest matches the current bytes. That match is not a
+signature verification: anyone who changes the bytes can recompute the digest.
+`posture` summarises v1-only signing (which cannot tell a v1-only APK from one
+whose v2/v3 block was removed), observed signer agreement across schemes,
+whether the v1 signer equals the block-scheme signer (`false` is notable),
+whether every differing signer appears in a declared (unverified) rotation
+lineage, and the strongest scheme.
+It never claims authenticity: no signature or lineage is cryptographically
+verified.
+
 ## Native targets
 
 After `open_binary`, use focused search, procedure, or function tools directly.
@@ -124,6 +266,85 @@ operation's full ordered list. Candidate capabilities describe possible
 providers. Top-level session capabilities describe the current route, and
 `analysis_provider_binding` identifies the selected provider.
 
+## Ghidra startup deadline recovery
+
+`provider_timeout` during the first Ghidra query can cover import, auto-analysis,
+bridge startup, and readiness. The default 330,000 ms is configurable through
+`REA_GHIDRA_STARTUP_TIMEOUT_MS`; a smaller query still needs the same cold import.
+
+Preserve useful Evidence before closing the session or restarting REA. Check
+`close_binary` and resolve reported cleanup failures. Set a larger startup budget
+in the environment that launches REA, then restart REA and reconnect the MCP
+client before reopening the same target. Closing and reopening a binary in the
+existing server keeps its original configuration and starts import again.
+Environment changes in an agent's shell do not update an already running MCP
+server. If its launch configuration cannot be changed, report that constraint
+and the exact retry setting instead of claiming a fixed technical limit.
+
+The client's tool-call deadline must allow startup plus the requested query;
+server settings and progress notifications cannot override it. CLI analysis
+avoids the MCP client deadline, but retains the provider deadline. For example:
+
+```sh
+REA_GHIDRA_STARTUP_TIMEOUT_MS=1800000 rea analyze /absolute/path/to/app.exe --provider ghidra --json
+```
+
+Thirty minutes is an example budget, not a completion guarantee. Keep host CPU
+and memory controls in place. Disable specific expensive analyzers only when
+their missing coverage is acceptable and retain the resulting limitations.
+Exported C and embedded resources remain useful independent evidence, but a
+timeout does not establish complete analysis or absence of code. See the
+[deadline and client guide](https://github.com/morluto/rea/blob/main/docs/mcp-contracts.md#ghidra-first-query-deadlines-and-recovery).
+
+## Bridge request ceiling
+
+Hopper and Cutter share `REA_BRIDGE_MAX_REQUEST_BYTES`. Unset, both sides use
+16777216 bytes. The payload budget excludes the terminating LF. A decimal
+integer from 7464 to 536870887 raises or lowers that ceiling. The floor fits
+the largest bounded symbol-name command. A tight Cutter status still returns
+its session id and up to 64 completion receipts. Its path is included when it
+fits, and is unknown when it does not, so an older descriptor path is not
+paired with the new generation. The maximum is the encoded payload, one byte under Node's string ceiling so
+the framing LF still fits. Cutter measures, signs, and writes a fitting
+response in bounded chunks. The bridge searches only the newly read bytes
+for the framing LF. The client unescapes each socket chunk into one response
+string before parsing it, so this ceiling does not keep the paged wire frame
+and another concatenated copy. Invalid values are configuration errors and do not fall back to 16777216. Set the variable in the environment that starts REA, restart REA,
+and start Hopper or Cutter from that same environment. An out-of-band bridge
+without the setting stays at 16777216 bytes. The client reports that mismatch
+during discovery only after the hello proves a live authenticated bridge, and
+does not send a status request or the command. A stale descriptor is ignored.
+Discovery fails if any such live ceiling differs, even when another live
+session matches. A command response discarded for that ceiling stays an
+unknown execution with `output_truncated`. A hello
+that never sends its authenticated frame does not hold the accept loop.
+At most 8 connection threads run; a further accept is closed, and a failure
+to start a thread does not stop the listener. `completed_connection` is the
+longest prefix of connections that have terminated. Dense scalar replies are
+encoded in bounded batches.
+The 25000 ms Cutter request deadline and the 256-candidate discovery budget
+still have no runtime override.
+
+## Static memory and original file bytes
+
+For Ghidra, `read_bytes` reads initialized analysis memory, including loader
+fixups; it does not read a running process. Check `returned_bytes` and `complete`
+before decoding a table. An incomplete read is missing coverage, not a zero-filled
+tail. Use `list_segments` to inspect the selected block's initialization and address-space
+facts, and retain unavailable facts as unknown.
+
+Use `address_to_file_offset` to anchor one observed address to original file
+bytes. A mapping for the first address does not establish a contiguous mapping
+for an entire table, and loader fixups can make analysis bytes differ from source
+bytes. Keep source ranges and analysis-memory observations separate.
+
+A pointer table can refer to writable slots populated by a constructor rather
+than directly to immutable resources. An uninitialized block or an observed zero
+does not establish the slot's later runtime value. Follow `xrefs` and the relevant
+writer's assembly or pseudocode; distinguish recorded assignments from inferred
+execution order. Do not replace unreadable bytes with zeros or conclude that a
+resource is absent without the necessary initialization evidence.
+
 ## DOS MZ with Ghidra
 
 For an admitted DOS MZ target, use `inspect_native_load_image` to check measured
@@ -133,10 +354,8 @@ reported static image, not runtime DOS or PC-98 hardware behavior.
 
 For Ghidra decompilation, `procedure_pseudo_code` also accepts `procedure: "file-offset:0x..."`; CLI `rea decompile <path> --file-offset=0x... --provider ghidra` resolves the source byte offset internally. An unresolved 0x-prefixed numeric procedure selector also falls back to a unique file mapping. Existing symbols and analysis addresses win; multiple mappings fail as ambiguous. Preserve the returned file-offset and address locations and raw selection observations. ELF virtual addresses are distinct from file offsets; never assume an arbitrary offset plus image base is an analysis address.
 
-Use `read_bytes` for initialized analysis memory and `address_to_file_offset` to
-anchor one observed address to original file bytes. Loader fixups can change a
-word; a source offset does not imply byte equality. Complete function body ranges
-use inclusive ends and can contain gaps. Do not treat the enclosing span as code.
+Complete function body ranges use inclusive ends and can contain gaps. Do not
+treat the enclosing span as code.
 For packed targets, retain the original and separately derived artifact identities;
 decompiling an unpacking stub does not recover the unpacked program.
 
@@ -226,8 +445,8 @@ every executable in the bundle. It also lists missing or weak loads and earlier
 `architecture`. System paths stay undetermined because the dyld shared cache
 provides them.
 
-Use `extract_artifact` when materialized files are needed. It extracts the active
-target and accepts optional `integrity_policy`; see the [MCP contracts](https://github.com/morluto/rea/blob/main/docs/mcp-contracts.md).
+Use `extract_artifact` when materialized files from a supported format are needed.
+It extracts the active target and accepts optional `integrity_policy`; see the [MCP contracts](https://github.com/morluto/rea/blob/main/docs/mcp-contracts.md).
 It materializes all regular files into a fresh temporary directory chosen by
 REA. Symlinks and encrypted entries are inventory facts, not extractable files.
 The result reports the materialized directory; callers do not choose the
@@ -236,23 +455,30 @@ destination. No separate permission grant is needed.
 Native DMG traversal automatically uses a read-only mount on supported macOS
 hosts, with an owned temporary mount directory and cleanup. It needs no approval
 flag. Unsupported hosts or mount failures remain explicit limitations; do not
-claim child inventory when only root identity is available. Extraction is a
-separate filesystem-writing operation.
+claim child inventory when only root identity is available. DMG extraction is
+unavailable through `extract_artifact`. To access DMG files on macOS, create a
+read-only mount with your own tools, select files from that mount for analysis,
+and detach the mount you created when finished. REA's inspection mount is
+temporary, not a retained export of the image's contents.
 
 For headerless DOS COM, explicitly open with `format: "dos-com"` (CLI
 `--target-format dos-com`). Inspect the load image before trusting function
 analysis: the entry is `0x10100`, file offset zero, with imposed real-mode segment
 context. PSP/stack/device state remains unmodeled. See [DOS guide](https://github.com/morluto/rea/blob/main/docs/ghidra-dos.md).
 
+Use `inspect_byte_entropy` to triage high- and low-entropy ranges before extraction.
+
 For other headerless images (embedded firmware without ELF/Mach-O/PE/MZ
 headers), open with `format: "raw-binary"`, `architecture` (`x86`, `x86_64`,
-`arm`, `arm64`, or `6502`), and `entry_point` (`0x` hex). File offset 0 is mapped at
+`arm`, `arm64`, `6502`, or `m68k`), and `entry_point` (`0x` hex). File offset 0 is mapped at
 that address. Hopper and IDA refuse this interpretation; Ghidra uses
-`BinaryLoader` with a little-endian generic language. Big-endian, Thumb-only,
-and Cortex-M variants are unsupported. `6502` selects NMOS 6502; the entry and
-complete file mapping must fit within `0x0000..0xffff`. 65C02, 65816, cartridge
-containers, bank switching, and mapper semantics are unsupported. Windows
-Ghidra P0 admits raw images only for `6502`. See [raw binary guide](https://github.com/morluto/rea/blob/main/docs/raw-binary.md).
+`BinaryLoader` with a generic language: big-endian for `m68k`, little-endian
+otherwise. Big-endian ARM, Thumb-only, and Cortex-M variants are unsupported.
+`6502` selects NMOS 6502; the entry and complete file mapping must fit within
+`0x0000..0xffff`. 65C02, 65816, cartridge containers, bank switching, and mapper
+semantics are unsupported. `m68k` requires an even entry and decodes the 68040
+superset of the 68000 family; Amiga/Atari containers and vector-table entry
+discovery are unsupported. Windows Ghidra P0 admits raw images only for `6502`. See [raw binary guide](https://github.com/morluto/rea/blob/main/docs/raw-binary.md).
 
 ### Function annotations in Ghidra
 
@@ -270,8 +496,37 @@ Other payloads are opaque, and this operation does not decode images or simulate
 Windows resource loading. Consult the budgets and supported tree profile in
 [offline binary diagnostics](https://github.com/morluto/rea/blob/main/docs/binary-diagnostics.md).
 
+## Offline PE and ELF export tables
+
+Use `inspect_native_exports` with an absolute `path`, or
+`rea inspect-native-exports ./libexample.so --json`, to read a PE export
+directory or ELF dynamic symbol table on any host without opening a native
+target. Ordinal-only, forwarded and hidden-version exports stay distinct;
+partial or unsupported coverage means the inventory is incomplete, not that
+exports are absent. Use `inspect_macho` for Mach-O. See
+[offline binary diagnostics](https://github.com/morluto/rea/blob/main/docs/binary-diagnostics.md).
+
 For exact core-module import/export differences, use `compare_wasm_interfaces`
 with two complete inspection Evidence records (or current-session retained
 references) and explicit scope. Preserve unknown types/duplicate names; static
 interface equality is not runtime compatibility. CLI accepts complete inline
 Evidence in `rea compare-wasm-interfaces input.json --json`.
+
+## Godot PCK
+
+For a standalone PCK, `open_binary` then `inspect_artifact` with
+`detail: "summary"` retains the complete static inventory. Use the returned
+`parent_evidence_id` with `inspect_analysis_view` inventory occurrences filtered
+by exact `logical_paths` or `occurrence_ids`, and join `artifact_id` to nodes
+filtered by `artifact_ids`. Select only needed fields; identity links remain.
+Prefer `inspect-artifact --detail summary --evidence_output PATH` for CLI tasks;
+the same CLI discovers embedded packs via their EOF trailer without opening a
+native analysis provider. Normal native opening remains unchanged.
+
+Distinguish declared MD5 from observed SHA-256 and graph/page completeness from
+`source_coverage`. Encrypted directories and sparse bundles have unknown entry
+coverage; encrypted members have unknown content identity. An empty filtered
+page cannot prove absence in unknown coverage. Retained views are historical,
+not fresh filesystem checks. No decryption, Godot resource decoding, GDScript
+decompilation or target execution is provided. See
+[Godot PCK](https://github.com/morluto/rea/blob/main/docs/godot-pck.md).

.agents/skills/reverse-engineer-anything/references/runtime-observation.md

@@ -69,6 +69,12 @@ with `inspect_analysis_view` and `view: {kind: "capture", detail: "preview", off
 Select complete record pages only for needed headers, metadata or payloads; views
 read the historical parent without repeating capture decoding.
 
+For selected protobuf/gRPC bodies, keep raw bytes as the source and use
+`inspect_analysis_view` kind `protobuf`, collection `bodies`, `frames` or
+`fields`. Batch exact record ordinals and field numbers/IDs, and select
+`schema_value` or `text` only when needed. Supply a local descriptor set and
+message type for schema interpretation; compressed payloads remain unknown.
+
 Inspect native `.saz` snapshots with `inspect_web_network_capture` and
 `format: "saz"`; encrypted archives are unsupported. For live
 traffic analysis, connect Everywhere's separate MCP server and use its

.github/actions/npm-cache/action.yml

@@ -1,5 +1,10 @@
 name: Cache npm package downloads
-description: Restore npm's download cache and save it only from trusted default-branch events.
+description: Restore npm's download cache. Trusted default-branch jobs that opt in may save it.
+inputs:
+  save:
+    description: Save the shared download cache when this event is allowed to write it.
+    required: false
+    default: "false"
 outputs:
   can-save:
     description: Whether this run may update the shared npm download cache.
@@ -23,20 +28,25 @@ runs:
           can_save=true
         fi
         cache_key="npm-download-${RUNNER_OS}-${runtime_arch}-${LOCKFILES_HASH}"
+        # Same platform only. A lockfile change can reuse tarballs from the previous key.
+        restore_key="npm-download-${RUNNER_OS}-${runtime_arch}-"
         {
           printf 'can-save=%s\n' "${can_save}"
           printf 'path=%s\n' "${cache_path}"
           printf 'key=%s\n' "${cache_key}"
+          printf 'restore-key=%s\n' "${restore_key}"
         } >> "${GITHUB_OUTPUT}"
     - name: Restore and save npm package downloads
-      if: steps.context.outputs.can-save == 'true'
+      if: steps.context.outputs.can-save == 'true' && inputs.save == 'true'
       uses: actions/cache@0057852bfaa89a56745cba8c7296529d2fc39830 # v4.3.0
       with:
         path: ${{ steps.context.outputs.path }}
         key: ${{ steps.context.outputs.key }}
+        restore-keys: ${{ steps.context.outputs.restore-key }}
     - name: Restore npm package downloads without saving
-      if: steps.context.outputs.can-save != 'true'
+      if: steps.context.outputs.can-save != 'true' || inputs.save != 'true'
       uses: actions/cache/restore@0057852bfaa89a56745cba8c7296529d2fc39830 # v4.3.0
       with:
         path: ${{ steps.context.outputs.path }}
         key: ${{ steps.context.outputs.key }}
+        restore-keys: ${{ steps.context.outputs.restore-key }}

.github/actions/portable-build-cache/action.yml

@@ -24,19 +24,7 @@ runs:
         BUILD_CACHE_PREFIX: rea-portable-build-v2-${{ runner.os }}-${{ runner.arch }}-${{ hashFiles('.npmrc', '.nvmrc', 'package-lock.json') }}
       run: |
         npm run --silent test:prepare -- --dry=json |
-          node --input-type=module -e '
-            import { createHash } from "node:crypto";
-            import { readFileSync } from "node:fs";
-            const { tasks } = JSON.parse(readFileSync(0, "utf8"));
-            const hashes = tasks.map(({ taskId, hash }) => {
-              if (typeof hash !== "string")
-                throw new Error(`Build task ${taskId} must have an eager Turbo hash`);
-              return `${taskId}:${hash}`;
-            }).sort();
-            const hash = createHash("sha256").update(hashes.join("\n")).digest("hex");
-            console.log(`key=${process.env.BUILD_CACHE_PREFIX}-${hash}`);
-            console.log(`prefix=${process.env.BUILD_CACHE_PREFIX}-`);
-          ' >> "$GITHUB_OUTPUT"
+          node scripts/ci/portable-build-cache.mjs key >> "$GITHUB_OUTPUT"
     - name: Restore portable build outputs
       if: inputs.build_cache_artifact == '' && inputs.save != 'true'
       uses: actions/cache/restore@0057852bfaa89a56745cba8c7296529d2fc39830 # v4.3.0

.github/pull_request_template.md

@@ -17,7 +17,10 @@ describe the use case and observable outcome. -->
 ## Change and scope
 
 <!-- Explain the chosen approach, why it fits REA's architecture, and what is
-intentionally not included. -->
+intentionally not included. If adding complexity for a speculative or vanishingly
+unlikely race or security edge case, record the concrete failure mode, likelihood,
+complexity cost, and the user's buy-in obtained before implementation. Do not
+silently expand scope for theoretical robustness. -->
 
 Follow the [contribution scope](https://github.com/morluto/rea/blob/main/CONTRIBUTING.md#contribution-scope).
 Unresolved scope, authority, side effects, or required verification block

.github/workflows/cache-retention.yml

@@ -0,0 +1,64 @@
+name: Cache retention
+
+on:
+  workflow_run:
+    workflows: ["CI"]
+    types: [completed]
+
+# Qualification calls ci.yml with actions: read. A nested workflow cannot raise
+# that token, so cache deletion stays in this workflow.
+permissions:
+  contents: read
+  actions: write
+
+concurrency:
+  group: cache-retention-${{ github.event.workflow_run.head_sha }}
+  cancel-in-progress: false
+
+jobs:
+  retain:
+    name: Retain Turbo snapshots and npm download caches
+    # Restrict the triggering-source checkout to same-repository default-branch
+    # CI events. The helper runs from this trusted workflow revision; triggering
+    # source is used only for cache identity and is never executed. Static and
+    # build save snapshots before later lanes finish, so failed, timed-out, or
+    # cancelled runs still retain those snapshots.
+    if: >-
+      (github.event.workflow_run.conclusion == 'success' ||
+      github.event.workflow_run.conclusion == 'failure' ||
+      github.event.workflow_run.conclusion == 'timed_out' ||
+      github.event.workflow_run.conclusion == 'cancelled') &&
+      github.event.workflow_run.head_repository.full_name == github.repository &&
+      github.event.workflow_run.head_branch == github.event.repository.default_branch &&
+      (github.event.workflow_run.event == 'push' ||
+      github.event.workflow_run.event == 'schedule' ||
+      github.event.workflow_run.event == 'workflow_dispatch')
+    runs-on: ubuntu-latest
+    timeout-minutes: 10
+    steps:
+      - uses: actions/checkout@11d5960a326750d5838078e36cf38b85af677262 # v4.4.0
+        with:
+          ref: ${{ github.workflow_sha }}
+          path: trusted-retention
+          persist-credentials: false
+      - uses: actions/checkout@11d5960a326750d5838078e36cf38b85af677262 # v4.4.0
+        with:
+          ref: ${{ github.event.workflow_run.head_sha }}
+          path: triggering-source
+          persist-credentials: false
+      - uses: actions/setup-node@49933ea5288caeca8642d1e84afbd3f7d6820020 # v4.4.0
+        with:
+          node-version-file: trusted-retention/.nvmrc
+      - name: Retain current Turbo snapshots and npm download caches
+        env:
+          GH_TOKEN: ${{ github.token }}
+          CI_REPOSITORY: ${{ github.repository }}
+          CACHE_PREFIX: rea-static-v1-${{ runner.os }}-${{ runner.arch }}-
+          CACHE_KEEP_KEY: rea-static-v1-${{ runner.os }}-${{ runner.arch }}-${{ hashFiles('triggering-source/.nvmrc', 'triggering-source/.npmrc', 'triggering-source/package-lock.json') }}-${{ github.event.workflow_run.head_sha }}
+          CACHE_DEEP_PREFIX: rea-static-deep-v1-${{ runner.os }}-${{ runner.arch }}-
+          CACHE_DEEP_KEEP_KEY: rea-static-deep-v1-${{ runner.os }}-${{ runner.arch }}-${{ hashFiles('triggering-source/.nvmrc', 'triggering-source/.npmrc', 'triggering-source/package-lock.json') }}-${{ github.event.workflow_run.head_sha }}
+          CACHE_KNIP_PREFIX: knip-meta-v1-${{ runner.os }}-
+          CACHE_KNIP_KEEP_KEY: knip-meta-v1-${{ runner.os }}-${{ hashFiles('triggering-source/knip.json', 'triggering-source/tsconfig.json', 'triggering-source/package.json', 'triggering-source/package-lock.json') }}-${{ github.event.workflow_run.head_sha }}
+          CACHE_PORTABLE_PREFIX: rea-portable-build-v2-${{ runner.os }}-${{ runner.arch }}-
+          DEDUPE_NPM_DOWNLOADS: "true"
+        run: node trusted-retention/scripts/ci/cache-retention.mjs

.github/workflows/ci.yml

@@ -15,6 +15,15 @@ on:
         description: Exact qualification source; must match the caller revision.
         required: true
         type: string
+      remote_cache_write:
+        description: Trusted release qualification may publish the remote cache.
+        required: false
+        type: boolean
+        default: false
+    secrets:
+      TURBO_TOKEN:
+        description: Remote cache token. Only trusted release qualification passes it.
+        required: false
 
 permissions:
   contents: read
@@ -29,6 +38,15 @@ concurrency:
 
 env:
   HUSKY: "0"
+  # Set Actions secret TURBO_TOKEN and variable TURBO_TEAM. OIDC can mint
+  # TURBO_TOKEN via vercel/setup-turborepo-remote-cache-action; do not commit a token.
+  # A called workflow sees the caller event, so workflow_call is not an event
+  # name here. Release qualification sets remote_cache_write and passes the
+  # token. Open pull requests get an empty token and the local cache only.
+  # Remote reads need a credential, and pull requests do not receive one.
+  TURBO_TOKEN: "${{ (inputs.remote_cache_write == true || (github.ref == format('refs/heads/{0}', github.event.repository.default_branch) && (github.event_name == 'push' || github.event_name == 'schedule' || github.event_name == 'workflow_dispatch'))) && secrets.TURBO_TOKEN || '' }}"
+  TURBO_TEAM: ${{ vars.TURBO_TEAM }}
+  TURBO_CACHE: "${{ (inputs.remote_cache_write == true || (github.ref == format('refs/heads/{0}', github.event.repository.default_branch) && (github.event_name == 'push' || github.event_name == 'schedule' || github.event_name == 'workflow_dispatch'))) && 'local:rw,remote:rw' || 'local:rw' }}"
 
 jobs:
   changes:
@@ -54,12 +72,15 @@ jobs:
       browser: ${{ steps.plan.outputs.browser }}
       binary_layout: ${{ steps.plan.outputs.binary_layout }}
       evm: ${{ steps.plan.outputs.evm }}
+      apk_signature: ${{ steps.plan.outputs.apk_signature }}
       javascript_recovery: ${{ steps.plan.outputs.javascript_recovery }}
       recorded_crash: ${{ steps.plan.outputs.recorded_crash }}
       web_network: ${{ steps.plan.outputs.web_network }}
       web_runtime: ${{ steps.plan.outputs.web_runtime }}
       web_source_map: ${{ steps.plan.outputs.web_source_map }}
+      agentenv: ${{ steps.plan.outputs.agentenv }}
       nativeaot: ${{ steps.plan.outputs.nativeaot }}
+      ghidra_bridge: ${{ steps.plan.outputs.ghidra_bridge }}
       website: ${{ steps.plan.outputs.website }}
       matrix: ${{ steps.plan.outputs.matrix }}
       test_selection: ${{ steps.plan.outputs.test_selection }}
@@ -126,6 +147,9 @@ jobs:
       - name: Cache npm package downloads
         id: npm-cache
         uses: ./.github/actions/npm-cache
+        with:
+          # Linux x64 writer. Parallel Linux jobs restore this key.
+          save: "true"
       - run: npm ci
       - name: Restore static task cache
         if: needs.changes.outputs.code == 'true'
@@ -139,19 +163,20 @@ jobs:
       - name: Validate workflow syntax and expressions
         if: needs.changes.outputs.workflows == 'true'
         run: go run github.com/rhysd/actionlint/cmd/actionlint@v1.7.7 -shellcheck= -pyflakes=
+      # Code tiers run this source-only policy test in their existing shards.
+      # Workflow-only changes need the guard, without preparing runtime artifacts.
+      - name: Check action pins for workflow-only changes
+        if: needs.changes.outputs.workflows == 'true' && needs.changes.outputs.code != 'true'
+        run: node scripts/run-exact-tests.mjs run tests/boundary/filesystem/workflowActionPinning.test.ts
       - name: Record static task source and execution identity
         continue-on-error: true
         if: needs.changes.outputs.code == 'true'
         env:
           CI_PR_HEAD_SHA: ${{ github.event.pull_request.head.sha }}
+          CI_STATIC_CONTEXT_SCRIPT: check:ci:fast:tasks
         run: npm run ci:static-context
       - if: needs.changes.outputs.code == 'true'
-        run: npm run check:ci
-      - if: needs.changes.outputs.code == 'true'
-        run: npm run verify:test-discovery
-      - name: Check LLDB producer retention and launch contracts
-        if: needs.changes.outputs.code == 'true'
-        run: npm run test:python
+        run: npm run check:ci:fast
       - if: needs.changes.outputs.code != 'true'
         run: npm run format:check
       - name: Save validated static task cache
@@ -173,6 +198,72 @@ jobs:
           if-no-files-found: ignore
           retention-days: 7
 
+  static-deep:
+    name: Deep static checks
+    needs: changes
+    # Full qualification only. Provider lanes wait on fast static, not this job.
+    if: needs.changes.outputs.tier == 'full' && needs.changes.outputs.code == 'true'
+    runs-on: ubuntu-latest
+    timeout-minutes: 15
+    steps:
+      - uses: actions/checkout@11d5960a326750d5838078e36cf38b85af677262 # v4.4.0
+      - uses: actions/setup-node@49933ea5288caeca8642d1e84afbd3f7d6820020 # v4.4.0
+        with:
+          node-version-file: .nvmrc
+      - name: Cache npm package downloads
+        id: npm-cache
+        uses: ./.github/actions/npm-cache
+      - run: npm ci
+      - name: Restore deep static task cache
+        id: static-cache
+        uses: actions/cache/restore@0057852bfaa89a56745cba8c7296529d2fc39830 # v4.3.0
+        with:
+          path: .turbo/cache/
+          key: rea-static-deep-v1-${{ runner.os }}-${{ runner.arch }}-${{ hashFiles('.nvmrc', '.npmrc', 'package-lock.json') }}-${{ github.sha }}
+          restore-keys: rea-static-deep-v1-${{ runner.os }}-${{ runner.arch }}-${{ hashFiles('.nvmrc', '.npmrc', 'package-lock.json') }}-
+      - name: Restore Knip analysis cache
+        id: knip-cache
+        uses: actions/cache/restore@0057852bfaa89a56745cba8c7296529d2fc39830 # v4.3.0
+        with:
+          path: node_modules/.cache/knip
+          # Each trusted commit publishes a new entry. Restore still matches the
+          # configuration prefix from the previous successful full run.
+          key: knip-meta-v1-${{ runner.os }}-${{ hashFiles('knip.json', 'tsconfig.json', 'package.json', 'package-lock.json') }}-${{ github.sha }}
+          restore-keys: knip-meta-v1-${{ runner.os }}-${{ hashFiles('knip.json', 'tsconfig.json', 'package.json', 'package-lock.json') }}-
+      - name: Record deep static task source and execution identity
+        continue-on-error: true
+        env:
+          CI_PR_HEAD_SHA: ${{ github.event.pull_request.head.sha }}
+          CI_STATIC_CONTEXT_SCRIPT: check:ci:deep:tasks
+        run: npm run ci:static-context
+      - name: Run deep static checks
+        run: npm run check:ci:deep
+      - name: Save validated deep static task cache
+        if: >-
+          success() && steps.npm-cache.outputs.can-save == 'true' &&
+          steps.static-cache.outputs.cache-hit != 'true'
+        uses: actions/cache/save@0057852bfaa89a56745cba8c7296529d2fc39830 # v4.3.0
+        with:
+          path: .turbo/cache/
+          key: ${{ steps.static-cache.outputs.cache-primary-key }}
+      - name: Save Knip analysis cache
+        if: >-
+          success() && steps.npm-cache.outputs.can-save == 'true' &&
+          steps.knip-cache.outputs.cache-hit != 'true'
+        uses: actions/cache/save@0057852bfaa89a56745cba8c7296529d2fc39830 # v4.3.0
+        with:
+          path: node_modules/.cache/knip
+          key: ${{ steps.knip-cache.outputs.cache-primary-key }}
+      - name: Upload complete deep static diagnostics
+        if: always()
+        uses: actions/upload-artifact@ea165f8d65b6e75b540449e92b4886f43607fa02 # v4.6.2
+        with:
+          name: static-deep-check-reports
+          path: .cache/static-checks/
+          include-hidden-files: true
+          if-no-files-found: ignore
+          retention-days: 7
+
   test-source:
     name: Source tests
     needs: changes
@@ -212,9 +303,11 @@ jobs:
       # every shard's diagnostics and coverage. Passing PRs run the full suite.
       fail-fast: ${{ github.event_name == 'pull_request' && needs.changes.outputs.tier == 'fast' }}
       matrix:
-        # Fast feedback spends the same six test jobs on one source group and
-        # five built groups; full qualification retains six complete shards.
-        shard: ${{ fromJSON(needs.changes.outputs.tier == 'fast' && '["1/5","2/5","3/5","4/5","5/5"]' || '["1/6","2/6","3/6","4/6","5/6","6/6"]') }}
+        # Fast feedback uses one source group and eight built groups so the
+        # slowest built shard job stays inside a ten-minute execution budget.
+        # Classification, build, and report merge sit outside that job.
+        # Full qualification retains six complete shards with coverage.
+        shard: ${{ fromJSON(needs.changes.outputs.tier == 'fast' && '["1/8","2/8","3/8","4/8","5/8","6/8","7/8","8/8"]' || '["1/6","2/6","3/6","4/6","5/6","6/6"]') }}
     steps:
       - uses: actions/checkout@11d5960a326750d5838078e36cf38b85af677262 # v4.4.0
       - uses: actions/setup-node@49933ea5288caeca8642d1e84afbd3f7d6820020 # v4.4.0
@@ -292,11 +385,13 @@ jobs:
       needs.changes.outputs.browser == 'true' ||
       needs.changes.outputs.binary_layout == 'true' ||
       needs.changes.outputs.evm == 'true' ||
+      needs.changes.outputs.apk_signature == 'true' ||
       needs.changes.outputs.javascript_recovery == 'true' ||
       needs.changes.outputs.recorded_crash == 'true' ||
       needs.changes.outputs.web_network == 'true' ||
       needs.changes.outputs.web_runtime == 'true' ||
-      needs.changes.outputs.web_source_map == 'true'
+      needs.changes.outputs.web_source_map == 'true' ||
+      needs.changes.outputs.agentenv == 'true'
     runs-on: ubuntu-latest
     timeout-minutes: 10
     permissions:
@@ -316,7 +411,9 @@ jobs:
         with:
           save: ${{ github.event_name != 'pull_request' && github.ref == format('refs/heads/{0}', github.event.repository.default_branch) }}
       - run: npm run test:prepare
-      - name: Find available trusted full-suite timing history
+      # Fast and full shards measure different file costs. Download only the
+      # matching history so one tier cannot pack the other with its weights.
+      - name: Find available trusted timing history
         id: history
         continue-on-error: true
         env:
@@ -324,13 +421,14 @@ jobs:
           CI_REPOSITORY: ${{ github.repository }}
           CI_REPOSITORY_ID: ${{ github.repository_id }}
           CI_DEFAULT_BRANCH: ${{ github.event.repository.default_branch }}
+          CI_TIMING_TIER: ${{ needs.changes.outputs.tier }}
         run: node scripts/ci/test-history.mjs
       - name: Download previous timing report
         if: steps.history.outputs.run-id != ''
         continue-on-error: true
         uses: actions/download-artifact@d3f86a106a0bac45b974a628896c90dbdf5c8093 # v4.3.0
         with:
-          name: test-timings
+          name: ${{ steps.history.outputs.artifact-name }}
           run-id: ${{ steps.history.outputs.run-id }}
           github-token: ${{ github.token }}
           path: .cache/ci-test-history
@@ -351,17 +449,23 @@ jobs:
       # test:prepare caches only portable JS, skill, and catalog outputs.
       # Native package jobs still run their build and prepack checks; Turbo
       # restores matching outputs or rebuilds when their inputs differ.
+      - name: Stage current portable build task cache
+        shell: bash
+        run: |
+          # A new Turbo dry run can evict the entries this build just used.
+          node scripts/ci/portable-build-cache.mjs snapshot < .cache/ci-build-plan.json
       - uses: actions/upload-artifact@ea165f8d65b6e75b540449e92b4886f43607fa02 # v4.6.2
         with:
           name: rea-build-cache
-          path: .turbo/cache/
+          path: .cache/ci-build-cache/
           include-hidden-files: true
           if-no-files-found: error
           retention-days: 7
 
   sqlite-runtime:
-    # Preserve all selected coverage, but admit expensive host/provider work
-    # only after static validation. Portable tests still provide early feedback.
+    # Admit expensive host/provider work after fast static validation
+    # (typecheck, lint, format). Deep static does not gate these lanes.
+    # Portable tests still provide early feedback.
     needs: [changes, static, build]
     if: needs.changes.outputs.sqlite == 'true'
     uses: ./.github/workflows/sqlite-runtime.yml
@@ -456,11 +560,22 @@ jobs:
           include-hidden-files: true
           if-no-files-found: error
           retention-days: 30
+      # Default-branch runs only. Pull-request artifacts share this name and
+      # the lookup reads one page, so publishing them would hide main.
       - name: Publish successful timing history
-        if: needs.changes.outputs.code == 'true' && needs.changes.outputs.tier == 'full' && success() && needs.test-shard.result == 'success'
+        if: >-
+          success() &&
+          needs.changes.outputs.code == 'true' &&
+          needs.test-shard.result == 'success' &&
+          (
+            needs.changes.outputs.tier == 'full' ||
+            (needs.changes.outputs.tier == 'fast' && needs.test-source.result == 'success')
+          ) &&
+          github.ref == format('refs/heads/{0}', github.event.repository.default_branch) &&
+          (github.event_name == 'push' || github.event_name == 'schedule' || github.event_name == 'workflow_dispatch')
         uses: actions/upload-artifact@ea165f8d65b6e75b540449e92b4886f43607fa02 # v4.6.2
         with:
-          name: test-timings
+          name: test-timings-${{ needs.changes.outputs.tier }}
           path: .cache/ci-test-history/test-results.json
           include-hidden-files: true
           if-no-files-found: error
@@ -564,6 +679,9 @@ jobs:
       - name: Cache npm package downloads
         id: npm-cache
         uses: ./.github/actions/npm-cache
+        with:
+          # ubuntu-latest is the Linux x64 writer above. Other hosts save their own key.
+          save: ${{ matrix.os != 'ubuntu-latest' }}
       - name: Verify native runtime identity
         env:
           EXPECTED_CI_PLATFORM: ${{ matrix.platform }}
@@ -642,9 +760,11 @@ jobs:
           tests/boundary/cli/nativePlistPrototypeKeyCli.test.ts
           tests/boundary/cli/nativePlistSelectionFailureCli.test.ts
           tests/boundary/cli/nativeDemangleCli.test.ts
+          tests/boundary/cli/nativeDispatchPropertyAttributes.test.ts
           tests/boundary/mcp/nativeDemangleMcp.test.ts
           tests/boundary/providers/native/nativeSignatureTargetBinding.test.ts
           tests/boundary/cli/nativeSignatureCli.test.ts
+          tests/boundary/cli/nativeMachoArm64_32Cli.test.ts
           tests/boundary/mcp/nativeSignatureTargetBindingMcp.test.ts
           src/process/capture/FilesystemSnapshot.test.ts
           tests/process-global/filesystemSnapshotFallbackCancellation.test.ts
@@ -699,17 +819,24 @@ jobs:
       - name: Cache npm package downloads
         id: npm-cache
         uses: ./.github/actions/npm-cache
+        with:
+          save: "true"
       - run: npm ci
+      - name: Typecheck with the official Windows compiler
+        env:
+          GOMAXPROCS: "2"
+          GOMEMLIMIT: "3GiB"
+        run: npm run typecheck -- --checkers 2
       - uses: actions/download-artifact@d3f86a106a0bac45b974a628896c90dbdf5c8093 # v4.3.0
         with:
           name: rea-build-cache
           path: .turbo/cache/
-      - run: npm run build:cached
       - uses: actions/download-artifact@d3f86a106a0bac45b974a628896c90dbdf5c8093 # v4.3.0
         with:
           name: windows-native-controls
           path: native/windows/build/
       - run: npm run verify:windows-native:artifact
+      - run: npm run build:cached
       - name: Record named Windows host capabilities
         run: node scripts/windows-capabilities.mjs | Out-File -Encoding utf8NoBOM windows-capabilities.json
       - name: Validate Windows host capability report
@@ -730,17 +857,26 @@ jobs:
           npx vitest run --coverage.enabled=false
           src/process/capture/FilesystemSnapshot.test.ts
           tests/process-global/filesystemSnapshotFallbackCancellation.test.ts
+          tests/boundary/filesystem/typescriptCompiler.test.ts
           tests/boundary/filesystem/binaryTarget.test.ts
+          src/contracts/sessionPathInputs.test.ts
+          tests/boundary/filesystem/directAnalysisSnapshotReplay.test.ts
+          tests/boundary/mcp/rawBinaryMcp.test.ts
           tests/boundary/filesystem/safeOutputTree.test.ts
           tests/boundary/providers/native/artifactProvider.archiveSafety.test.ts
           tests/boundary/providers/native/artifactProvider.extractionCompleteness.test.ts
           tests/boundary/filesystem/setupUnavailableProvider.test.ts
           tests/boundary/filesystem/windowsNpxRegistration.test.ts
+          tests/boundary/filesystem/clientConfigurationEnvironment.test.ts
+          tests/boundary/filesystem/clientConfigurationLayeredEnvironment.test.ts
+          tests/acceptance/setup/registrationRuntimeEnvironmentCli.test.ts
           tests/boundary/android/workspaceGuardianRetry.test.ts
           tests/acceptance/setup/homeDirectoryFallback.test.ts
           tests/boundary/filesystem/javascriptApplicationWorkflows.test.ts
           tests/boundary/filesystem/webScriptManifestPaths.test.ts
           tests/boundary/process/ProcessEvidence.host.test.ts
+          tests/boundary/process/packageNpmInvocation.test.ts
+          tests/boundary/process/packedBridgeVerification.test.ts
           tests/boundary/process/windowsOutputCapture.test.ts
           tests/acceptance/webModuleCli.test.ts
           tests/acceptance/webSourceLocationCli.test.ts
@@ -763,6 +899,8 @@ jobs:
           src/process/ProcessOwnership.validation.test.ts
           src/application/Update.test.ts
           tests/process-global/updateMaintenance.test.ts
+          tests/process-global/runnerCancellation.test.ts
+          tests/process-global/windowsOwnedCommand.test.ts
           tests/conformance/providers/ghidra/realGhidraWindowsWorkflow.test.ts
       - uses: actions/download-artifact@d3f86a106a0bac45b974a628896c90dbdf5c8093 # v4.3.0
         with:
@@ -795,7 +933,7 @@ jobs:
           tests/boundary/filesystem/javascriptArtifactContentIdentity.test.ts
       - uses: actions/setup-node@49933ea5288caeca8642d1e84afbd3f7d6820020 # v4.4.0
         with:
-          node-version: 22.19.0
+          node-version: 22.23.4
       - name: Verify native path boundaries on Node 22
         run: >-
           npx vitest run --coverage.enabled=false
@@ -837,67 +975,117 @@ jobs:
             skills/reverse-engineer-anything
           if-no-files-found: error
 
+  # Provider lanes wait for fast static and the portable build, not deep static.
   browser:
     needs: [changes, static, build]
     if: needs.changes.outputs.browser == 'true'
+    secrets:
+      TURBO_TOKEN: ${{ secrets.TURBO_TOKEN }}
     uses: ./.github/workflows/real-browser.yml
     with:
       build_cache_artifact: rea-build-cache
+      remote_cache_write: ${{ inputs.remote_cache_write == true || (github.ref == format('refs/heads/{0}', github.event.repository.default_branch) && (github.event_name == 'push' || github.event_name == 'schedule' || github.event_name == 'workflow_dispatch')) }}
 
   binary_layout:
     needs: [changes, static, build]
     if: needs.changes.outputs.binary_layout == 'true'
+    secrets:
+      TURBO_TOKEN: ${{ secrets.TURBO_TOKEN }}
     uses: ./.github/workflows/real-binary-layout.yml
     with:
       build_cache_artifact: rea-build-cache
+      remote_cache_write: ${{ inputs.remote_cache_write == true || (github.ref == format('refs/heads/{0}', github.event.repository.default_branch) && (github.event_name == 'push' || github.event_name == 'schedule' || github.event_name == 'workflow_dispatch')) }}
 
   evm:
     needs: [changes, static, build]
     if: needs.changes.outputs.evm == 'true'
+    secrets:
+      TURBO_TOKEN: ${{ secrets.TURBO_TOKEN }}
     uses: ./.github/workflows/real-evm-interface.yml
     with:
       build_cache_artifact: rea-build-cache
+      remote_cache_write: ${{ inputs.remote_cache_write == true || (github.ref == format('refs/heads/{0}', github.event.repository.default_branch) && (github.event_name == 'push' || github.event_name == 'schedule' || github.event_name == 'workflow_dispatch')) }}
+
+  apk_signature:
+    needs: [changes, static, build]
+    if: needs.changes.outputs.apk_signature == 'true'
+    secrets:
+      TURBO_TOKEN: ${{ secrets.TURBO_TOKEN }}
+    uses: ./.github/workflows/real-apk-signature.yml
+    with:
+      build_cache_artifact: rea-build-cache
+      remote_cache_write: ${{ inputs.remote_cache_write == true || (github.ref == format('refs/heads/{0}', github.event.repository.default_branch) && (github.event_name == 'push' || github.event_name == 'schedule' || github.event_name == 'workflow_dispatch')) }}
 
   javascript_recovery:
     needs: [changes, static, build]
     if: needs.changes.outputs.javascript_recovery == 'true'
+    secrets:
+      TURBO_TOKEN: ${{ secrets.TURBO_TOKEN }}
     uses: ./.github/workflows/real-javascript-recovery.yml
     with:
       build_cache_artifact: rea-build-cache
+      remote_cache_write: ${{ inputs.remote_cache_write == true || (github.ref == format('refs/heads/{0}', github.event.repository.default_branch) && (github.event_name == 'push' || github.event_name == 'schedule' || github.event_name == 'workflow_dispatch')) }}
 
   recorded_crash:
     needs: [changes, static, build]
     if: needs.changes.outputs.recorded_crash == 'true'
+    secrets:
+      TURBO_TOKEN: ${{ secrets.TURBO_TOKEN }}
     uses: ./.github/workflows/real-recorded-crash.yml
     with:
       build_cache_artifact: rea-build-cache
+      remote_cache_write: ${{ inputs.remote_cache_write == true || (github.ref == format('refs/heads/{0}', github.event.repository.default_branch) && (github.event_name == 'push' || github.event_name == 'schedule' || github.event_name == 'workflow_dispatch')) }}
 
   web_network:
     needs: [changes, static, build]
     if: needs.changes.outputs.web_network == 'true'
+    secrets:
+      TURBO_TOKEN: ${{ secrets.TURBO_TOKEN }}
     uses: ./.github/workflows/real-web-network-captures.yml
     with:
       build_cache_artifact: rea-build-cache
+      remote_cache_write: ${{ inputs.remote_cache_write == true || (github.ref == format('refs/heads/{0}', github.event.repository.default_branch) && (github.event_name == 'push' || github.event_name == 'schedule' || github.event_name == 'workflow_dispatch')) }}
 
   web_runtime:
     needs: [changes, static, build]
     if: needs.changes.outputs.web_runtime == 'true'
+    secrets:
+      TURBO_TOKEN: ${{ secrets.TURBO_TOKEN }}
     uses: ./.github/workflows/real-web-runtime.yml
     with:
       build_cache_artifact: rea-build-cache
+      remote_cache_write: ${{ inputs.remote_cache_write == true || (github.ref == format('refs/heads/{0}', github.event.repository.default_branch) && (github.event_name == 'push' || github.event_name == 'schedule' || github.event_name == 'workflow_dispatch')) }}
 
   web_source_map:
     needs: [changes, static, build]
     if: needs.changes.outputs.web_source_map == 'true'
+    secrets:
+      TURBO_TOKEN: ${{ secrets.TURBO_TOKEN }}
     uses: ./.github/workflows/real-web-source-map.yml
     with:
       build_cache_artifact: rea-build-cache
+      remote_cache_write: ${{ inputs.remote_cache_write == true || (github.ref == format('refs/heads/{0}', github.event.repository.default_branch) && (github.event_name == 'push' || github.event_name == 'schedule' || github.event_name == 'workflow_dispatch')) }}
+
+  agentenv:
+    needs: [changes, static, build]
+    if: needs.changes.outputs.agentenv == 'true'
+    secrets:
+      TURBO_TOKEN: ${{ secrets.TURBO_TOKEN }}
+    uses: ./.github/workflows/real-agentenv.yml
+    with:
+      build_cache_artifact: rea-build-cache
+      remote_cache_write: ${{ inputs.remote_cache_write == true || (github.ref == format('refs/heads/{0}', github.event.repository.default_branch) && (github.event_name == 'push' || github.event_name == 'schedule' || github.event_name == 'workflow_dispatch')) }}
 
   nativeaot:
     needs: [changes, static]
     if: needs.changes.outputs.nativeaot == 'true'
     uses: ./.github/workflows/nativeaot-fixtures.yml
 
+  ghidra_bridge:
+    needs: [changes, static]
+    if: needs.changes.outputs.ghidra_bridge == 'true'
+    uses: ./.github/workflows/ghidra-bridge-sources.yml
+
   website:
     needs: [changes, static]
     if: needs.changes.outputs.website == 'true'
@@ -911,6 +1099,7 @@ jobs:
       [
         changes,
         static,
+        static-deep,
         build,
         reverse-tools-quality,
         test-shard,
@@ -927,12 +1116,15 @@ jobs:
         browser,
         binary_layout,
         evm,
+        apk_signature,
         javascript_recovery,
         recorded_crash,
         web_network,
         web_runtime,
         web_source_map,
+        agentenv,
         nativeaot,
+        ghidra_bridge,
         website,
       ]
     runs-on: ubuntu-24.04

.github/workflows/ghidra-bridge-sources.yml

@@ -0,0 +1,64 @@
+name: Ghidra bridge source compilation
+
+on:
+  workflow_dispatch:
+  workflow_call:
+
+permissions:
+  contents: read
+
+env:
+  HUSKY: "0"
+
+jobs:
+  ghidra-bridge-sources:
+    name: Compile packaged Ghidra bridge sources
+    runs-on: ubuntu-24.04
+    timeout-minutes: 15
+    env:
+      GHIDRA_RELEASE_URL: https://github.com/NationalSecurityAgency/ghidra/releases/download/Ghidra_12.1.4_build/ghidra_12.1.4_PUBLIC_20260921.zip
+      GHIDRA_RELEASE_SHA256: ddac49f903da9d5bac833e5cc79395098b9c33cfd3279be5f31bd00387d2d4db
+    steps:
+      - uses: actions/checkout@11d5960a326750d5838078e36cf38b85af677262 # v4.4.0
+      - uses: actions/setup-node@49933ea5288caeca8642d1e84afbd3f7d6820020 # v4.4.0
+        with:
+          node-version-file: .nvmrc
+      - name: Validate the SDK pin against the provider's verified release
+        run: node scripts/verify-ghidra-bridge-sources.mjs --check-sdk-pin >> "$GITHUB_ENV"
+      # The bridge probe imports only source-owned ARM/Thumb bytes and never
+      # executes target code. All SDK/project state stays in runner storage.
+      - name: Acquire checksum-verified Ghidra SDK in isolated runner storage
+        shell: bash
+        run: |
+          ghidra_root="${RUNNER_TEMP}/rea-ghidra"
+          mkdir -p "${ghidra_root}"
+          curl -fsSL --retry 3 --write-out 'ghidra_download_bytes=%{size_download} ghidra_download_seconds=%{time_total}\n' -o "${ghidra_root}/ghidra.zip" "${GHIDRA_RELEASE_URL}"
+          echo "${GHIDRA_RELEASE_SHA256}  ${ghidra_root}/ghidra.zip" | sha256sum -c -
+          unzip -q "${ghidra_root}/ghidra.zip" -d "${ghidra_root}"
+          rm "${ghidra_root}/ghidra.zip"
+          properties="$(find "${ghidra_root}" -path '*/Ghidra/application.properties' -print -quit)"
+          test -n "${properties}"
+          echo "GHIDRA_INSTALL_DIR=$(dirname "$(dirname "${properties}")")" >> "${GITHUB_ENV}"
+      # JAVA_HOME_<major>_X64 comes from the runner image, not the env context.
+      - name: Compile bridge sources with the verified runner JDK
+        shell: bash
+        run: |
+          java_home_variable="JAVA_HOME_${GHIDRA_JAVA_MAJOR}_X64"
+          test -n "${!java_home_variable}"
+          JAVA_HOME="${!java_home_variable}" npm run verify:ghidra:bridge-sources -- --verified-sdk
+      - name: Inspect real ARM and Thumb instruction context through the bridge
+        shell: bash
+        run: |
+          java_home_variable="JAVA_HOME_${GHIDRA_JAVA_MAJOR}_X64"
+          export JAVA_HOME="${!java_home_variable}"
+          probe_root="${RUNNER_TEMP}/rea-arm-context"
+          mkdir -p "${probe_root}/project"
+          printf '\036\377\057\341\160\107\000\000' > "${probe_root}/arm-thumb.bin"
+          timeout 180s "${GHIDRA_INSTALL_DIR}/support/analyzeHeadless" "${probe_root}/project" ArmContext \
+            -import "${probe_root}/arm-thumb.bin" -loader BinaryLoader -processor ARM:LE:32:v7 \
+            -noanalysis -readOnly -deleteProject \
+            -scriptPath "${GITHUB_WORKSPACE}/tests/conformance/ghidra;${GITHUB_WORKSPACE}/bridge/ghidra" \
+            -postScript ReaArmContextProbe.java "${GITHUB_WORKSPACE}/bridge/ghidra" \
+            > "${probe_root}/output.log" 2>&1 || { cat "${probe_root}/output.log"; exit 1; }
+          cat "${probe_root}/output.log"
+          grep -F 'REA_ARM_CONTEXT_OK arm=0 thumb=1 unavailable=null (injected provider value)' "${probe_root}/output.log"

.github/workflows/mutation.yml

@@ -14,6 +14,11 @@ concurrency:
 
 env:
   HUSKY: "0"
+  # Writes only on the default branch for push, schedule, manual dispatch,
+  # and workflow_call. Pull requests get an empty token and remote read-only mode.
+  TURBO_TOKEN: "${{ (github.ref == format('refs/heads/{0}', github.event.repository.default_branch) && (github.event_name == 'push' || github.event_name == 'schedule' || github.event_name == 'workflow_dispatch' || github.event_name == 'workflow_call')) && secrets.TURBO_TOKEN || '' }}"
+  TURBO_TEAM: ${{ vars.TURBO_TEAM }}
+  TURBO_CACHE: "${{ (github.ref == format('refs/heads/{0}', github.event.repository.default_branch) && (github.event_name == 'push' || github.event_name == 'schedule' || github.event_name == 'workflow_dispatch' || github.event_name == 'workflow_call')) && 'local:rw,remote:rw' || 'local:rw,remote:r' }}"
 
 jobs:
   mutation:

.github/workflows/real-agentenv.yml

@@ -0,0 +1,98 @@
+name: Real agentenv environment bundle
+
+on:
+  workflow_dispatch:
+  workflow_call:
+    inputs:
+      build_cache_artifact:
+        description: Portable build cache artifact from the calling CI run.
+        required: false
+        type: string
+        default: ""
+      remote_cache_write:
+        description: Trusted release qualification may publish the remote cache.
+        required: false
+        type: boolean
+        default: false
+    secrets:
+      TURBO_TOKEN:
+        description: Remote cache token. Only trusted callers pass it.
+        required: false
+
+permissions:
+  contents: read
+
+env:
+  HUSKY: "0"
+  # A called workflow sees the caller event, so workflow_call is not an event
+  # name here. Writes when remote_cache_write is set, or on the default branch
+  # for push, schedule, and manual dispatch. Open pull requests stay empty-token
+  # and use the local cache only.
+  TURBO_TOKEN: "${{ (inputs.remote_cache_write == true || (github.ref == format('refs/heads/{0}', github.event.repository.default_branch) && (github.event_name == 'push' || github.event_name == 'schedule' || github.event_name == 'workflow_dispatch'))) && secrets.TURBO_TOKEN || '' }}"
+  TURBO_TEAM: ${{ vars.TURBO_TEAM }}
+  TURBO_CACHE: "${{ (inputs.remote_cache_write == true || (github.ref == format('refs/heads/{0}', github.event.repository.default_branch) && (github.event_name == 'push' || github.event_name == 'schedule' || github.event_name == 'workflow_dispatch'))) && 'local:rw,remote:rw' || 'local:rw' }}"
+  NODE_OPTIONS: --max-old-space-size=768
+  PIP_DISABLE_PIP_VERSION_CHECK: "1"
+
+jobs:
+  real-agentenv:
+    name: Export and run an environment bundle with the agentenv runner
+    runs-on: ubuntu-24.04
+    timeout-minutes: 30
+    steps:
+      - uses: actions/checkout@11d5960a326750d5838078e36cf38b85af677262 # v4.4.0
+      - uses: actions/setup-node@49933ea5288caeca8642d1e84afbd3f7d6820020 # v4.4.0
+        with:
+          node-version-file: .nvmrc
+      - name: Cache npm package downloads
+        id: npm-cache
+        uses: ./.github/actions/npm-cache
+      - run: npm ci
+      - uses: ./.github/actions/portable-build-cache
+        with:
+          build_cache_artifact: ${{ inputs.build_cache_artifact }}
+      - uses: actions/setup-python@a26af69be951a213d495a4c3e4e4022e16d87065 # v5.6.0
+        with:
+          python-version: "3.12"
+      # Public ECR caps anonymous pulls by data per source IP and hosted runners share IPs, so the
+      # framework's image builds can fail here for hours. mirror.gcr.io serves the same Docker
+      # official images under the same digests, so each base image the pinned framework builds from
+      # is pulled there and named as its Dockerfile names it, as the framework's own CI does.
+      # Digest-pinned names are set in containerd, whose images Docker reads only on the containerd
+      # image store; switching stores restarts dockerd, so it happens before anything uses Docker.
+      - name: Use the containerd image store
+        run: |
+          [ -s /etc/docker/daemon.json ] || echo '{}' | sudo tee /etc/docker/daemon.json >/dev/null
+          jq '.features["containerd-snapshotter"] = true' /etc/docker/daemon.json > "$RUNNER_TEMP/daemon.json"
+          sudo mv "$RUNNER_TEMP/daemon.json" /etc/docker/daemon.json
+          sudo systemctl restart docker
+          docker info --format '{{json .DriverStatus}}' | grep -q io.containerd.snapshotter
+      - name: Acquire the pinned agentenv runner in isolated runner storage and check host Docker
+        shell: bash
+        run: |
+          runner_root="${RUNNER_TEMP}/rea-agentenv-runner"
+          python -m venv "${runner_root}"
+          "${runner_root}/bin/python" -m pip install --require-virtualenv --no-cache-dir 'agentenv-framework==0.9.1334'
+          "${runner_root}/bin/agent-env" --version
+          docker version --format '{{.Server.Version}}'
+          echo "REA_AGENT_ENV_COMMAND=${runner_root}/bin/agent-env" >> "${GITHUB_ENV}"
+          echo "REA_DOCKER_COMMAND=$(command -v docker)" >> "${GITHUB_ENV}"
+      - name: Pull the pinned framework's public ECR base images from mirror.gcr.io
+        shell: bash
+        run: |
+          package="$("${REA_AGENT_ENV_COMMAND%/agent-env}/python" -c 'import agent_env, pathlib; print(pathlib.Path(agent_env.__file__).parent)')"
+          grep -rhoE '^FROM +public\.ecr\.aws/docker/library/[^ ]+' --include='Dockerfile*' "${package}" | awk '{print $2}' | sort -u |
+            while read -r image; do
+              name=${image#public.ecr.aws/docker/library/}
+              source=mirror.gcr.io/library/$name
+              case $name in *@*) source=mirror.gcr.io/library/${name%%[:@]*}@${name#*@} ;; esac
+              for attempt in 1 2 3; do
+                docker pull --quiet "$source" && break
+                if [ "$attempt" = 3 ]; then echo "::warning::couldn't pull $source; $image's build will pull it"; continue 2; fi
+                sleep $((attempt * 10))
+              done
+              case ${name%@*} in *:*) docker tag "$source" "${image%@*}" ;; esac
+              case $image in *@*) sudo ctr --namespace moby images tag "$source" "$image" >/dev/null ;; esac
+            done
+      - name: Export a bundle from the source-owned capture and run it with the agentenv runner
+        run: npm run verify:agentenv:environment

.github/workflows/real-apk-signature.yml

@@ -0,0 +1,97 @@
+name: Real APK signature inspection
+
+on:
+  workflow_call:
+    inputs:
+      build_cache_artifact:
+        description: Portable build cache artifact from the calling CI run.
+        required: false
+        type: string
+        default: ""
+      remote_cache_write:
+        description: Trusted release qualification may publish the remote cache.
+        required: false
+        type: boolean
+        default: false
+    secrets:
+      TURBO_TOKEN:
+        description: Remote cache token. Only trusted callers pass it.
+        required: false
+  workflow_dispatch:
+
+permissions:
+  contents: read
+
+env:
+  HUSKY: "0"
+  # A called workflow sees the caller event, so workflow_call is not an event
+  # name here. Writes when remote_cache_write is set, or on the default branch
+  # for push, schedule, and manual dispatch. Open pull requests stay empty-token
+  # and use the local cache only.
+  TURBO_TOKEN: "${{ (inputs.remote_cache_write == true || (github.ref == format('refs/heads/{0}', github.event.repository.default_branch) && (github.event_name == 'push' || github.event_name == 'schedule' || github.event_name == 'workflow_dispatch'))) && secrets.TURBO_TOKEN || '' }}"
+  TURBO_TEAM: ${{ vars.TURBO_TEAM }}
+  TURBO_CACHE: "${{ (inputs.remote_cache_write == true || (github.ref == format('refs/heads/{0}', github.event.repository.default_branch) && (github.event_name == 'push' || github.event_name == 'schedule' || github.event_name == 'workflow_dispatch'))) && 'local:rw,remote:rw' || 'local:rw' }}"
+  NODE_OPTIONS: "--max-old-space-size=768"
+  RAYON_NUM_THREADS: "1"
+  GOMAXPROCS: "2"
+  UV_THREADPOOL_SIZE: "2"
+  JAVA_TOOL_OPTIONS: "-Xmx512m -XX:ActiveProcessorCount=2"
+
+jobs:
+  real-apk-signature:
+    runs-on: ubuntu-latest
+    timeout-minutes: 12
+    steps:
+      - uses: actions/checkout@11d5960a326750d5838078e36cf38b85af677262 # v4.4.0
+      - uses: actions/setup-node@49933ea5288caeca8642d1e84afbd3f7d6820020 # v4.4.0
+        with:
+          node-version-file: .nvmrc
+      - name: Cache npm package downloads
+        id: npm-cache
+        uses: ./.github/actions/npm-cache
+      - run: npm ci
+      - uses: ./.github/actions/portable-build-cache
+        with:
+          build_cache_artifact: ${{ inputs.build_cache_artifact }}
+      - name: Prepare pinned SDK oracle without installing host software
+        shell: bash
+        run: |
+          python3 - <<'PYTHON'
+          import hashlib, os, urllib.request, zipfile
+          from pathlib import Path
+          root = Path(os.environ["RUNNER_TEMP"]) / "rea-apk-signature-sdk"
+          packages = [
+              ("build-tools_r36.1_linux.zip", "a7b5889e4a79fcf3b0976bef40d401f4240fb1eed891d9d91169da1111e11d78", 64195204, "build-tools/36.1.0", {"apksigner", "aapt2", "lib/apksigner.jar", "source.properties"}),
+              ("platform-36_r02.zip", "37607369a28c5b640b3a7998868d45898ebcb777565a0e85f9acf36f29631d2e", 65878410, "platforms/android-36", {"android.jar", "source.properties"}),
+          ]
+          for name, expected, expected_size, directory, selected in packages:
+              archive = root / name
+              archive.parent.mkdir(parents=True, exist_ok=True)
+              digest = hashlib.sha256()
+              total = 0
+              with urllib.request.urlopen("https://dl.google.com/android/repository/" + name, timeout=60) as response, archive.open("wb") as output:
+                  while chunk := response.read(1024 * 1024):
+                      total += len(chunk)
+                      if total > expected_size:
+                          raise RuntimeError("SDK archive exceeds the pinned package size budget")
+                      digest.update(chunk)
+                      output.write(chunk)
+              if total != expected_size or digest.hexdigest() != expected:
+                  raise RuntimeError("SDK oracle archive digest mismatch")
+              extracted = set()
+              with zipfile.ZipFile(archive) as source:
+                  for entry in source.infolist():
+                      relative = entry.filename.partition("/")[2]
+                      if relative not in selected:
+                          continue
+                      target = root / directory / relative
+                      target.parent.mkdir(parents=True, exist_ok=True)
+                      target.write_bytes(source.read(entry))
+                      target.chmod((entry.external_attr >> 16) & 0o777 or 0o644)
+                      extracted.add(relative)
+              if extracted != selected:
+                  raise RuntimeError("Pinned SDK package is missing required oracle files")
+          with Path(os.environ["GITHUB_ENV"]).open("a") as environment:
+              environment.write("ANDROID_HOME=" + str(root) + "\n")
+          PYTHON
+      - run: npm run verify:apk-signature -- --require-tools

.github/workflows/real-binary-layout.yml

@@ -14,6 +14,15 @@ on:
         required: false
         type: string
         default: ""
+      remote_cache_write:
+        description: Trusted release qualification may publish the remote cache.
+        required: false
+        type: boolean
+        default: false
+    secrets:
+      TURBO_TOKEN:
+        description: Remote cache token. Only trusted callers pass it.
+        required: false
 
 permissions:
   contents: read
@@ -21,6 +30,13 @@ permissions:
 env:
   NODE_OPTIONS: --max-old-space-size=768
   HUSKY: "0"
+  # A called workflow sees the caller event, so workflow_call is not an event
+  # name here. Writes when remote_cache_write is set, or on the default branch
+  # for push, schedule, and manual dispatch. Open pull requests stay empty-token
+  # and use the local cache only.
+  TURBO_TOKEN: "${{ (inputs.remote_cache_write == true || (github.ref == format('refs/heads/{0}', github.event.repository.default_branch) && (github.event_name == 'push' || github.event_name == 'schedule' || github.event_name == 'workflow_dispatch'))) && secrets.TURBO_TOKEN || '' }}"
+  TURBO_TEAM: ${{ vars.TURBO_TEAM }}
+  TURBO_CACHE: "${{ (inputs.remote_cache_write == true || (github.ref == format('refs/heads/{0}', github.event.repository.default_branch) && (github.event_name == 'push' || github.event_name == 'schedule' || github.event_name == 'workflow_dispatch'))) && 'local:rw,remote:rw' || 'local:rw' }}"
   REA_VERIFY_LARGE_ELF_MCP: ${{ inputs.large_mcp && '1' || '0' }}
 
 jobs:

.github/workflows/real-browser.yml

@@ -8,10 +8,26 @@ on:
         required: false
         type: string
         default: ""
+      remote_cache_write:
+        description: Trusted release qualification may publish the remote cache.
+        required: false
+        type: boolean
+        default: false
+    secrets:
+      TURBO_TOKEN:
+        description: Remote cache token. Only trusted callers pass it.
+        required: false
   workflow_dispatch:
 
 env:
   HUSKY: "0"
+  # A called workflow sees the caller event, so workflow_call is not an event
+  # name here. Writes when remote_cache_write is set, or on the default branch
+  # for push, schedule, and manual dispatch. Open pull requests stay empty-token
+  # and use the local cache only.
+  TURBO_TOKEN: "${{ (inputs.remote_cache_write == true || (github.ref == format('refs/heads/{0}', github.event.repository.default_branch) && (github.event_name == 'push' || github.event_name == 'schedule' || github.event_name == 'workflow_dispatch'))) && secrets.TURBO_TOKEN || '' }}"
+  TURBO_TEAM: ${{ vars.TURBO_TEAM }}
+  TURBO_CACHE: "${{ (inputs.remote_cache_write == true || (github.ref == format('refs/heads/{0}', github.event.repository.default_branch) && (github.event_name == 'push' || github.event_name == 'schedule' || github.event_name == 'workflow_dispatch'))) && 'local:rw,remote:rw' || 'local:rw' }}"
 
 permissions:
   contents: read

.github/workflows/real-cutter.yml

@@ -8,6 +8,14 @@ on:
 permissions:
   contents: read
 
+env:
+  HUSKY: "0"
+  # Writes only on the default branch for push, schedule, manual dispatch,
+  # and workflow_call. Pull requests get an empty token and remote read-only mode.
+  TURBO_TOKEN: "${{ (github.ref == format('refs/heads/{0}', github.event.repository.default_branch) && (github.event_name == 'push' || github.event_name == 'schedule' || github.event_name == 'workflow_dispatch' || github.event_name == 'workflow_call')) && secrets.TURBO_TOKEN || '' }}"
+  TURBO_TEAM: ${{ vars.TURBO_TEAM }}
+  TURBO_CACHE: "${{ (github.ref == format('refs/heads/{0}', github.event.repository.default_branch) && (github.event_name == 'push' || github.event_name == 'schedule' || github.event_name == 'workflow_dispatch' || github.event_name == 'workflow_call')) && 'local:rw,remote:rw' || 'local:rw,remote:r' }}"
+
 jobs:
   windows-native-artifact:
     name: Build Windows bridge security boundary
@@ -22,7 +30,7 @@ jobs:
           sudo apt-get update
           sudo apt-get install --no-install-recommends -y g++-mingw-w64-x86-64
       - run: npm run build:windows-native
-      - uses: actions/upload-artifact@v4
+      - uses: actions/upload-artifact@ea165f8d65b6e75b540449e92b4886f43607fa02 # v4.6.2
         with:
           name: cutter-windows-native
           path: native/windows/build/
@@ -53,7 +61,7 @@ jobs:
         uses: ./.github/actions/npm-cache
       - run: npm ci
       - if: matrix.platform == 'Windows'
-        uses: actions/download-artifact@v4
+        uses: actions/download-artifact@d3f86a106a0bac45b974a628896c90dbdf5c8093 # v4.3.0
         with:
           name: cutter-windows-native
           path: native/windows/build/
@@ -67,7 +75,7 @@ jobs:
         env:
           EXPECTED_PLATFORM: ${{ matrix.platform }}
         run: npm run verify:cutter
-      - uses: actions/upload-artifact@v4
+      - uses: actions/upload-artifact@ea165f8d65b6e75b540449e92b4886f43607fa02 # v4.6.2
         if: always()
         with:
           name: cutter-upstream-${{ matrix.platform }}

.github/workflows/real-evm-interface.yml

@@ -9,6 +9,15 @@ on:
         required: false
         type: string
         default: ""
+      remote_cache_write:
+        description: Trusted release qualification may publish the remote cache.
+        required: false
+        type: boolean
+        default: false
+    secrets:
+      TURBO_TOKEN:
+        description: Remote cache token. Only trusted callers pass it.
+        required: false
 
 permissions:
   contents: read
@@ -16,6 +25,13 @@ permissions:
 env:
   NODE_OPTIONS: --max-old-space-size=768
   HUSKY: "0"
+  # A called workflow sees the caller event, so workflow_call is not an event
+  # name here. Writes when remote_cache_write is set, or on the default branch
+  # for push, schedule, and manual dispatch. Open pull requests stay empty-token
+  # and use the local cache only.
+  TURBO_TOKEN: "${{ (inputs.remote_cache_write == true || (github.ref == format('refs/heads/{0}', github.event.repository.default_branch) && (github.event_name == 'push' || github.event_name == 'schedule' || github.event_name == 'workflow_dispatch'))) && secrets.TURBO_TOKEN || '' }}"
+  TURBO_TEAM: ${{ vars.TURBO_TEAM }}
+  TURBO_CACHE: "${{ (inputs.remote_cache_write == true || (github.ref == format('refs/heads/{0}', github.event.repository.default_branch) && (github.event_name == 'push' || github.event_name == 'schedule' || github.event_name == 'workflow_dispatch'))) && 'local:rw,remote:rw' || 'local:rw' }}"
 
 jobs:
   real-evm-interface:

.github/workflows/real-ghidra-windows.yml

@@ -6,6 +6,11 @@ on:
 
 env:
   HUSKY: "0"
+  # Writes only on the default branch for push, schedule, manual dispatch,
+  # and workflow_call. Pull requests get an empty token and remote read-only mode.
+  TURBO_TOKEN: "${{ (github.ref == format('refs/heads/{0}', github.event.repository.default_branch) && (github.event_name == 'push' || github.event_name == 'schedule' || github.event_name == 'workflow_dispatch' || github.event_name == 'workflow_call')) && secrets.TURBO_TOKEN || '' }}"
+  TURBO_TEAM: ${{ vars.TURBO_TEAM }}
+  TURBO_CACHE: "${{ (github.ref == format('refs/heads/{0}', github.event.repository.default_branch) && (github.event_name == 'push' || github.event_name == 'schedule' || github.event_name == 'workflow_dispatch' || github.event_name == 'workflow_call')) && 'local:rw,remote:rw' || 'local:rw,remote:r' }}"
 
 permissions:
   contents: read

.github/workflows/real-go-binary.yml

@@ -13,20 +13,8 @@ on:
       - "src/composition/go.ts"
       - "src/cli/goCommands.ts"
       - "src/server/registerGoTools.ts"
-      - "src/server/createServer.ts"
-      - "src/cli.ts"
-      - "src/cliCommandNames.ts"
-      - "src/contracts/toolContracts.ts"
-      - "src/contracts/toolEffects.ts"
-      - "src/artifacts/readStableArtifact*.ts"
-      - "src/artifacts/ArtifactReader.ts"
-      - "src/process/PrivateRuntimeRoot.ts"
       - "scripts/verify-go-binary.mjs"
       - "scripts/lib/mcp-verifier-results.mjs"
-      - "scripts/rea.mjs"
-      - "package.json"
-      - "package-lock.json"
-      - ".nvmrc"
 
 permissions:
   contents: read
@@ -38,6 +26,11 @@ concurrency:
 env:
   NODE_OPTIONS: --max-old-space-size=768
   HUSKY: "0"
+  # Writes only on the default branch for push, schedule, manual dispatch,
+  # and workflow_call. Pull requests get an empty token and remote read-only mode.
+  TURBO_TOKEN: "${{ (github.ref == format('refs/heads/{0}', github.event.repository.default_branch) && (github.event_name == 'push' || github.event_name == 'schedule' || github.event_name == 'workflow_dispatch' || github.event_name == 'workflow_call')) && secrets.TURBO_TOKEN || '' }}"
+  TURBO_TEAM: ${{ vars.TURBO_TEAM }}
+  TURBO_CACHE: "${{ (github.ref == format('refs/heads/{0}', github.event.repository.default_branch) && (github.event_name == 'push' || github.event_name == 'schedule' || github.event_name == 'workflow_dispatch' || github.event_name == 'workflow_call')) && 'local:rw,remote:rw' || 'local:rw,remote:r' }}"
 
 jobs:
   real-go-binary:

.github/workflows/real-hopper-linux.yml

@@ -5,6 +5,11 @@ on:
 
 env:
   HUSKY: "0"
+  # Writes only on the default branch for push, schedule, manual dispatch,
+  # and workflow_call. Pull requests get an empty token and remote read-only mode.
+  TURBO_TOKEN: "${{ (github.ref == format('refs/heads/{0}', github.event.repository.default_branch) && (github.event_name == 'push' || github.event_name == 'schedule' || github.event_name == 'workflow_dispatch' || github.event_name == 'workflow_call')) && secrets.TURBO_TOKEN || '' }}"
+  TURBO_TEAM: ${{ vars.TURBO_TEAM }}
+  TURBO_CACHE: "${{ (github.ref == format('refs/heads/{0}', github.event.repository.default_branch) && (github.event_name == 'push' || github.event_name == 'schedule' || github.event_name == 'workflow_dispatch' || github.event_name == 'workflow_call')) && 'local:rw,remote:rw' || 'local:rw,remote:r' }}"
 
 permissions:
   contents: read

.github/workflows/real-hopper.yml

@@ -5,6 +5,11 @@ on:
 
 env:
   HUSKY: "0"
+  # Writes only on the default branch for push, schedule, manual dispatch,
+  # and workflow_call. Pull requests get an empty token and remote read-only mode.
+  TURBO_TOKEN: "${{ (github.ref == format('refs/heads/{0}', github.event.repository.default_branch) && (github.event_name == 'push' || github.event_name == 'schedule' || github.event_name == 'workflow_dispatch' || github.event_name == 'workflow_call')) && secrets.TURBO_TOKEN || '' }}"
+  TURBO_TEAM: ${{ vars.TURBO_TEAM }}
+  TURBO_CACHE: "${{ (github.ref == format('refs/heads/{0}', github.event.repository.default_branch) && (github.event_name == 'push' || github.event_name == 'schedule' || github.event_name == 'workflow_dispatch' || github.event_name == 'workflow_call')) && 'local:rw,remote:rw' || 'local:rw,remote:r' }}"
 
 permissions:
   contents: read

.github/workflows/real-javascript-recovery.yml

@@ -8,6 +8,15 @@ on:
         required: false
         type: string
         default: ""
+      remote_cache_write:
+        description: Trusted release qualification may publish the remote cache.
+        required: false
+        type: boolean
+        default: false
+    secrets:
+      TURBO_TOKEN:
+        description: Remote cache token. Only trusted callers pass it.
+        required: false
   workflow_dispatch:
 
 permissions:
@@ -15,6 +24,13 @@ permissions:
 
 env:
   HUSKY: "0"
+  # A called workflow sees the caller event, so workflow_call is not an event
+  # name here. Writes when remote_cache_write is set, or on the default branch
+  # for push, schedule, and manual dispatch. Open pull requests stay empty-token
+  # and use the local cache only.
+  TURBO_TOKEN: "${{ (inputs.remote_cache_write == true || (github.ref == format('refs/heads/{0}', github.event.repository.default_branch) && (github.event_name == 'push' || github.event_name == 'schedule' || github.event_name == 'workflow_dispatch'))) && secrets.TURBO_TOKEN || '' }}"
+  TURBO_TEAM: ${{ vars.TURBO_TEAM }}
+  TURBO_CACHE: "${{ (inputs.remote_cache_write == true || (github.ref == format('refs/heads/{0}', github.event.repository.default_branch) && (github.event_name == 'push' || github.event_name == 'schedule' || github.event_name == 'workflow_dispatch'))) && 'local:rw,remote:rw' || 'local:rw' }}"
   NODE_OPTIONS: "--max-old-space-size=768"
   RAYON_NUM_THREADS: "1"
 

.github/workflows/real-packet-capture.yml

@@ -20,21 +20,17 @@ on:
       - "src/composition/captures.ts"
       - "src/cli/captureCommands.ts"
       - "src/server/registerCaptureTools.ts"
-      - "src/server/createServer.ts"
-      - "src/server/sessionAvailabilityPolicy.ts"
-      - "src/application/CapabilityInventory.ts"
-      - "src/contracts/toolContracts.ts"
-      - "src/contracts/toolEffects.ts"
-      - "src/cli.ts"
-      - "src/cliCommandNames.ts"
-      - "src/process/**"
-      - "src/artifacts/readStableArtifact.ts"
       - "scripts/verify-packet-capture.mjs"
       - ".github/workflows/real-packet-capture.yml"
-      - "package*.json"
-      - ".nvmrc"
 permissions:
   contents: read
+env:
+  HUSKY: "0"
+  # Writes only on the default branch for push, schedule, manual dispatch,
+  # and workflow_call. Pull requests get an empty token and remote read-only mode.
+  TURBO_TOKEN: "${{ (github.ref == format('refs/heads/{0}', github.event.repository.default_branch) && (github.event_name == 'push' || github.event_name == 'schedule' || github.event_name == 'workflow_dispatch' || github.event_name == 'workflow_call')) && secrets.TURBO_TOKEN || '' }}"
+  TURBO_TEAM: ${{ vars.TURBO_TEAM }}
+  TURBO_CACHE: "${{ (github.ref == format('refs/heads/{0}', github.event.repository.default_branch) && (github.event_name == 'push' || github.event_name == 'schedule' || github.event_name == 'workflow_dispatch' || github.event_name == 'workflow_call')) && 'local:rw,remote:rw' || 'local:rw,remote:r' }}"
 concurrency:
   group: real-packet-capture-${{ github.event.pull_request.number || github.ref }}
   cancel-in-progress: true
@@ -53,7 +49,7 @@ jobs:
       - run: sudo apt-get update && sudo DEBIAN_FRONTEND=noninteractive apt-get install -y tshark util-linux python3
       - run: npm ci
       - run: npm run build:cached
-      - run: npm run test:focused -- tests/boundary/captures/fieldDiscovery.test.ts src/process/LinuxResourceLimits.test.ts tests/boundary/captures/captureProvider.test.ts tests/boundary/captures/captureDescriptorCleanup.test.ts tests/boundary/captures/resourceLimits.test.ts tests/boundary/captures/packetOutput.test.ts tests/boundary/mcp/packetCaptureMcp.test.ts tests/boundary/cli/captureCleanup.test.ts
+      - run: npm run test:focused -- tests/boundary/captures/fieldDiscovery.test.ts src/process/LinuxResourceLimits.test.ts tests/boundary/captures/captureProvider.test.ts tests/boundary/captures/captureDescriptorCleanup.test.ts tests/boundary/pcap/pcapProvider.test.ts tests/boundary/pcap/pcapProviderCrossPlatform.test.ts tests/boundary/captures/resourceLimits.test.ts tests/boundary/captures/packetOutput.test.ts tests/boundary/mcp/packetCaptureMcp.test.ts tests/boundary/cli/captureCleanup.test.ts
       - run: node scripts/verify-packet-capture.mjs
         env:
           REA_TSHARK_COMMAND: /usr/bin/tshark

.github/workflows/real-recorded-crash.yml

@@ -9,6 +9,15 @@ on:
         required: false
         type: string
         default: ""
+      remote_cache_write:
+        description: Trusted release qualification may publish the remote cache.
+        required: false
+        type: boolean
+        default: false
+    secrets:
+      TURBO_TOKEN:
+        description: Remote cache token. Only trusted callers pass it.
+        required: false
 
 permissions:
   contents: read
@@ -16,6 +25,13 @@ permissions:
 env:
   NODE_OPTIONS: --max-old-space-size=768
   HUSKY: "0"
+  # A called workflow sees the caller event, so workflow_call is not an event
+  # name here. Writes when remote_cache_write is set, or on the default branch
+  # for push, schedule, and manual dispatch. Open pull requests stay empty-token
+  # and use the local cache only.
+  TURBO_TOKEN: "${{ (inputs.remote_cache_write == true || (github.ref == format('refs/heads/{0}', github.event.repository.default_branch) && (github.event_name == 'push' || github.event_name == 'schedule' || github.event_name == 'workflow_dispatch'))) && secrets.TURBO_TOKEN || '' }}"
+  TURBO_TEAM: ${{ vars.TURBO_TEAM }}
+  TURBO_CACHE: "${{ (inputs.remote_cache_write == true || (github.ref == format('refs/heads/{0}', github.event.repository.default_branch) && (github.event_name == 'push' || github.event_name == 'schedule' || github.event_name == 'workflow_dispatch'))) && 'local:rw,remote:rw' || 'local:rw' }}"
   UV_CONCURRENT_BUILDS: "1"
   UV_CONCURRENT_INSTALLS: "1"
   UV_PYTHON_DOWNLOADS: never
@@ -24,7 +40,19 @@ jobs:
   real-recorded-crash:
     runs-on: ubuntu-24.04
     timeout-minutes: 15
+    strategy:
+      fail-fast: false
+      matrix:
+        image: ["ubuntu:24.04", "debian:trixie-slim"]
+    container: ${{ matrix.image }}
     steps:
+      - name: Install fixture and debugger prerequisites in the disposable container
+        shell: bash
+        run: |
+          apt-get update -qq
+          apt-get install --yes --no-install-recommends \
+            ca-certificates git python3-venv gcc libc6-dev strace gdb
+
       - uses: actions/checkout@11d5960a326750d5838078e36cf38b85af677262 # v4.4.0
       - uses: actions/setup-node@49933ea5288caeca8642d1e84afbd3f7d6820020 # v4.4.0
         with:
@@ -36,11 +64,6 @@ jobs:
       - uses: ./.github/actions/portable-build-cache
         with:
           build_cache_artifact: ${{ inputs.build_cache_artifact }}
-      - name: Install GDB on the disposable CI runner
-        shell: bash
-        run: |
-          sudo apt-get update -qq
-          sudo apt-get install --yes --no-install-recommends gdb
       - name: Acquire exact unchanged upstreams in isolated runner storage
         shell: bash
         run: |

.github/workflows/real-signal-recording.yml

@@ -13,21 +13,17 @@ on:
       - "src/composition/signals.ts"
       - "src/cli/signalCommands.ts"
       - "src/server/registerSignalTools.ts"
-      - "src/server/createServer.ts"
-      - "src/server/sessionAvailabilityPolicy.ts"
-      - "src/application/CapabilityInventory.ts"
-      - "src/contracts/toolContracts.ts"
-      - "src/contracts/toolEffects.ts"
-      - "src/cli.ts"
-      - "src/cliCommandNames.ts"
-      - "src/process/**"
-      - "src/artifacts/readStableArtifact.ts"
       - "scripts/verify-signal-recording.mjs"
       - ".github/workflows/real-signal-recording.yml"
-      - "package*.json"
-      - ".nvmrc"
 permissions:
   contents: read
+env:
+  HUSKY: "0"
+  # Writes only on the default branch for push, schedule, manual dispatch,
+  # and workflow_call. Pull requests get an empty token and remote read-only mode.
+  TURBO_TOKEN: "${{ (github.ref == format('refs/heads/{0}', github.event.repository.default_branch) && (github.event_name == 'push' || github.event_name == 'schedule' || github.event_name == 'workflow_dispatch' || github.event_name == 'workflow_call')) && secrets.TURBO_TOKEN || '' }}"
+  TURBO_TEAM: ${{ vars.TURBO_TEAM }}
+  TURBO_CACHE: "${{ (github.ref == format('refs/heads/{0}', github.event.repository.default_branch) && (github.event_name == 'push' || github.event_name == 'schedule' || github.event_name == 'workflow_dispatch' || github.event_name == 'workflow_call')) && 'local:rw,remote:rw' || 'local:rw,remote:r' }}"
 concurrency:
   group: real-signal-recording-${{ github.event.pull_request.number || github.ref }}
   cancel-in-progress: true

.github/workflows/real-termux.yml

@@ -12,6 +12,11 @@ permissions:
 
 env:
   HUSKY: "0"
+  # Writes only on the default branch for push, schedule, manual dispatch,
+  # and workflow_call. Pull requests get an empty token and remote read-only mode.
+  TURBO_TOKEN: "${{ (github.ref == format('refs/heads/{0}', github.event.repository.default_branch) && (github.event_name == 'push' || github.event_name == 'schedule' || github.event_name == 'workflow_dispatch' || github.event_name == 'workflow_call')) && secrets.TURBO_TOKEN || '' }}"
+  TURBO_TEAM: ${{ vars.TURBO_TEAM }}
+  TURBO_CACHE: "${{ (github.ref == format('refs/heads/{0}', github.event.repository.default_branch) && (github.event_name == 'push' || github.event_name == 'schedule' || github.event_name == 'workflow_dispatch' || github.event_name == 'workflow_call')) && 'local:rw,remote:rw' || 'local:rw,remote:r' }}"
 
 jobs:
   real-termux:

.github/workflows/real-web-network-captures.yml

@@ -9,12 +9,28 @@ on:
         required: false
         type: string
         default: ""
+      remote_cache_write:
+        description: Trusted release qualification may publish the remote cache.
+        required: false
+        type: boolean
+        default: false
+    secrets:
+      TURBO_TOKEN:
+        description: Remote cache token. Only trusted callers pass it.
+        required: false
 
 permissions:
   contents: read
 
 env:
   HUSKY: "0"
+  # A called workflow sees the caller event, so workflow_call is not an event
+  # name here. Writes when remote_cache_write is set, or on the default branch
+  # for push, schedule, and manual dispatch. Open pull requests stay empty-token
+  # and use the local cache only.
+  TURBO_TOKEN: "${{ (inputs.remote_cache_write == true || (github.ref == format('refs/heads/{0}', github.event.repository.default_branch) && (github.event_name == 'push' || github.event_name == 'schedule' || github.event_name == 'workflow_dispatch'))) && secrets.TURBO_TOKEN || '' }}"
+  TURBO_TEAM: ${{ vars.TURBO_TEAM }}
+  TURBO_CACHE: "${{ (inputs.remote_cache_write == true || (github.ref == format('refs/heads/{0}', github.event.repository.default_branch) && (github.event_name == 'push' || github.event_name == 'schedule' || github.event_name == 'workflow_dispatch'))) && 'local:rw,remote:rw' || 'local:rw' }}"
   NODE_OPTIONS: --max-old-space-size=768
 
 jobs:

.github/workflows/real-web-runtime.yml

@@ -7,11 +7,27 @@ on:
         required: false
         type: string
         default: ""
+      remote_cache_write:
+        description: Trusted release qualification may publish the remote cache.
+        required: false
+        type: boolean
+        default: false
+    secrets:
+      TURBO_TOKEN:
+        description: Remote cache token. Only trusted callers pass it.
+        required: false
   workflow_dispatch:
 permissions:
   contents: read
 env:
   HUSKY: "0"
+  # A called workflow sees the caller event, so workflow_call is not an event
+  # name here. Writes when remote_cache_write is set, or on the default branch
+  # for push, schedule, and manual dispatch. Open pull requests stay empty-token
+  # and use the local cache only.
+  TURBO_TOKEN: "${{ (inputs.remote_cache_write == true || (github.ref == format('refs/heads/{0}', github.event.repository.default_branch) && (github.event_name == 'push' || github.event_name == 'schedule' || github.event_name == 'workflow_dispatch'))) && secrets.TURBO_TOKEN || '' }}"
+  TURBO_TEAM: ${{ vars.TURBO_TEAM }}
+  TURBO_CACHE: "${{ (inputs.remote_cache_write == true || (github.ref == format('refs/heads/{0}', github.event.repository.default_branch) && (github.event_name == 'push' || github.event_name == 'schedule' || github.event_name == 'workflow_dispatch'))) && 'local:rw,remote:rw' || 'local:rw' }}"
 jobs:
   real-web-runtime:
     runs-on: ubuntu-latest

.github/workflows/real-web-source-map.yml

@@ -7,11 +7,27 @@ on:
         required: false
         type: string
         default: ""
+      remote_cache_write:
+        description: Trusted release qualification may publish the remote cache.
+        required: false
+        type: boolean
+        default: false
+    secrets:
+      TURBO_TOKEN:
+        description: Remote cache token. Only trusted callers pass it.
+        required: false
   workflow_dispatch:
 permissions:
   contents: read
 env:
   HUSKY: "0"
+  # A called workflow sees the caller event, so workflow_call is not an event
+  # name here. Writes when remote_cache_write is set, or on the default branch
+  # for push, schedule, and manual dispatch. Open pull requests stay empty-token
+  # and use the local cache only.
+  TURBO_TOKEN: "${{ (inputs.remote_cache_write == true || (github.ref == format('refs/heads/{0}', github.event.repository.default_branch) && (github.event_name == 'push' || github.event_name == 'schedule' || github.event_name == 'workflow_dispatch'))) && secrets.TURBO_TOKEN || '' }}"
+  TURBO_TEAM: ${{ vars.TURBO_TEAM }}
+  TURBO_CACHE: "${{ (inputs.remote_cache_write == true || (github.ref == format('refs/heads/{0}', github.event.repository.default_branch) && (github.event_name == 'push' || github.event_name == 'schedule' || github.event_name == 'workflow_dispatch'))) && 'local:rw,remote:rw' || 'local:rw' }}"
 jobs:
   real-web-source-map:
     runs-on: ubuntu-latest

.github/workflows/release.yml

@@ -21,6 +21,12 @@ on:
 
 env:
   HUSKY: "0"
+  # Remote cache stays local when either value is empty.
+  # Writes only on the default branch for push, schedule, manual dispatch,
+  # and workflow_call. Pull requests get an empty token and remote read-only mode.
+  TURBO_TOKEN: "${{ (github.ref == format('refs/heads/{0}', github.event.repository.default_branch) && (github.event_name == 'push' || github.event_name == 'schedule' || github.event_name == 'workflow_dispatch' || github.event_name == 'workflow_call')) && secrets.TURBO_TOKEN || '' }}"
+  TURBO_TEAM: ${{ vars.TURBO_TEAM }}
+  TURBO_CACHE: "${{ (github.ref == format('refs/heads/{0}', github.event.repository.default_branch) && (github.event_name == 'push' || github.event_name == 'schedule' || github.event_name == 'workflow_dispatch' || github.event_name == 'workflow_call')) && 'local:rw,remote:rw' || 'local:rw,remote:r' }}"
 
 concurrency:
   group: release-${{ inputs.release_branch || github.event.pull_request.merge_commit_sha || github.ref_name }}
@@ -301,11 +307,17 @@ jobs:
       contents: read
       actions: read
     uses: ./.github/workflows/ci.yml
+    secrets:
+      TURBO_TOKEN: ${{ secrets.TURBO_TOKEN }}
     with:
       source_sha: ${{ needs.merged-release.outputs.sha || needs.release-please.outputs.sha }}
+      remote_cache_write: true
 
   rehearsal:
     name: Rehearse qualified release on a clean runner
+    env:
+      TURBO_TOKEN: ""
+      TURBO_CACHE: local:rw
     needs: [merged-release, release-please, qualification]
     if: >-
       always() && !cancelled() && needs.qualification.result == 'success'
@@ -339,6 +351,9 @@ jobs:
       always() && !cancelled() &&
       needs.qualification.result == 'success' && needs.rehearsal.result == 'success'
     runs-on: ubuntu-latest
+    env:
+      TURBO_TOKEN: ""
+      TURBO_CACHE: local:rw
     permissions:
       contents: read
       id-token: write
@@ -347,41 +362,56 @@ jobs:
         with:
           ref: ${{ needs.merged-release.outputs.sha || needs.release-please.outputs.sha }}
           persist-credentials: false
-
       - uses: actions/setup-node@249970729cb0ef3589644e2896645e5dc5ba9c38 # v6.5.0
         with:
           node-version-file: .nvmrc
           package-manager-cache: false
           registry-url: https://registry.npmjs.org
+      - name: Download the original qualified package
+        uses: actions/download-artifact@d3f86a106a0bac45b974a628896c90dbdf5c8093 # v4.3.0
+        with:
+          name: verified-release-package
+          path: .cache/verified-package/
+      - name: Bind publication to qualified intent
+        id: package
+        run: node scripts/ci/prepare-npm-publication.mjs
+      - name: Publish
+        if: steps.package.outputs.published == 'false'
+        env:
+          NPM_RELEASE_TAG: ${{ steps.package.outputs.tag }}
+        run: >-
+          npm publish .cache/verified-package/package.tgz --ignore-scripts
+          --registry=https://registry.npmjs.org --access=public --tag="$NPM_RELEASE_TAG"
 
+  verify-npm:
+    name: Verify published npm package
+    needs: [merged-release, release-please, publish]
+    if: >-
+      always() && !cancelled() && needs.publish.result == 'success'
+    runs-on: ubuntu-latest
+    env:
+      TURBO_TOKEN: ""
+      TURBO_CACHE: local:rw
+    permissions:
+      contents: read
+      actions: read
+    steps:
+      - uses: actions/checkout@d23441a48e516b6c34aea4fa41551a30e30af803 # v6.1.0
+        with:
+          ref: ${{ needs.merged-release.outputs.sha || needs.release-please.outputs.sha }}
+          persist-credentials: false
+      - uses: actions/setup-node@249970729cb0ef3589644e2896645e5dc5ba9c38 # v6.5.0
+        with:
+          node-version-file: .nvmrc
+          package-manager-cache: false
       - run: npm ci
-
       - uses: ./.github/actions/prepare-release-package
         id: package
-
-      - name: Inspect existing npm version
-        id: registry
-        env:
-          VERIFIED_PACKAGE_TARBALL: ${{ steps.package.outputs.tarball }}
-          VERIFIED_PACKAGE_VERSION: ${{ steps.package.outputs.version }}
-        run: node scripts/ci/npm-release.mjs inspect "$VERIFIED_PACKAGE_TARBALL" "$VERIFIED_PACKAGE_VERSION"
-
-      - name: Publish
-        if: steps.registry.outputs.published == 'false'
-        shell: bash
-        env:
-          VERIFIED_PACKAGE_TARBALL: ${{ steps.package.outputs.tarball }}
-          VERIFIED_PACKAGE_VERSION: ${{ steps.package.outputs.version }}
-        run: |
-          tag="$(node scripts/release-npm-tag.mjs "$VERIFIED_PACKAGE_VERSION")"
-          npm publish "${VERIFIED_PACKAGE_TARBALL}" --access public --tag "${tag}"
-
       - name: Verify published npm bytes and dist-tag
         env:
           VERIFIED_PACKAGE_TARBALL: ${{ steps.package.outputs.tarball }}
           VERIFIED_PACKAGE_VERSION: ${{ steps.package.outputs.version }}
         run: node scripts/ci/npm-release.mjs verify "$VERIFIED_PACKAGE_TARBALL" "$VERIFIED_PACKAGE_VERSION"
-
       - name: Verify published CLI from npm
         env:
           VERIFIED_PACKAGE_VERSION: ${{ steps.package.outputs.version }}
@@ -389,9 +419,12 @@ jobs:
 
   publish-mcp:
     name: Publish to MCP Registry
-    needs: [merged-release, release-please, publish]
+    env:
+      TURBO_TOKEN: ""
+      TURBO_CACHE: local:rw
+    needs: [merged-release, release-please, verify-npm]
     if: >-
-      always() && !cancelled() && needs.publish.result == 'success' &&
+      always() && !cancelled() && needs.verify-npm.result == 'success' &&
       ((needs.merged-release.result == 'success' && needs.merged-release.outputs.release_created == 'true') ||
       (needs.release-please.result == 'success' && inputs.phase == 'publish' && needs.release-please.outputs.release_created == 'true'))
     runs-on: ubuntu-latest

.github/workflows/sqlite-runtime.yml

@@ -20,7 +20,7 @@ jobs:
         # Exercise unavailable-runtime diagnostics, the first supported release,
         # and new patches on both supported lines without hand-editing versions.
         # The repository-pinned Node 24 version also runs the complete test suite.
-        node: ["22.x", "24.15.0", "24.x", "26.x"]
+        node: ["22.x", "24.22.1", "24.x", "26.x"]
     steps:
       - uses: actions/checkout@11d5960a326750d5838078e36cf38b85af677262 # v4.4.0
       - uses: actions/setup-node@49933ea5288caeca8642d1e84afbd3f7d6820020 # v4.4.0

.nvmrc

@@ -1 +1 @@
-24.18.0
+24.22.1

.release-please-manifest.json

@@ -1,3 +1,3 @@
 {
-  ".": "6.5.0"
+  ".": "6.7.0"
 }

AGENTS.md

@@ -20,6 +20,10 @@ Give each interpretation and resource a clear owner. Consumers must use that own
 
 Keep proof of ownership separate from validity, editability, and readiness. Malformed owned data may still be safely removable. Unresolved identity or applicability must remain unknown rather than be treated as absent, foreign, or aligned.
 
+For lifecycle changes, follow [ownership and completion invariants](docs/tool-design.md#own-the-lifecycle) and select the applicable [lifecycle regressions](docs/testing.md#lifecycle-regressions). Include late handshakes, cancellation during final cleanup, mixed child outcomes, and cleanup retry where those stages exist; active-execution cancellation alone does not cover finalization.
+
+Before adding complexity for a speculative or vanishingly unlikely race or security edge case, explain the concrete failure mode, likelihood, and complexity cost to the user and get their buy-in. Do not silently expand scope for theoretical robustness.
+
 Assess brittleness in touched code during implementation, debugging, review, and maintenance. Fix the producing representation or ownership rule before adding downstream repairs. Normalize once at the owning boundary using format-aware parsers. Remove superseded representations, aliases, wrappers, and redundant tests after verifying affected consumers. Introduce abstractions only when they remove observed complexity.
 
 When fixing a bug, inspect the producing boundary and affected consumers for duplicated interpretation or unclear ownership. Complete any directly related refactor in the same change and remove superseded code. Passing regression tests alone does not establish that the brittleness is resolved. If related cleanup remains, explain what remains and why.
@@ -57,7 +61,7 @@ do not establish fewer tokens or faster analysis.
 
 ## Development
 
-Use [.nvmrc](.nvmrc) and `package.json#packageManager` for the pinned toolchain. Cursor Cloud may place older Node.js first on `PATH`; environment setup installs the pinned toolchain under `/usr/local/bin`.
+Use [.nvmrc](.nvmrc) and `package.json#packageManager` for the pinned toolchain. Cursor Cloud may place older Node.js first on `PATH`; environment setup installs the pinned toolchain under `/usr/local/bin`. Prefix `PATH="/usr/local/bin:$PATH"` before `node` and `npm`.
 
 - `npm ci`: install locked dependencies.
 - `npm run build:cached`: build the runtime and packaged skill.
@@ -68,6 +72,8 @@ Use [.nvmrc](.nvmrc) and `package.json#packageManager` for the pinned toolchain.
 
 Use relevant checks for the change. The complete local gate, `npm run check:pr`, is optional for broad changes or CI diagnosis. See [CONTRIBUTING.md](CONTRIBUTING.md) for development, generated-file ownership, and release conventions.
 
+Coordinate expensive checks across agents. If a test unexpectedly consumes heavy CPU, stop its owned process family and profile a bounded representative case before retrying. Inspect production work, fixture generation, subprocesses, and repeated setup; preserve regression coverage when optimizing. Follow the [test resource guidance](docs/testing.md#developer-commands).
+
 `docs/public/product-catalog.json`, `docs/verification/managed-conformance-*.json`, and `skills/` are ignored outputs. Edit source contracts and `.agents/skills/reverse-engineer-anything/`, then regenerate as needed. `.cache/mcp-tool-catalog.json` is test metadata; runtime builds and source checking do not consume it. Never commit binaries, provider project documents, credentials, `dist/`, `node_modules/`, or local planning artifacts.
 
 ## Tool and Test Changes

CHANGELOG.md

@@ -1,5 +1,93 @@
 # Changelog
 
+## [6.7.0](https://github.com/morluto/rea/compare/rea-agents-6.6.0...rea-agents-6.7.0) (2026-10-18)
+
+### Highlights
+
+- Client configuration edits follow the duplicate-member semantics used by MCP clients.
+
+### Changes
+
+- Includes the [6.6.0 changes and migration notes](https://github.com/morluto/rea/releases/tag/rea-agents-6.6.0). That version was tagged but was not published to npm.
+
+### Fixes
+
+- **Client configuration:** Setup edits the last repeated JSON/JSONC member, matching the value clients actually read. Setup and uninstall refuse a removal that would reactivate a shadowed duplicate, preserving the original configuration. ([#4957](https://github.com/morluto/rea/pull/4957), [morluto/rea#4952](https://github.com/morluto/rea/issues/4952)). Thanks @kaoru0822-kitauji, @xu91102.
+
+Full contribution record: [4 commits and 1 merged PR](https://github.com/morluto/rea/blob/rea-agents-6.7.0/docs/releases/6.7.0.contributions.json).
+
+## [6.6.0](https://github.com/morluto/rea/compare/rea-agents-6.5.0...rea-agents-6.6.0) (2026-10-18)
+
+### Highlights
+
+- Add Kuna as a caller-supplied static-analysis provider.
+- Query selected JavaScript facts and multiple APK locale tables without repeating the underlying analysis.
+- Extend Ghidra analysis to Windows ARM PE targets and explicit raw m68k images.
+- Preserve completed Evidence, cancellation reasons and retryable cleanup across provider and CLI finalization.
+
+### Changes
+
+- **Kuna analysis:** Use a caller-installed Kuna for procedure listings, pseudocode, assembly, call relationships, exact-address xrefs and cached string searches. Project exports retain source spans and digests; repeated requests reuse completed work. Recorded real-provider coverage uses Kuna 1.833 on Linux; Windows, PE and universal Mach-O are unsupported. See [the provider guide](docs/kuna-provider.md). ([#3832](https://github.com/morluto/rea/pull/3832), [#3854](https://github.com/morluto/rea/pull/3854), [#4071](https://github.com/morluto/rea/pull/4071), [#4095](https://github.com/morluto/rea/pull/4095)). Thanks @gensokyo-dev, @kaoru0822-kitauji, @N0zoM1z0.
+- **Retained JavaScript facts:** Query exact application paths, selected properties and filtered module references through `inspect_analysis_view`, with paging and complete row Evidence. These views reuse completed analysis and preserve source selection and unknown coverage. ([#4265](https://github.com/morluto/rea/pull/4265), [morluto/rea#4190](https://github.com/morluto/rea/issues/4190), [#4097](https://github.com/morluto/rea/pull/4097), [morluto/rea#4092](https://github.com/morluto/rea/issues/4092)). Thanks @kaoru0822-kitauji, @N0zoM1z0.
+- **APK locale tables:** Request several resource locales in one decode with repeatable CLI `--locales` or MCP `locales`. Each table reports its status and coverage; existing single-locale fields remain available. ([#4804](https://github.com/morluto/rea/pull/4804), [morluto/rea#4777](https://github.com/morluto/rea/issues/4777)). Thanks @abhay-2108, @N0zoM1z0.
+- **Windows ARM binaries:** Ghidra admits ARMNT/Thumb-2 and ARM64 PE applications and DLLs on the Windows x64 host lane. Recorded Ghidra 12.1.4 verification uses `WindowsResourceReference` disabled and raised startup budgets; default-analyzer completion, ARM64EC/ARM64X and Windows-on-Arm hosts remain unverified or unsupported. See [the exact scope](docs/windows-ghidra-p0.md#native-authority-and-target-identity). ([#3965](https://github.com/morluto/rea/pull/3965), [morluto/rea#2186](https://github.com/morluto/rea/issues/2186)). Thanks @N0zoM1z0, @WestLabsAI.
+- **Raw m68k images:** Analyze explicitly selected big-endian m68k raw binaries using Ghidra’s fixed 68000 profile and a caller-supplied aligned entry. Real CLI/MCP verification used Ghidra 12.1.4 on macOS arm64; Linux is unverified and Windows P0 does not admit m68k. See [mapping requirements](docs/raw-binary.md). ([#4018](https://github.com/morluto/rea/pull/4018)). Thanks @icedac.
+- **JADX deadlines:** Set `REA_JADX_OPERATION_TIMEOUT_MS` to extend APK loading and upstream operation deadlines together. Doctor reports the effective value and setup forwards an explicit selection. ([#4466](https://github.com/morluto/rea/pull/4466), [morluto/rea#3918](https://github.com/morluto/rea/issues/3918)). Thanks @jaychou813, @SandstromPL.
+- **Electron window options:** Resolve stable local constant `BrowserWindow` options while retaining source provenance. Relevant writes and escapes keep the result unknown. ([#4400](https://github.com/morluto/rea/pull/4400), [morluto/rea#4366](https://github.com/morluto/rea/issues/4366)). Thanks @kaoru0822-kitauji, @N0zoM1z0.
+- **Language editions:** Add the Greek website edition and Malayalam documentation, with Korean typography and translation corrections across the site. ([#3520](https://github.com/morluto/rea/pull/3520), [#4792](https://github.com/morluto/rea/pull/4792), [#3821](https://github.com/morluto/rea/pull/3821), [morluto/rea#3820](https://github.com/morluto/rea/issues/3820), [#3799](https://github.com/morluto/rea/pull/3799), [morluto/rea#3798](https://github.com/morluto/rea/issues/3798), [#4344](https://github.com/morluto/rea/pull/4344), [#3856](https://github.com/morluto/rea/pull/3856), [#4166](https://github.com/morluto/rea/pull/4166)). Thanks @gensokyo-dev, @hi2242, @kalayciburak, @kaoru0822-kitauji, @kraken-503, @kukinghan, @N0zoM1z0, @yeseong0412.
+
+### Fixes
+
+- **Cancellation and cleanup:** Keep completed observations, exact cancellation reasons and the original physical failure through finalization. Failed cleanup retains its process or file owner for retry; recovered resources disappear from pending-cleanup reports without replacing the original error. ([#4238](https://github.com/morluto/rea/pull/4238), [morluto/rea#4232](https://github.com/morluto/rea/issues/4232), [#4274](https://github.com/morluto/rea/pull/4274), [morluto/rea#3923](https://github.com/morluto/rea/issues/3923), [3e82366](https://github.com/morluto/rea/commit/3e8236639abffa39deb5d5e939ec8baf6aaf3ca3), [16924e2](https://github.com/morluto/rea/commit/16924e275e939d0543ba01493324959a9e1d731c), [bbfe6ed](https://github.com/morluto/rea/commit/bbfe6edd60290f9bfbec68c8a0bcf605d8bbf5b3), [1f23e97](https://github.com/morluto/rea/commit/1f23e977b50301e91f02a4a17c9f8d48bb7b60f3), [08fbeb8](https://github.com/morluto/rea/commit/08fbeb8bdab27118b6c57039aef7c0fa8b6463bc)). Thanks @coygeek.
+- **Provider startup:** Join prerequisite probes and in-flight readiness work before shutdown, preserve typed timeout and JVM-memory failures, and retain startup diagnostics through cleanup. Ghidra recognizes keg-only Homebrew JDKs and names the setting responsible for a startup timeout. ([376eeca](https://github.com/morluto/rea/commit/376eeca770820b31f7f8d4c5fcc397315621643e), [ceeff5a](https://github.com/morluto/rea/commit/ceeff5a41788e47e5aee96a4baf191a6ce7fde4b), [3ad1032](https://github.com/morluto/rea/commit/3ad1032b11edea2dcb054e78f951cc43159ae83d), [964decd](https://github.com/morluto/rea/commit/964decd989acfa84c4bfc3295362c4c46877fc20), [3fa972a](https://github.com/morluto/rea/commit/3fa972af7c7cb6cf014ffc426bdf7130dc6f2fd2), [#4557](https://github.com/morluto/rea/pull/4557), [morluto/rea#4556](https://github.com/morluto/rea/issues/4556), [#4590](https://github.com/morluto/rea/pull/4590), [morluto/rea#4241](https://github.com/morluto/rea/issues/4241)). Thanks @DanielCvz1, @J0UH, @Rinkia, @RishTheFish21, @vadlachary2002.
+- **Client configuration:** Preserve the selected launcher, destination file, caller timeout choices and unresolved configuration evidence. Setup/uninstall recovery stays bound to the exact published file, keeps current verified backups, and transfers failed cleanup ownership for retry. ([d0253c9](https://github.com/morluto/rea/commit/d0253c99d65afbbd12732a2fdaf711640574e17f), [7e4eed9](https://github.com/morluto/rea/commit/7e4eed9dd48485b6a7100300aa98401811c2deb9), [2fb5e20](https://github.com/morluto/rea/commit/2fb5e204a35723f6f843d56ca47b0db94d8fcc44), [8235210](https://github.com/morluto/rea/commit/823521085a28f825e7fee1929d852d420042d1f7), [d818c3d](https://github.com/morluto/rea/commit/d818c3d2639d3115edb8a5269a754101aea6c3bd), [19329e4](https://github.com/morluto/rea/commit/19329e43bb38ca9d64512d801b25abcb188d2512), [dadf909](https://github.com/morluto/rea/commit/dadf909dce2e8b3d055541de69b4118d21276316), [333e0ab](https://github.com/morluto/rea/commit/333e0ab0dfc126a9915907d5079ed6195190e38e), [c15c206](https://github.com/morluto/rea/commit/c15c2062ad815cc9a3222ccfaeb7ef53cacd8b72)). Thanks @Adolanium.
+- **Configuration formats:** Read signed 64-bit TOML integers without rounding, handle large TOML documents without argument-spread failures, preserve JSONC trivia and Hermes YAML scalar text, and reject NUL bytes in registration environment values. ([#4937](https://github.com/morluto/rea/pull/4937), [morluto/rea#4917](https://github.com/morluto/rea/issues/4917), [c43bc0b](https://github.com/morluto/rea/commit/c43bc0bf66a4e245a0cafa40def0d0207bcbd73c), [4fd28fc](https://github.com/morluto/rea/commit/4fd28fc9e0c4f8358147af38c72ab9e814f52d61), [#3568](https://github.com/morluto/rea/pull/3568), [#4335](https://github.com/morluto/rea/pull/4335)). Thanks @alfredo1491, @gensokyo-dev, @kaoru0822-kitauji, @N0zoM1z0, @ShuaixinHuang, @xu91102.
+- **Claude Code configuration:** Setup, doctor and uninstall now accept only the strict JSON format Claude Code reads, including a leading UTF-8 BOM. Comments and trailing commas are refused before backup or writing; clients that support JSONC retain their existing behavior. ([#4944](https://github.com/morluto/rea/pull/4944), [morluto/rea#4940](https://github.com/morluto/rea/issues/4940)). Thanks @kaoru0822-kitauji, @xu91102.
+- **CLI output and dispatch:** Accept `--flag=value` and spaced global output values, keep array positions in filtered JSON, validate integer tokens before coercion, and exit quietly when the output reader closes. Cancellation remains effective through command finalization; app bundles and static native inspections select the intended workflow. ([#3811](https://github.com/morluto/rea/pull/3811), [#4823](https://github.com/morluto/rea/pull/4823), [2ae4e16](https://github.com/morluto/rea/commit/2ae4e1608f290c95ff7cff28c4d16a79b301d03d), [8d82579](https://github.com/morluto/rea/commit/8d8257928bbe8d43fedffc4db5fca26c8108820f), [#3781](https://github.com/morluto/rea/pull/3781), [279c590](https://github.com/morluto/rea/commit/279c590ef6bb737d4a48ea696a0ae5eb30010b3d), [#3511](https://github.com/morluto/rea/pull/3511), [morluto/rea#3321](https://github.com/morluto/rea/issues/3321), [f31b231](https://github.com/morluto/rea/commit/f31b231a92a61be6bd1800b16be0b3210ae82944)). Thanks @arjunprabhulal, @coygeek, @kaoru0822-kitauji, @ShuaixinHuang, @Sriram-PR, @Tapraj-Pathak.
+- **Updates:** Route `--update` through the verified updater and bound retained package-manager diagnostics. ([4f7b5fb](https://github.com/morluto/rea/commit/4f7b5fb6023ce2ec449561782d09e3c7e5ac7c7b), [01cdbbd](https://github.com/morluto/rea/commit/01cdbbdc556262c47d04a5b3391dd9c17946d3fc)).
+- **MCP contracts:** Advertise self-contained tool inputs while preserving tuple, array, Boolean and nested-JSON constraints across schema profiles. Schema-reference handling no longer rewrites literal metadata; catalog identity includes the output-schema profile. ([5e30969](https://github.com/morluto/rea/commit/5e30969cc5f1255d44ec7a1f1ca8dbac0af5e7c4), [10aa74d](https://github.com/morluto/rea/commit/10aa74dfedc3ae776f50d1fa1eb83b64ae82d3d6), [204b9f3](https://github.com/morluto/rea/commit/204b9f3ac140a993c3e458bb0e117de31818ab29), [b3389a5](https://github.com/morluto/rea/commit/b3389a5b0834265fe7a40391be315d810a53a97e), [a2e3016](https://github.com/morluto/rea/commit/a2e30168a046f3b79319eb7fbaf807b802104199), [ba98130](https://github.com/morluto/rea/commit/ba98130049fa40ade66b03611ff70c7ddc011a1d), [c5fb86d](https://github.com/morluto/rea/commit/c5fb86d1948af0a287a581b29b2b1d554743c607)). Thanks @akushonkamen, @HarsH-1877, @kaoru0822-kitauji, @xingzihai.
+- **Session ordering:** Keep snapshot admission ahead of later calls on the same connection, retain shutdown signals until cleanup succeeds, and preserve inspection authority through investigation publication. ([#3661](https://github.com/morluto/rea/pull/3661), [morluto/rea#3334](https://github.com/morluto/rea/issues/3334), [8873b39](https://github.com/morluto/rea/commit/8873b39770e89eae22b7693bc492d7e423b22bea), [663d07f](https://github.com/morluto/rea/commit/663d07f02cd62c99d81c59391d2c6d745c875335)). Thanks @coygeek, @rudycelekli, @Swatantra-66.
+- **JavaScript exports and resolution:** Retain CommonJS exports created with `Object.defineProperty` and track export receiver lifetimes. Resolve scoped packages, package `main`, JSON requires and `.mts`/`.cts` imports using the owning loader; overwritten `require` bindings and erased type-only imports no longer create false runtime links. ([#4725](https://github.com/morluto/rea/pull/4725), [morluto/rea#4724](https://github.com/morluto/rea/issues/4724), [#3567](https://github.com/morluto/rea/pull/3567), [morluto/rea#3541](https://github.com/morluto/rea/issues/3541), [#4838](https://github.com/morluto/rea/pull/4838), [morluto/rea#4632](https://github.com/morluto/rea/issues/4632), [#3844](https://github.com/morluto/rea/pull/3844), [morluto/rea#3383](https://github.com/morluto/rea/issues/3383), [#4164](https://github.com/morluto/rea/pull/4164), [morluto/rea#3915](https://github.com/morluto/rea/issues/3915), [#4223](https://github.com/morluto/rea/pull/4223), [#3801](https://github.com/morluto/rea/pull/3801), [morluto/rea#3393](https://github.com/morluto/rea/issues/3393), [#4165](https://github.com/morluto/rea/pull/4165), [morluto/rea#3932](https://github.com/morluto/rea/issues/3932)). Thanks @Abishek0070, @coygeek, @JNTMTMTM, @kaoru0822-kitauji, @prasad2512, @rodricksz4h5, @xu91102.
+- **JavaScript value flow:** Preserve values captured before later assignments, guarded factory fallthrough and nested destructuring. Track exact property keys and setter writes; direct `eval`, compound assignments and incoming callbacks preserve unresolved effects instead of inventing certainty. ([#4196](https://github.com/morluto/rea/pull/4196), [morluto/rea#3881](https://github.com/morluto/rea/issues/3881), [#4341](https://github.com/morluto/rea/pull/4341), [morluto/rea#3963](https://github.com/morluto/rea/issues/3963), [bfb7ed0](https://github.com/morluto/rea/commit/bfb7ed0bd43edc605e00fc52c6e21e735b285904), [#3480](https://github.com/morluto/rea/pull/3480), [#3831](https://github.com/morluto/rea/pull/3831), [morluto/rea#3312](https://github.com/morluto/rea/issues/3312), [9ac76d3](https://github.com/morluto/rea/commit/9ac76d3d9486856d25395b413d7ac5ffa744cad0), [#4342](https://github.com/morluto/rea/pull/4342), [morluto/rea#4206](https://github.com/morluto/rea/issues/4206), [#4243](https://github.com/morluto/rea/pull/4243), [morluto/rea#3880](https://github.com/morluto/rea/issues/3880)). Thanks @coygeek, @kaoru0822-kitauji, @maxazure, @N0zoM1z0, @ShuaixinHuang.
+- **JavaScript relationships:** Distinguish `Map` reads from network endpoints, associate response endpoints only with writes that reach the read, and preserve child-process receiver ownership and effective stdio choices. Version comparisons keep relocation and unknown relationships distinct from application changes. ([fbfdb6a](https://github.com/morluto/rea/commit/fbfdb6abe9d663be0be146a15a489e33ccba0f68), [#4759](https://github.com/morluto/rea/pull/4759), [morluto/rea#4606](https://github.com/morluto/rea/issues/4606), [#3819](https://github.com/morluto/rea/pull/3819), [morluto/rea#3362](https://github.com/morluto/rea/issues/3362), [#4336](https://github.com/morluto/rea/pull/4336), [morluto/rea#3355](https://github.com/morluto/rea/issues/3355), [556504e](https://github.com/morluto/rea/commit/556504edcbb0f8cd9eaf81a91b18ff891b8f5457), [4d980d0](https://github.com/morluto/rea/commit/4d980d017622d3175d86aaf3b0dcfbdf4ca3d091)). Thanks @Abdooo2235, @coygeek, @kaoru0822-kitauji, @Xinglan233.
+- **JavaScript source admission:** Parse JavaScript with its own grammar, refuse FIFO recovery inputs without blocking, and report oversized or resource-skipped files with their actual capacity and coverage state. Malformed runtime events stay excluded rather than becoming policy-filtered facts. ([#3782](https://github.com/morluto/rea/pull/3782), [morluto/rea#3780](https://github.com/morluto/rea/issues/3780), [#3788](https://github.com/morluto/rea/pull/3788), [morluto/rea#3332](https://github.com/morluto/rea/issues/3332), [#4757](https://github.com/morluto/rea/pull/4757), [morluto/rea#4629](https://github.com/morluto/rea/issues/4629), [#3852](https://github.com/morluto/rea/pull/3852), [morluto/rea#3670](https://github.com/morluto/rea/issues/3670), [#4755](https://github.com/morluto/rea/pull/4755), [morluto/rea#4685](https://github.com/morluto/rea/issues/4685)). Thanks @andresaragon, @coygeek, @kaoru0822-kitauji, @sx4im.
+- **Browser capture:** Bound CDP handshakes and command responses, preserve malformed inventory uncertainty and partial cleanup diagnostics, and report cancellation after content reads and final cleanup. Accessibility comparisons require observed coverage; attached emulation teardown limits remain explicit. ([#3555](https://github.com/morluto/rea/pull/3555), [morluto/rea#3553](https://github.com/morluto/rea/issues/3553), [#3901](https://github.com/morluto/rea/pull/3901), [morluto/rea#3379](https://github.com/morluto/rea/issues/3379), [4cc044b](https://github.com/morluto/rea/commit/4cc044be68f813e96907e6cedcf422d0a3170de3), [#4406](https://github.com/morluto/rea/pull/4406), [morluto/rea#4403](https://github.com/morluto/rea/issues/4403), [#2893](https://github.com/morluto/rea/pull/2893), [0123d59](https://github.com/morluto/rea/commit/0123d5976828bfe621bd51daa4cf39fee48a1349), [2ff0867](https://github.com/morluto/rea/commit/2ff086796ad7b8f60397956750584e170dbe2910)). Thanks @coygeek, @dakjdakd, @kaoru0822-kitauji, @maxazure, @N0zoM1z0, @YunzhouLi-hub.
+- **Browser page facts:** Preserve URL path bytes when removing credentials, repeated document `Link` headers and complete COOP/COEP structured items. WebMCP discovery reads the actual schema properties and exact receiver path. ([#4731](https://github.com/morluto/rea/pull/4731), [df2fc25](https://github.com/morluto/rea/commit/df2fc251e318a1255309dba232f9f73c700777b6), [c360cfc](https://github.com/morluto/rea/commit/c360cfcd6d7901514b16dd6113eba2e0b0fdf422), [ebd5a82](https://github.com/morluto/rea/commit/ebd5a82c439f6deb3a06062225ce009eb03257cf), [38f92d1](https://github.com/morluto/rea/commit/38f92d108e3148cdf47e4ffbfe3de8a76e795d51)). Thanks @Abdooo2235, @geekwolf, @ShuaixinHuang.
+- **Electron runtime capture:** Keep native IPC listener and one-shot dispatch behavior, block synchronous shell effects during capture, refuse active scenarios when the inspect fuse is disabled, and preserve fuse/transport cleanup ownership. ([#4389](https://github.com/morluto/rea/pull/4389), [morluto/rea#4390](https://github.com/morluto/rea/issues/4390), [#4605](https://github.com/morluto/rea/pull/4605), [morluto/rea#4593](https://github.com/morluto/rea/issues/4593), [#4562](https://github.com/morluto/rea/pull/4562), [morluto/rea#4393](https://github.com/morluto/rea/issues/4393), [#3605](https://github.com/morluto/rea/pull/3605), [morluto/rea#3453](https://github.com/morluto/rea/issues/3453), [f9de413](https://github.com/morluto/rea/commit/f9de41371f7536e2e22ce6e43c9314fa2f238a30), [e7df00e](https://github.com/morluto/rea/commit/e7df00e80777757b329bbe8aba6d421d8892b9d5)). Thanks @Adolanium, @gensokyo-dev, @gkobayashi0, @HarsH-1877, @kaoru0822-kitauji, @keeponlight, @N0zoM1z0, @Nitin13-git, @xu91102.
+- **APK resource accuracy:** Retain completed tables and decode facts after interrupted or failed reads. Derive locale limits from resource configuration fields, disclose unselectable qualified tables, distinguish document failures from omitted entries, and bind Evidence to the decoded snapshot. ([34a88d6](https://github.com/morluto/rea/commit/34a88d6c1674710953e2c94bd0a17332cd69ca6f), [f5fcf52](https://github.com/morluto/rea/commit/f5fcf5216ffb3a266b8c00f6d9a0943051084ab5), [6385ef4](https://github.com/morluto/rea/commit/6385ef4355d7a794b3154ee38529cbddee583336), [98886ad](https://github.com/morluto/rea/commit/98886adddfcbd7b4f4236b3a9825f91878c3cb11), [dc2e811](https://github.com/morluto/rea/commit/dc2e8117b60c4481372ed89b077e6c1cf81ee653), [#4771](https://github.com/morluto/rea/pull/4771), [morluto/rea#4768](https://github.com/morluto/rea/issues/4768)). Thanks @abhay-2108, @gensokyo-dev, @Iyanuoluwa007, @kaoru0822-kitauji, @N0zoM1z0.
+- **Apktool guidance:** Retain the observed version and disclose the matching platform/release advisory windows without treating a version string as proof of decoder safety. Missing prerequisites and decode refusals identify the selected APK or launcher. ([d38f099](https://github.com/morluto/rea/commit/d38f099692394ec90c582f9f0a14b036b421b160), [63696d2](https://github.com/morluto/rea/commit/63696d277328f556c971d1fc9f81ae83128d468b), [cc038da](https://github.com/morluto/rea/commit/cc038da218cada68c87700477aba3db4739791ff)). Thanks @GovinduSasanka, @peter-research.
+- **Android method bodies:** Recover original bytecode bodies when Java source is omitted while preserving genuine native/abstract body absence and primary failures during retryable cleanup. ([63c65a2](https://github.com/morluto/rea/commit/63c65a22430a4bbc0c1807a0298edb867775001f), [742b6f5](https://github.com/morluto/rea/commit/742b6f587f3c4ef1b8212e7376d6f3a305b194dc), [9fa6d2b](https://github.com/morluto/rea/commit/9fa6d2b254cc390665974efb52463386a029e9fc)). Thanks @balkan16, @Chessing234.
+- **ADB observations and transfers:** Preserve display-scoped focus, distinguish missing declarations from absent focus, retain numeric physical density and mark unknown permission coverage partial. Transfers use consumed snapshot identity and refuse unknown destination existence; uninstall requests explicitly select user zero. ([7086768](https://github.com/morluto/rea/commit/7086768a27242a15daa93486aefbaacc3c891e78), [3d028db](https://github.com/morluto/rea/commit/3d028dbc74a465f6f8df1813a931fbb5362e7dc3), [#4074](https://github.com/morluto/rea/pull/4074), [morluto/rea#3942](https://github.com/morluto/rea/issues/3942), [#4663](https://github.com/morluto/rea/pull/4663), [#4665](https://github.com/morluto/rea/pull/4665), [1342127](https://github.com/morluto/rea/commit/1342127e91d4db8b4113a7a4e85be909bc96e1db), [61e6880](https://github.com/morluto/rea/commit/61e688070f48016b6d178cd04c02fcc2f33e19ad), [be0b2d1](https://github.com/morluto/rea/commit/be0b2d138c46bb951fc3c8ee57ef7b0d57fed312)). Thanks @Copilot, @coygeek, @gongfpp, @Tapraj-Pathak, @yankikucuk.
+- **Apple metadata:** Preserve exact plist/keyed-archive integers, negative zero, non-finite real identity and dictionary keys. Reject malformed XML before value decoding, retain undecodable dates as unknown and report unsupported OpenStep entries explicitly. ([#4324](https://github.com/morluto/rea/pull/4324), [morluto/rea#4292](https://github.com/morluto/rea/issues/4292), [2cf9c28](https://github.com/morluto/rea/commit/2cf9c28f27261b7e64af2ba7ecb4e23ea7714aee), [#4839](https://github.com/morluto/rea/pull/4839), [morluto/rea#4834](https://github.com/morluto/rea/issues/4834), [f2b7bbb](https://github.com/morluto/rea/commit/f2b7bbb31279a820f4cf32d439570f8e5381d010), [6c98708](https://github.com/morluto/rea/commit/6c98708793c23615de8053b2cd14d9f164b960c5), [78cf610](https://github.com/morluto/rea/commit/78cf6100f33ad2412c0a3d60b83f8a44a43fbbb5), [c7e9e09](https://github.com/morluto/rea/commit/c7e9e096d3e2444e4c251123b8e914e9c0941f4e), [#3793](https://github.com/morluto/rea/pull/3793), [#4105](https://github.com/morluto/rea/pull/4105), [morluto/rea#3892](https://github.com/morluto/rea/issues/3892)). Thanks @coygeek, @kaoru0822-kitauji, @mertkayacs, @pablodiazjorge, @pollychen-lab, @xu91102.
+- **Apple application structure:** Derive bundle components and launchd domains from their owning directories, retain unavailable components, and select only exact `Assets.car` names as asset catalogs. Interface Builder keeps unresolved classes and counts hierarchy records beyond the returned object page. ([149a0fc](https://github.com/morluto/rea/commit/149a0fc06c3c67ee45624e6da14dfb0754473758), [5d6029e](https://github.com/morluto/rea/commit/5d6029ea62a65ec34fd739ee40223c77c6fdb63b), [c192c58](https://github.com/morluto/rea/commit/c192c58815ddeaf5cf30159e7082c742c0bc36b7), [#4094](https://github.com/morluto/rea/pull/4094), [morluto/rea#3914](https://github.com/morluto/rea/issues/3914), [#4118](https://github.com/morluto/rea/pull/4118), [morluto/rea#3958](https://github.com/morluto/rea/issues/3958), [#4162](https://github.com/morluto/rea/pull/4162), [morluto/rea#3950](https://github.com/morluto/rea/issues/3950)). Thanks @coygeek, @kaoru0822-kitauji.
+- **Native names and layout:** Keep valid Objective-C records beside malformed neighbors, preserve leading U+FEFF and repeated load-command fields, and select exact arm64e.x1 slices. PE resource ownership and NativeAOT reads stay within declared image/header ranges. ([#3833](https://github.com/morluto/rea/pull/3833), [morluto/rea#3336](https://github.com/morluto/rea/issues/3336), [#4549](https://github.com/morluto/rea/pull/4549), [morluto/rea#4540](https://github.com/morluto/rea/issues/4540), [morluto/rea#4541](https://github.com/morluto/rea/issues/4541), [#4730](https://github.com/morluto/rea/pull/4730), [#4048](https://github.com/morluto/rea/pull/4048), [morluto/rea#4008](https://github.com/morluto/rea/issues/4008), [#3500](https://github.com/morluto/rea/pull/3500), [morluto/rea#3389](https://github.com/morluto/rea/issues/3389), [#4463](https://github.com/morluto/rea/pull/4463), [morluto/rea#4373](https://github.com/morluto/rea/issues/4373), [morluto/rea#4455](https://github.com/morluto/rea/issues/4455), [3fc0075](https://github.com/morluto/rea/commit/3fc007534bb2104d662b43cc712281f9d1d6a7d3), [4a076f3](https://github.com/morluto/rea/commit/4a076f368c62bdfb19415accad342e03a38d1005)). Thanks @coygeek, @Iyanuoluwa007, @jayinvers, @kaoru0822-kitauji, @kevinjosh10, @maxazure, @ShuaixinHuang.
+- **Native analysis providers:** Keep Ghidra byte matches within examined coverage, retain PE PDB/function-boundary limitations and preserve configured MIPS choices. Attached IDA refuses universal Mach-O rather than silently choosing a slice; request timeout identity and captured output survive adapter handling. ([#4531](https://github.com/morluto/rea/pull/4531), [#2933](https://github.com/morluto/rea/pull/2933), [morluto/rea#2259](https://github.com/morluto/rea/issues/2259), [01a1b3f](https://github.com/morluto/rea/commit/01a1b3fda6a7dacd24a0b10619d41d12153080d4), [50dd9fc](https://github.com/morluto/rea/commit/50dd9fcad1efb47cb404192c30bdf190dcdaca58), [#4765](https://github.com/morluto/rea/pull/4765), [morluto/rea#3871](https://github.com/morluto/rea/issues/3871), [c0cfccb](https://github.com/morluto/rea/commit/c0cfccb45fe43e7eeff18e56c64caa5b121988f9)). Thanks @BeLit-Development-2, @coygeek, @kaoru0822-kitauji, @N0zoM1z0, @ShuaixinHuang.
+- **Recorded crashes:** Prevent GDB from discovering an executable named by an untrusted core, reject core mapping paths that escape the isolated sysroot, and preserve pwntools worker ownership across launcher exec. Completed decode facts and cancellation reasons remain available through failed final cleanup. ([5bf4fff](https://github.com/morluto/rea/commit/5bf4fffc1b835130ddb16f96db9ce13b639591db), [b33c0df](https://github.com/morluto/rea/commit/b33c0df02e93fabca00a44f2de625b1a63637ee9), [#4435](https://github.com/morluto/rea/pull/4435), [morluto/rea#4011](https://github.com/morluto/rea/issues/4011)). Thanks @coygeek, @N0zoM1z0.
+- **Cutter and Rizin:** Read Cutter’s active document from core configuration and attribute its host version correctly. Keep interrupted Rizin frames and output whitespace before completion markers, with exact command and cleanup failure attribution. ([#4507](https://github.com/morluto/rea/pull/4507), [morluto/rea#4497](https://github.com/morluto/rea/issues/4497), [#4514](https://github.com/morluto/rea/pull/4514), [morluto/rea#4498](https://github.com/morluto/rea/issues/4498), [#4459](https://github.com/morluto/rea/pull/4459), [morluto/rea#4015](https://github.com/morluto/rea/issues/4015), [d5fc7e6](https://github.com/morluto/rea/commit/d5fc7e6ebf9af050081fc4c6e18b9079cfaa5cdc), [199396f](https://github.com/morluto/rea/commit/199396f61c8f61da326d059606a992304a760147)). Thanks @coygeek, @N0zoM1z0, @Xander-AJ.
+- **Windows cleanup:** Retain unpublished runtimes after failed removal, handle read-only attributes and sharing conflicts, preserve primary snapshot failures and honor caller environment overrides. ([f78d137](https://github.com/morluto/rea/commit/f78d137b42741a0ed5fb8ac161286f9a50220c4d), [d69acc3](https://github.com/morluto/rea/commit/d69acc3455c33f90a2fc4381db43d38b6af50622), [0679a79](https://github.com/morluto/rea/commit/0679a7972b4f2b02eee6a86a44f50f4c2967d3cb), [5f9ac2c](https://github.com/morluto/rea/commit/5f9ac2c98a6fe11988c41afb737df9b6d5c3a1b3), [1bbf4bd](https://github.com/morluto/rea/commit/1bbf4bd16ca134cdb2187eaf29291cae90d14a57)). Thanks @DeryFerd.
+- **Managed metadata:** Compare nominal signatures and declaring identities across builds, preserve empty names without display substitutions, verify ABI-prefixed Mach-O exports exactly, and charge cached metadata once. Metadata errors report artifact file offsets; mixed-mode assemblies retain their actual boundary classification. ([#3059](https://github.com/morluto/rea/pull/3059), [#3079](https://github.com/morluto/rea/pull/3079), [#3853](https://github.com/morluto/rea/pull/3853), [morluto/rea#3352](https://github.com/morluto/rea/issues/3352), [#3974](https://github.com/morluto/rea/pull/3974), [morluto/rea#3961](https://github.com/morluto/rea/issues/3961), [#3491](https://github.com/morluto/rea/pull/3491), [morluto/rea#3449](https://github.com/morluto/rea/issues/3449), [#4418](https://github.com/morluto/rea/pull/4418), [morluto/rea#4372](https://github.com/morluto/rea/issues/4372), [#3516](https://github.com/morluto/rea/pull/3516), [morluto/rea#3366](https://github.com/morluto/rea/issues/3366)). Thanks @8rulerstar, @coygeek, @duckhunt113, @Hahaa909, @JayOfTheKeyboard, @kaoru0822-kitauji, @keeponlight, @N0zoM1z0, @tomyrms.
+- **Archive and filesystem identity:** Reject colliding extraction paths before writing, classify ZIP offsets beyond EOF as malformed, normalize ZIP current-directory prefixes once, and preserve symlink and unreadable-path provenance. A directory named `*.map` is no longer mistaken for a source-map file. ([862a0b2](https://github.com/morluto/rea/commit/862a0b2631785492f8dc20c8acd396646c545fd1), [c209dcb](https://github.com/morluto/rea/commit/c209dcb209118069f123e789949eeba1ef78853e), [6b1c36e](https://github.com/morluto/rea/commit/6b1c36e336eeacbb7159b211e687b4b84f99b9f5), [25c2222](https://github.com/morluto/rea/commit/25c22222abe7a681491bdb4d7316250fdcf32973), [d6e4fe4](https://github.com/morluto/rea/commit/d6e4fe4a4dd8914ec3db50d2a7fc3d0802649262), [#4131](https://github.com/morluto/rea/pull/4131), [morluto/rea#3909](https://github.com/morluto/rea/issues/3909)). Thanks @Adkid-Zephyr, @coygeek, @jackie-cqz, @kaoru0822-kitauji, @maxazure, @XY-code1.
+- **Evidence and reconstruction:** Authenticate all retained inventory substeps, preserve source-attributed caveats in composed workflows, and require comparable cited Evidence before declaring reconstruction coverage or readiness. Missing Evidence and residual replay uncertainty stay explicit. ([9ac91e0](https://github.com/morluto/rea/commit/9ac91e0c1e13e91cd21341635ad5be89b0e8642d), [c4608a2](https://github.com/morluto/rea/commit/c4608a29620ea3e7bb8259b47357b942fe4e4709), [#4075](https://github.com/morluto/rea/pull/4075), [morluto/rea#3357](https://github.com/morluto/rea/issues/3357), [#3526](https://github.com/morluto/rea/pull/3526), [morluto/rea#3363](https://github.com/morluto/rea/issues/3363), [153017d](https://github.com/morluto/rea/commit/153017d750322188c1a112b2284c1b8065c67320)). Thanks @coygeek, @gensokyo-dev, @kaoru0822-kitauji, @N0zoM1z0, @rudycelekli.
+- **Offline data:** Preserve exact UTC packet-capture instants, SQLite column identity and snapshot admission failures, and cooked WASM directive literals. Cleanup cancellation retains completed EVM/WASM observations rather than losing them. ([796faa0](https://github.com/morluto/rea/commit/796faa0e4e6bf745bcdc33acef65af8ccfc3622d), [748ac7f](https://github.com/morluto/rea/commit/748ac7f0df340f0d0569eca2e803a59cd51ffdbd), [bc51d0e](https://github.com/morluto/rea/commit/bc51d0ea5df7dcee1fcc42dfae3ab0c62c9cca5e), [4f87f02](https://github.com/morluto/rea/commit/4f87f02bb19bf77349a683a5bcb2da2d1da427d0), [d694aef](https://github.com/morluto/rea/commit/d694aef227e90164e6654ea9e5ce6a854d099404), [321cacb](https://github.com/morluto/rea/commit/321cacb84a020f447e1110df5c58fa71d99b7bf1)). Thanks @himamaheswarreddy, @mertkayacs, @rollroyces, @viniciusdsandrade.
+- **Kuna recovery:** Share concurrent single-function requests and cache repeats, validate warm exports once, stream string inventories beyond the former fixed buffer ceiling, and keep unproven processes in their run slots until close confirms release. Failure messages distinguish final buffer refusals from configurable deadlines. ([#4045](https://github.com/morluto/rea/pull/4045), [#4169](https://github.com/morluto/rea/pull/4169), [#4049](https://github.com/morluto/rea/pull/4049), [#4318](https://github.com/morluto/rea/pull/4318), [#4052](https://github.com/morluto/rea/pull/4052), [#4098](https://github.com/morluto/rea/pull/4098), [#4343](https://github.com/morluto/rea/pull/4343)). Thanks @kaoru0822-kitauji.
+- **Flutter snapshots:** Check symbol-table ranges against loadable segments and retain completed inspection facts when cancellation occurs during snapshot cleanup. ([#3859](https://github.com/morluto/rea/pull/3859), [morluto/rea#3346](https://github.com/morluto/rea/issues/3346), [#4766](https://github.com/morluto/rea/pull/4766), [morluto/rea#4717](https://github.com/morluto/rea/issues/4717), [#4762](https://github.com/morluto/rea/pull/4762), [morluto/rea#4640](https://github.com/morluto/rea/issues/4640)). Thanks @coygeek, @kaoru0822-kitauji.
+- **Native metadata recovery:** Keep unknown Objective-C category classes, empty protocol names, undecoded Swift aliases and unfinished property types explicit instead of discarding neighboring valid records. Preserve unnamed segment identity and exclude a dylib’s own install name from dependencies. ([#4093](https://github.com/morluto/rea/pull/4093), [morluto/rea#4012](https://github.com/morluto/rea/issues/4012), [#4078](https://github.com/morluto/rea/pull/4078), [morluto/rea#3873](https://github.com/morluto/rea/issues/3873), [#4085](https://github.com/morluto/rea/pull/4085), [morluto/rea#3934](https://github.com/morluto/rea/issues/3934), [#3825](https://github.com/morluto/rea/pull/3825), [morluto/rea#3345](https://github.com/morluto/rea/issues/3345), [#4051](https://github.com/morluto/rea/pull/4051), [morluto/rea#3875](https://github.com/morluto/rea/issues/3875), [#4050](https://github.com/morluto/rea/pull/4050), [morluto/rea#3872](https://github.com/morluto/rea/issues/3872)). Thanks @coygeek, @kaoru0822-kitauji.
+- **Document selection:** Preserve explicitly selected plist paths and bundle defaults, and only use document fallback after confirming its applicability. Finalization keeps completed selection facts and cancellation outcomes. ([e6c4877](https://github.com/morluto/rea/commit/e6c487799bb64f559e66b69be1863e6ec5187548), [#3774](https://github.com/morluto/rea/pull/3774), [9df559b](https://github.com/morluto/rea/commit/9df559b7e712298a761ff26e9caae8d659b7e54b)). Thanks @kaoru0822-kitauji.
+- **Capture and comparison accuracy:** Admit schema-valid HAR numeric values while retaining exact-publication limits, preserve empty URL path segments, and keep unrelated process traces and unnamed reference endpoints distinct or unknown. ([#3159](https://github.com/morluto/rea/pull/3159), [morluto/rea#3145](https://github.com/morluto/rea/issues/3145), [#3792](https://github.com/morluto/rea/pull/3792), [morluto/rea#3307](https://github.com/morluto/rea/issues/3307), [#4178](https://github.com/morluto/rea/pull/4178), [morluto/rea#3931](https://github.com/morluto/rea/issues/3931), [#3809](https://github.com/morluto/rea/pull/3809), [morluto/rea#3388](https://github.com/morluto/rea/issues/3388)). Thanks @coygeek, @gensokyo-dev, @kaoru0822-kitauji, @N0zoM1z0, @pablodiazjorge, @xu91102.
+- **Evidence publication:** Recheck staged output identity after its final write, classify missing review Evidence and duplicate ledger entries as caller input, and preserve the selected freshness requirement in comparison parameters. ([#4764](https://github.com/morluto/rea/pull/4764), [morluto/rea#4613](https://github.com/morluto/rea/issues/4613), [#4452](https://github.com/morluto/rea/pull/4452), [morluto/rea#4444](https://github.com/morluto/rea/issues/4444), [#4065](https://github.com/morluto/rea/pull/4065), [morluto/rea#4000](https://github.com/morluto/rea/issues/4000)). Thanks @amyanger, @coygeek, @kaoru0822-kitauji, @Rinkia, @Swatantra-66.
+
+### ⚠ BREAKING CHANGES
+
+- **Saved Evidence:** Evidence and analysis snapshots now require explicit nullable `relationship_coverage` in their identity commitments. Artifact inventories require scoped coverage and explicit `integrity_contradictions`; obsolete `coverage_status` and prose-based coverage inference are removed. Regenerate old Evidence/snapshots and inventory observations with this release: saved files are not migrated and omitted facts are not filled with guessed defaults. ([db8128d](https://github.com/morluto/rea/commit/db8128dc0d1c4f731f3d19877fa00d25cf17f657), [0405c9d](https://github.com/morluto/rea/commit/0405c9d361de6d904dfa16cc0b6d8350407f2c9b), [a9588be](https://github.com/morluto/rea/commit/a9588be4ebdd20eeda975cb63ffcdb1cd868b340)).
+- **Browser capture results:** Consumers must retain resolved capture limits, event families and omission facts. Set desired budgets through the supported capture options; omitted caller inputs still select defaults, but stored observations must report the resolved values. Admitted content reads queue behind bounded concurrency. ([57a23f0](https://github.com/morluto/rea/commit/57a23f0a2da0fdc3835e203718ded8acea5a34e6)).
+- **Android version codes:** `inspect_adb_package.version_code` is a decimal string rather than a JSON number. Read it as a string, or use `BigInt` for numeric comparison. ([#3485](https://github.com/morluto/rea/pull/3485), [morluto/rea#3368](https://github.com/morluto/rea/issues/3368), [7666a48](https://github.com/morluto/rea/commit/7666a482b69dc7cabdbac3bbaf89a43e32300192)). Thanks @coygeek, @kaoru0822-kitauji.
+- **Native UI failures:** Failed or cancelled native UI scenarios now return typed errors with the collected report in `partial_observation`. Handle the error path and read that report instead of expecting a successful result containing failed steps. ([d909919](https://github.com/morluto/rea/commit/d9099190ab3a819bd3a9fd62a8714cef398016d5)).
+- **Provider status:** `binary_session` candidate availability can be unknown during owner finalization. Preserve that state rather than treating it as unavailable; internal consumers must await the asynchronous session status API. ([150e996](https://github.com/morluto/rea/commit/150e996a86f2bec4982ba87f37bf502a8d5e8214)).
+- **Swift symbol locations:** `swift_symbols` entries may have a null `address` when the provider supplies no location. Handle unknown addresses; optional `evidence.reported_address` retains the original provider value. ([8116dc0](https://github.com/morluto/rea/commit/8116dc06512b4609d8db63cebc9f2ff8cb964151)).
+- **Linux Hopper admission:** Older builds previously admitted by a fixed allowlist need retained publisher-signed package evidence or an approved update to the current public demo. Update explicit `HOPPER_LAUNCHER_PATH` selections when migrating; the live feed offers only its current release. The official RPM metadata mismatch remains unresolved and is refused. ([#3590](https://github.com/morluto/rea/pull/3590), [morluto/rea#3577](https://github.com/morluto/rea/issues/3577), [70cd4cb](https://github.com/morluto/rea/commit/70cd4cbfb849b61f05d97d2690372a9782831426)). Thanks @gensokyo-dev, @kaoru0822-kitauji, @N0zoM1z0.
+
+Full contribution record: [571 commits and 236 merged PRs](https://github.com/morluto/rea/blob/rea-agents-6.6.0/docs/releases/6.6.0.contributions.json).
+
 ## [6.5.0](https://github.com/morluto/rea/compare/rea-agents-6.4.0...rea-agents-6.5.0) (2026-10-13)
 
 Provider prerequisites and recorded verification scope are listed in [the testing guide](docs/testing.md).

CONTRIBUTING.md

@@ -21,8 +21,13 @@ For analysis workflows and agent interfaces, follow the
 [performance priorities](AGENTS.md#performance). Explain the selected output
 scope, reuse/cache decisions, and relevant before/after measurements when these
 costs change.
+Use the [analysis optimization skill](.agents/skills/rea-analysis-optimization/SKILL.md)
+to locate provider and shared-workflow bottlenecks, preserve Evidence, and
+compare equivalent real tasks before claiming an improvement.
 
 When adding or changing an MCP tool, follow the [tool design guide](docs/tool-design.md) and preserve the canonical contracts and generated catalog.
+When adding or changing an admission ceiling, budget, or deadline, follow
+[admission limits and recovery](docs/design/limits-and-recovery.md).
 
 ## Reporting issues
 
@@ -225,7 +230,10 @@ manifest. A source checkout does not: `native/windows/build/` is ignored by
 Git, and neither `npm ci` nor `npm run build:cached` (or `npm run build`)
 produces `rea-windows-x64.node` or `manifest.json` in that directory.
 Windows workflows and tests that require native controls need these artifacts;
-without them, those controls are unavailable even after a successful runtime build.
+this includes `run-exclusive` commands, which use the native Job Object to keep
+the artifact lock until every command descendant has exited. Without the
+artifacts, those controls are unavailable even after a successful runtime
+build.
 
 Follow the [Windows native build guide](native/windows/README.md#build) before
 running those workflows from source. The artifact build requires Linux,
@@ -233,6 +241,19 @@ MinGW-w64, and `tar`; the build script does not run on Windows. See that guide
 for the build and artifact-verification commands and the separate real Windows
 verification prerequisites. Keep the generated addon and manifest uncommitted.
 
+The general source suite (`npm run test:fast`, `npm run test:changed`, and
+`npm run test:local`) runs on Linux in CI, and Windows CI runs a curated lane.
+On a Windows checkout, some source tests fail because fixtures use POSIX paths,
+`#!/bin/sh` tool stubs, or POSIX rename semantics. Compare failures against the
+base revision, or run the suite on Linux or WSL, before attributing them to a
+change. See [testing](docs/testing.md#developer-commands) for command scopes
+and lane selection.
+
+`git submodule update --init --recursive` fails at the pinned Unblob submodule
+on Windows and can leave later submodules unchecked out. Apply the
+[Unblob checkout workaround](docs/firmware-analysis.md#windows-submodule-checkout),
+then rerun the command.
+
 ## Documentation website
 
 The VitePress site uses the Markdown files in `docs/`. Run `npm run docs:dev`
@@ -255,6 +276,23 @@ The separate GitHub Pages host uses only the manual
 **Settings → Pages → Build and deployment → Source** must be set to
 **GitHub Actions** before deploying there.
 
+### Translation corrections
+
+README editions and website editions have separate translation sources:
+
+- Find the language's `readme` path in the
+  [language inventory](website/locales/languages.json) and edit that file.
+  Locale codes and README suffixes can differ. Use the
+  [English README](README.md) to check commands, requirements, and analysis claims.
+- Edit `msgstr` values in `website/locales/<language>/messages.po` for website
+  text. Follow the [website translation guide](website/locales/README.md) for
+  the Python environment, catalog validation, and local preview. Preserve source
+  strings, contexts, placeholders, commands, paths, and recorded evidence.
+
+Before submitting a website translation, run the catalog and website checks in
+that guide. Translated HTML in `website/_site/` is generated; edit the PO catalog
+rather than the generated pages.
+
 ## Development feedback and PR verification
 
 For routine merges, wait for the required checks to pass for the current PR

README.md

@@ -1,7 +1,7 @@
 <div align="center">
 
 **English** · [简体中文](docs/readme/README_zh.md) · [繁體中文](docs/readme/README_zh-TW.md) · [日本語](docs/readme/README_ja.md) · [한국어](docs/readme/README_ko.md) · [Türkçe](docs/readme/README_tr.md) · [Русский](docs/readme/README_ru.md) · [Tiếng Việt](docs/readme/README_vi.md) · [ไทย](docs/readme/README_th.md) · [Deutsch](docs/readme/README_de.md) · [Español](docs/readme/README_es.md) · [Français](docs/readme/README_fr.md) · [Українська](docs/readme/README_uk.md) · [Polski](docs/readme/README_pl.md) · [Português (Brasil)](docs/readme/README_pt-BR.md) · [العربية](docs/readme/README_ar.md) · [فارسی](docs/readme/README_fa.md) · [Bahasa Indonesia](docs/readme/README_id.md) · [Italiano](docs/readme/README_it.md)
-· [Ελληνικά](docs/readme/README_el.md)
+· [Ελληνικά](docs/readme/README_el.md) · [മലയാളം](docs/readme/README_ml.md) · [Roman Urdu](docs/readme/README_ur-Latn.md)
 
 # REA: Reverse Engineer Anything
 
@@ -90,6 +90,12 @@ Setup supports Claude Code, Codex, Cursor, Gemini CLI, Grok Build and
 [installation and setup](docs/installation.md) for provider configuration and
 manual MCP registration.
 
+MCP uses compact input schemas by default to reduce context overhead. Set
+`REA_MCP_INPUT_SCHEMA_PROFILE=full` in your MCP registration when you need the
+complete advertised field guidance and nested validation details. Both profiles
+keep the full tool set and canonical server validation. See
+"Compact input schema profile" in the MCP contracts guide.
+
 ### Ask your agent
 
 ```text
@@ -132,6 +138,14 @@ For a CLI installed with `npm install --global` or `bun add --global`:
 rea update
 ```
 
+`rea --update` is an alias for this verified workflow and accepts the same global
+output flags. Use `rea update` when combining it with other command controls.
+
+Updater commands retain at most 4 MiB of stdout and stderr combined. Oversized
+metadata replies are rejected after the command exits; verbose installs continue
+with bounded diagnostics. Human progress output respects terminal backpressure
+and waits for submitted writes to finish.
+
 Bun global updates require Bun 1.2.17 or later. With an older Bun release,
 upgrade Bun manually before running `rea update`.
 
@@ -163,34 +177,44 @@ CLI commands use the same workflows.
 
 ## What you can analyze
 
-REA requires Node.js 22.x (>=22.19), 24.x (>=24.11), or 26+, plus npm.
+REA requires Node.js 22.x (>=22.23.4), 24.x (>=24.22.1), or 26.12.1+, plus npm.
 Additional tools and host support depend on the target:
 
-| Target                 | What REA returns                                                                     | Requirements and guide                                                                                                                |
-| ---------------------- | ------------------------------------------------------------------------------------ | ------------------------------------------------------------------------------------------------------------------------------------- |
-| Native binaries        | Pseudocode, assembly, strings, symbols, calls and references                         | Hopper, Ghidra or IDA; [native analysis](docs/native-investigation.md)                                                                |
-| Offline ELF layout     | Sections, segments, original symbols/relocations and static mitigation candidates    | Caller-supplied pwntools on Linux x64; [binary diagnostics](docs/binary-diagnostics.md)                                               |
-| Go binaries            | Embedded compiler version, modules, replacements and build settings                  | Built-in static ELF/PE/thin Mach-O inspection; no Go installation required                                                            |
-| Rust binaries          | Embedded cargo-auditable crate names, versions, sources and dependency edges         | Built-in static ELF/PE/thin Mach-O inspection; [Rust binaries](docs/rust-binaries.md)                                                 |
-| WASM artifact          | Validation, sections, exact import/export WAT forms and decoded WAT                  | Caller-supplied WABT 1.0.42; `inspect_wasm_artifact`                                                                                  |
-| EVM bytecode           | Dispatch selectors, byte offsets, inferred arguments and mutability                  | Local raw/hex carrier on Linux x64; [offline EVM guide](docs/evm-bytecode.md)                                                         |
-| SQLite databases       | Tables, columns, indexes, views, triggers and selected ordinary-table rows           | Local database/WAL snapshot; Node.js 24.x >=24.15 or 26+ with native SQLite limits                                                    |
-| Recorded Linux crashes | Raw notes, every recorded thread's registers/signals and optional mapping candidates | Caller-supplied pwntools on Linux x64; optional GDB/pwndbg; [recorded crashes](docs/recorded-crashes.md)                              |
-| JavaScript / Electron  | Modules, imports, source maps, routes, IPC and native add-on relationships           | Node.js and npm; Linux: procps-compatible `ps`; [application analysis](docs/javascript-application-workflows.md)                      |
-| Websites               | Page structure, scripts, network observations and requested screenshots              | A Chrome-family browser; [browser analysis](docs/browser-observation.md)                                                              |
-| Saved network captures | Requests, responses, exposed payloads and source locations                           | HAR or Fiddler Everywhere SAZ; mitmdump on Linux for native mitmproxy; [capture guide](docs/web-network-captures.md)                  |
-| .NET assemblies        | Metadata, CIL instructions, declared native dependencies and build comparisons       | None (static only); [managed-code guide](docs/managed-code-analysis.md)                                                               |
-| Android APKs           | Manifest declarations, classes, decompiled methods and references                    | Headless JADX and a full JDK on Linux/macOS/Windows x64; [Android guide](docs/android-analysis.md)                                    |
-| Android devices        | Devices, packages, processes, logs, dumps, screen captures, and APK/file transfers   | Caller-supplied adb (emulator or hardware); [ADB guide](docs/adb-device-analysis.md)                                                  |
-| JEB projects           | Engine-backed project units, artifact digests and type/method pseudo-code            | A caller-started JEB client serving MCP, selected with REA_JEB_MCP_URL; [JEB guide](docs/jeb-analysis.md)                             |
-| Android resources      | Decoded manifest, string tables, locales and version facts                           | Caller-supplied Apktool launcher; [resource guide](docs/apktool-resource-analysis.md)                                                 |
-| Flutter apps           | Dart snapshot hash, per-ABI library digests and engine build provenance              | None (static APK parsing); [Flutter guide](docs/flutter-analysis.md)                                                                  |
-| Firmware               | Regions, extraction results and native-analysis handoffs                             | Binwalk / Unblob on Linux; [firmware guide](docs/firmware-analysis.md)                                                                |
-| Recorded packets       | Selected fields, repeated values, filters and decode-as rules                        | Linux: TShark 4.x + prlimit; [capture guide](docs/packet-captures.md)                                                                 |
-| Packet summaries       | Endpoints, DNS, TLS SNI, protocol distribution and HTTP requests                     | Caller-supplied tshark and capinfos (Linux-verified); [pcap guide](docs/pcap-analysis.md)                                             |
-| Recorded logic signals | Decoder annotations, sample ranges and diagnostics                                   | Linux: sigrok-cli 0.7.2 + prlimit; [signal recordings](docs/signal-recordings.md)                                                     |
-| Packages and resources | File inventories, digests, plists, Apple bundle anatomy and extracted resources      | [Artifact and JavaScript guide](docs/javascript-artifact-reconstruction.md), [Apple applications](docs/apple-application-analysis.md) |
-| Process behavior       | Terminal output, interactions, exit and filesystem observations, and run comparisons | Linux/macOS with a native PTY; [process capture](docs/process-capture.md)                                                             |
+| Target                  | What REA returns                                                                                  | Requirements and guide                                                                                                                |
+| ----------------------- | ------------------------------------------------------------------------------------------------- | ------------------------------------------------------------------------------------------------------------------------------------- |
+| Desktop installations   | Runtime candidates, supporting files, component markers and next probes                           | Built-in bounded directory/header inspection; [desktop applications](docs/desktop-applications.md)                                    |
+| Native binaries         | Pseudocode, assembly, strings, symbols, calls and references                                      | Hopper, Ghidra or IDA; [native analysis](docs/native-investigation.md)                                                                |
+| Offline ELF layout      | Sections, segments, original symbols/relocations and static mitigation candidates                 | Caller-supplied pwntools on Linux x64; [binary diagnostics](docs/binary-diagnostics.md)                                               |
+| Go binaries             | Embedded compiler version, modules, replacements and build settings                               | Built-in static ELF/PE/thin Mach-O inspection; no Go installation required                                                            |
+| GPU code objects        | CUDA/HIP targets, selected kernel parameters and resources, retained PTX text                     | Built-in static ELF/PE/fatbin/CUBIN/PTX inspection; [GPU code objects](docs/gpu-code-objects.md)                                      |
+| Model artifacts         | GGUF / safetensors metadata, tensor declarations and derived totals                               | Built-in header inspection; optional full-file hash; [model files](docs/model-artifacts.md)                                           |
+| IL2CPP metadata         | Unity global-metadata.dat assembly images, type and method names and index ranges                 | Built-in static decoding of versions 24, 27, 29, 31 and 39; no Unity required                                                         |
+| Godot PCK               | Archive inventory and extraction, original header/directory facts and selected content identities | Built-in static formats 2, 3 and 4; no Godot required                                                                                 |
+| Rust binaries           | Embedded cargo-auditable crate names, versions, sources and dependency edges                      | Built-in static ELF/PE/thin Mach-O inspection; [Rust binaries](docs/rust-binaries.md)                                                 |
+| PE certificates         | Authenticode WIN_CERTIFICATE entries and PKCS#7 signer/X.509 identity fields                      | Built-in static PE inspection; [PE certificates](docs/pe-certificates.md)                                                             |
+| ELF notes               | PT_NOTE/SHT_NOTE owners, types, GNU build-id and ABI-tag fields                                   | Built-in static ELF inspection; [ELF notes](docs/elf-notes.md)                                                                        |
+| Erlang BEAM             | IFF chunks, atoms, exports, imports, lambdas and the Code header                                  | Built-in static BEAM inspection; [Erlang BEAM](docs/erlang-beam.md)                                                                   |
+| WASM artifact           | Validation, sections, exact import/export WAT forms and decoded WAT                               | Caller-supplied WABT 1.0.42; `inspect_wasm_artifact`                                                                                  |
+| EVM bytecode            | Dispatch selectors, byte offsets, inferred arguments and mutability                               | Local raw/hex carrier on Linux x64; [offline EVM guide](docs/evm-bytecode.md)                                                         |
+| SQLite databases        | Tables, columns, indexes, views, triggers and selected ordinary-table rows                        | Local database/WAL snapshot; Node.js 24.x >=24.15 or 26+ with native SQLite limits                                                    |
+| Recorded Linux crashes  | Raw notes, every recorded thread's registers/signals and optional mapping candidates              | Caller-supplied pwntools on Linux x64; optional GDB/pwndbg; [recorded crashes](docs/recorded-crashes.md)                              |
+| JavaScript / Electron   | Modules, imports, source maps, routes, IPC and native add-on relationships                        | Node.js and npm; Linux: procps-compatible `ps`; [application analysis](docs/javascript-application-workflows.md)                      |
+| Websites                | Page structure, scripts, network observations and requested screenshots                           | A Chrome-family browser; [browser analysis](docs/browser-observation.md)                                                              |
+| Saved network captures  | Requests, responses, exposed payloads and source locations                                        | HAR or Fiddler Everywhere SAZ; mitmdump on Linux for native mitmproxy; [capture guide](docs/web-network-captures.md)                  |
+| .NET assemblies         | Metadata, CIL instructions, declared native dependencies and build comparisons                    | None (static only); [managed-code guide](docs/managed-code-analysis.md)                                                               |
+| Android APKs            | Manifest declarations, classes, decompiled methods and references                                 | Headless JADX and a full JDK on Linux/macOS/Windows x64; [Android guide](docs/android-analysis.md)                                    |
+| APK signatures          | v1/v2/v3/v3.1 signers, certificate fingerprints, key rotation and digest self-consistency         | None (static only); `inspect_apk_signature`                                                                                           |
+| Android devices         | Devices, packages, processes, logs, dumps, screen captures, and APK/file transfers                | Caller-supplied adb (emulator or hardware); [ADB guide](docs/adb-device-analysis.md)                                                  |
+| JEB projects            | Engine-backed project units, artifact digests and type/method pseudo-code                         | A caller-started JEB client serving MCP, selected with REA_JEB_MCP_URL; [JEB guide](docs/jeb-analysis.md)                             |
+| Android resources       | Decoded manifest, string tables, locales and version facts                                        | Caller-supplied Apktool launcher; [resource guide](docs/apktool-resource-analysis.md)                                                 |
+| Flutter apps            | Dart snapshot hash, per-ABI library digests and engine build provenance                           | None (static APK parsing); [Flutter guide](docs/flutter-analysis.md)                                                                  |
+| Firmware                | Regions, extraction results and native-analysis handoffs                                          | Binwalk / Unblob on Linux; [firmware guide](docs/firmware-analysis.md)                                                                |
+| Recorded packets        | Selected fields, repeated values, filters and decode-as rules                                     | Linux: TShark 4.x + prlimit; [capture guide](docs/packet-captures.md)                                                                 |
+| Packet summaries        | Endpoints, DNS, TLS SNI, protocol distribution and HTTP requests                                  | Caller-supplied tshark and capinfos (Linux-verified); [pcap guide](docs/pcap-analysis.md)                                             |
+| Recorded logic signals  | Decoder annotations, sample ranges and diagnostics                                                | Linux: sigrok-cli 0.7.2 + prlimit; [signal recordings](docs/signal-recordings.md)                                                     |
+| Packages and resources  | File inventories, digests, plists, Apple bundle anatomy and extracted resources                   | [Artifact and JavaScript guide](docs/javascript-artifact-reconstruction.md), [Apple applications](docs/apple-application-analysis.md) |
+| Process behavior        | Terminal output, interactions, exit and filesystem observations, and run comparisons              | Linux/macOS with a native PTY; [process capture](docs/process-capture.md)                                                             |
+| PyInstaller executables | CArchive members, Python version, entry scripts and extracted bytecode                            | None (static only); [Python packaged applications](docs/python-packaged-applications.md)                                              |
 
 Static JavaScript and .NET inspection read the supplied files without running
 the application. Runtime capture runs or interacts with the selected target
@@ -345,7 +369,7 @@ Report vulnerabilities through [SECURITY.md](SECURITY.md).
 
 ## Star history
 
-🎉 **250,000 GitHub stars — thank you!**
+🎉 **300,000 GitHub stars — thank you!**
 
 Thanks to everyone using REA, reporting bugs, requesting features, testing builds, and contributing fixes.
 

bridge/android/ReaJadxBridge.java

@@ -10,16 +10,24 @@ import jadx.api.plugins.input.insns.Opcode;
 import jadx.core.dex.info.MethodInfo;
 import jadx.core.dex.nodes.*;
 import java.io.PrintStream;
+import java.io.StringReader;
 import java.nio.charset.StandardCharsets;
 import java.util.*;
 import java.util.concurrent.CountDownLatch;
 import java.util.stream.Collectors;
 import kotlin.Unit;
+import kotlin.ResultKt;
+import kotlin.coroutines.Continuation;
+import kotlin.coroutines.CoroutineContext;
 import kotlin.coroutines.EmptyCoroutineContext;
+import kotlin.coroutines.intrinsics.IntrinsicsKt;
 import kotlinx.coroutines.BuildersKt;
 import kotlinx.io.CoreKt;
 import kotlinx.io.JvmCoreKt;
-import kotlinx.serialization.json.JsonObject;
+import javax.xml.parsers.DocumentBuilderFactory;
+import org.w3c.dom.Element;
+import org.w3c.dom.Node;
+import org.xml.sax.InputSource;
 
 /** REA-owned metadata boundary for the unmodified jadx-headless-mcp 0.7.1 JAR. */
 class ReaJadxBridge {
@@ -27,9 +35,13 @@ class ReaJadxBridge {
   private final SessionHolder holder;
   private JadxSession indexedSession;
   private Map<String, List<ClassSnapshot>> classes = Map.of();
+  private JadxSession manifestSession;
+  private ManifestSnapshot manifestSnapshot;
 
-  private record MethodSnapshot(String name, String shortId, Map<String, Object> summary) {}
+  private record MethodSnapshot(String name, String rawName, String shortId, List<String> bodyAbsenceMarkers,
+      Map<String, Object> summary) {}
   private record ClassSnapshot(JavaClass node, List<MethodSnapshot> methods, Map<String, Object> summary) {}
+  private record ManifestSnapshot(String text, Map<String, Object> summary) {}
 
   private static final class InputFailure extends IllegalArgumentException {
     final String field;
@@ -44,15 +56,94 @@ class ReaJadxBridge {
 
   private static CallToolResult json(Object value) { return text(JSON.toJson(value), false); }
 
+  /** Translate the typed load outcome before upstream drops its exception identity. */
+  static Object loadResult(Object result) {
+    try { ResultKt.throwOnFailure(result); }
+    catch (Throwable failure) {
+      if (failure instanceof java.util.concurrent.CancellationException
+          || failure instanceof Error && !(failure instanceof OutOfMemoryError)) return result;
+      if (failure instanceof OutOfMemoryError) {
+        String reason = failure.getMessage();
+        boolean heap = "Java heap space".equals(reason) || "GC overhead limit exceeded".equals(reason);
+        Map<String, Object> reply = new LinkedHashMap<>();
+        reply.put("kind", "jvm-memory-failure");
+        reply.put("memory_area", heap ? "heap" : "unknown");
+        reply.put("exception_type", "java.lang.OutOfMemoryError");
+        if (reason != null) reply.put("message", reason);
+        reply.put("max_heap_bytes", Runtime.getRuntime().maxMemory());
+        return text(JSON.toJson(reply), true);
+      }
+      return text("load failed: " + failure.getMessage(), true);
+    }
+    SessionHolder.LoadResult loaded = (SessionHolder.LoadResult) result;
+    Map<String, Object> reply = new LinkedHashMap<>();
+    reply.put("state", "LOADED");
+    reply.put("apk_path", loaded.getApkPath());
+    reply.put("class_count", loaded.getClassCount());
+    reply.put("raw_class_count", loaded.getRawClassCount());
+    reply.put("resource_count", loaded.getResourceCount());
+    reply.put("load_duration_ms", loaded.getLoadDurationMs());
+    reply.put("decompile_timeout_ms", loaded.getDecompileTimeoutMs());
+    reply.put("threads", loaded.getThreads());
+    reply.put("code_cache_size", loaded.getCodeCacheSize());
+    reply.put("resources", loaded.getResourceMode());
+    if (!loaded.getIncludePackages().isEmpty()) reply.put("include_packages", loaded.getIncludePackages());
+    if (!loaded.getExcludePackages().isEmpty()) reply.put("exclude_packages", loaded.getExcludePackages());
+    return json(reply);
+  }
+
+  private Object load(kotlinx.serialization.json.JsonObject arguments, SessionConfig defaults,
+      Continuation<? super CallToolResult> continuation) {
+    JsonObject args = JsonParser.parseString(arguments.toString()).getAsJsonObject();
+    String path = string(args, "path");
+    SessionConfig options = defaults.withLoadOverrides(
+        optionalString(args, "include_packages"), optionalString(args, "exclude_packages"),
+        args.has("threads") ? integer(args, "threads", 1) : null,
+        args.has("code_cache_size") ? integer(args, "code_cache_size", 64) : null,
+        optionalString(args, "resources"));
+    clearManifest();
+    return loadCall(next -> holder.load(path, options, next), continuation);
+  }
+
+  static Object loadCall(kotlin.jvm.functions.Function1<Continuation<SessionHolder.LoadResult>, Object> load,
+      Continuation<? super CallToolResult> continuation) {
+    var forwarded = new java.util.concurrent.atomic.AtomicBoolean();
+    Continuation<SessionHolder.LoadResult> resumed = new Continuation<>() {
+      public CoroutineContext getContext() { return continuation.getContext(); }
+      public void resumeWith(Object result) {
+        if (!forwarded.compareAndSet(false, true)) throw new IllegalStateException("Load resumed twice");
+        Object translated;
+        try { translated = loadResult(result); }
+        catch (Throwable failure) { translated = ResultKt.createFailure(failure); }
+        // Keep the consumer outside the translation catch, including inline completion.
+        continuation.resumeWith(translated);
+      }
+    };
+    Object result;
+    try { result = load.invoke(resumed); }
+    catch (Throwable failure) {
+      if (forwarded.get()) ResultKt.throwOnFailure(ResultKt.createFailure(failure));
+      result = ResultKt.createFailure(failure);
+    }
+    if (result == IntrinsicsKt.getCOROUTINE_SUSPENDED()) return result;
+    Object translated = loadResult(result);
+    ResultKt.throwOnFailure(translated);
+    return translated;
+  }
+
+  private static String optionalString(JsonObject args, String name) {
+    return args.has(name) ? string(args, name) : null;
+  }
+
   private static String string(JsonObject args, String name) {
-    JsonElement value = JsonParser.parseString(args.toString()).getAsJsonObject().get(name);
+    JsonElement value = args.get(name);
     if (value == null || !value.isJsonPrimitive() || !value.getAsJsonPrimitive().isString())
       throw new IllegalArgumentException(name + " must be a string");
     return value.getAsString();
   }
 
   private static int integer(JsonObject args, String name, int fallback) {
-    JsonElement value = JsonParser.parseString(args.toString()).getAsJsonObject().get(name);
+    JsonElement value = args.get(name);
     if (value == null) return fallback;
     if (!value.isJsonPrimitive() || !value.getAsJsonPrimitive().isNumber())
       throw new IllegalArgumentException(name + " must be a nonnegative integer");
@@ -71,6 +162,77 @@ class ReaJadxBridge {
     return current;
   }
 
+  private void clearManifest() {
+    manifestSession = null;
+    manifestSnapshot = null;
+  }
+
+  /** Original archive identity, never a deobfuscated basename or name suffix. */
+  private ManifestSnapshot manifest(JadxSession current) {
+    if (manifestSession != current || manifestSnapshot == null) {
+      ResourceFile selected = null;
+      for (ResourceFile resource : current.getResources()) {
+        if (resource.getType() != ResourceType.MANIFEST
+            || !resource.getOriginalName().equals("AndroidManifest.xml")) continue;
+        if (selected != null)
+          throw new IllegalArgumentException("Root AndroidManifest.xml resource identity is ambiguous");
+        selected = resource;
+      }
+      if (selected == null) throw new IllegalArgumentException("Root AndroidManifest.xml not found");
+      var content = selected.loadContent();
+      if (content == null || content.getDataType() != jadx.core.xmlgen.ResContainer.DataType.TEXT)
+        throw new IllegalArgumentException("Root AndroidManifest.xml did not decode as text");
+      String xml = content.getText().getCodeStr();
+      ManifestSnapshot snapshot = new ManifestSnapshot(xml, manifestSummary(xml));
+      manifestSnapshot = snapshot;
+      manifestSession = current;
+    }
+    return manifestSnapshot;
+  }
+
+  /** One parsed XML owner supplies namespace-aware values alongside the exact decoded text. */
+  static Map<String, Object> manifestSummary(String xml) {
+    Element root;
+    try {
+      // Match REA's existing JDK XML setup: document types are unsupported.
+      // Parse an in-memory string; XInclude and external DTD/schema I/O stay disabled.
+      DocumentBuilderFactory factory = DocumentBuilderFactory.newDefaultNSInstance();
+      factory.setFeature("http://apache.org/xml/features/disallow-doctype-decl", true);
+      factory.setXIncludeAware(false);
+      factory.setAttribute(javax.xml.XMLConstants.ACCESS_EXTERNAL_DTD, "");
+      factory.setAttribute(javax.xml.XMLConstants.ACCESS_EXTERNAL_SCHEMA, "");
+      root = factory.newDocumentBuilder().parse(new InputSource(new StringReader(xml)))
+          .getDocumentElement();
+    } catch (javax.xml.parsers.ParserConfigurationException | org.xml.sax.SAXException
+        | java.io.IOException failure) {
+      throw new IllegalArgumentException("Root AndroidManifest.xml is not supported well-formed XML", failure);
+    }
+    if (!root.getTagName().equals("manifest") || root.getNamespaceURI() != null)
+      throw new IllegalArgumentException("AndroidManifest.xml root must be an unnamespaced <manifest>");
+    String android = "http://schemas.android.com/apk/res/android";
+    String minSdk = "";
+    String targetSdk = "";
+    List<String> permissions = new ArrayList<>();
+    for (Node child = root.getFirstChild(); child != null; child = child.getNextSibling()) {
+      if (!(child instanceof Element element) || element.getNamespaceURI() != null) continue;
+      if (element.getTagName().equals("uses-sdk")) {
+        minSdk = element.getAttributeNS(android, "minSdkVersion");
+        targetSdk = element.getAttributeNS(android, "targetSdkVersion");
+      } else if (element.getTagName().equals("uses-permission")) {
+        String permission = element.getAttributeNS(android, "name");
+        if (!permission.isEmpty()) permissions.add(permission);
+      }
+    }
+    Map<String, Object> summary = new LinkedHashMap<>();
+    summary.put("package", root.getAttribute("package"));
+    summary.put("version_name", root.getAttributeNS(android, "versionName"));
+    summary.put("version_code", root.getAttributeNS(android, "versionCode"));
+    summary.put("min_sdk", minSdk);
+    summary.put("target_sdk", targetSdk);
+    summary.put("permissions", List.copyOf(permissions));
+    return Collections.unmodifiableMap(summary);
+  }
+
   private ClassSnapshot selected(JsonObject args) {
     JadxSession current = session();
     if (indexedSession != current) {
@@ -90,10 +252,13 @@ class ReaJadxBridge {
             if (!shortId.startsWith(rawName + "("))
               throw new IllegalStateException("JADX method ID has no raw-name prefix: " + shortId);
             String descriptor = shortId.substring(rawName.length());
+            List<String> bodyAbsenceMarkers = new ArrayList<>();
+            if (method.getAccessFlags().isNative()) bodyAbsenceMarkers.add("native_method");
+            if (method.getAccessFlags().isAbstract()) bodyAbsenceMarkers.add("abstract_method");
             String signature = name + "(" + method.getArgTypes().stream()
                 .map(Object::toString).collect(Collectors.joining(", "))
                 + "): " + method.getReturnType().toString();
-            return new MethodSnapshot(name, shortId, Map.of(
+            return new MethodSnapshot(name, rawName, shortId, List.copyOf(bodyAbsenceMarkers), Map.of(
                 "name", name, "signature", signature,
                 "dex_descriptor", descriptor,
                 "is_constructor", method.isConstructor(), "def_pos", 0));
@@ -140,16 +305,52 @@ class ReaJadxBridge {
 
   private JadxSession.SmartCode methodBody(JadxSession current, ClassSnapshot cls,
       MethodSnapshot selected, JavaMethod method) {
-    if (!selected.shortId().equals("<clinit>()V"))
-      return current.getMethodBodySmart(method, current.getMaxSourceBytes(), true);
+    // Access flags belong to the parsed member, before code generation rewrites it.
+    // These methods have no Java body; source slicing can include unrelated members.
+    if (!selected.bodyAbsenceMarkers().isEmpty())
+      return new JadxSession.SmartCode("", "java", false, selected.bodyAbsenceMarkers());
     // Java code generation can fold all or part of <clinit> into field declarations.
     // Its method text cannot establish coverage, even when a static block remains.
-    JadxSession.MethodSmali smali = current.getMethodSmali(cls.node(), method.getName());
-    if (!smali.getFound() || smali.getBlocks().size() != 1)
-      throw new IllegalArgumentException("Static initializer disassembly requires one exact method: " + selected.shortId());
+    if (selected.shortId().equals("<clinit>()V"))
+      return exactMethodSmali(current, cls, selected, List.of("static_initializer_disassembly"));
+    String java = current.getMethodSource(method);
+    if (java.startsWith("// ERROR:")) throw new IllegalArgumentException(java);
+    // The upstream placeholder means only that Java code was omitted, including
+    // methods merged into synthetic bridges. Access flags established absence above.
+    if (java.isEmpty() || java.equals(
+        "// method exists but has no decompiled body (native/abstract)"))
+      return exactMethodSmali(current, cls, selected, List.of("java_body_unavailable"));
+    List<String> markers = DecompileFailureKt.detectDecompileFailure(java);
+    if (markers.stream().anyMatch(DecompileFailureKt.getSTRONG_FAILURE_MARKERS()::contains))
+      return exactMethodSmali(current, cls, selected, markers);
     return new JadxSession.SmartCode(
-        TextUtilKt.truncateToBytes(smali.getBlocks().get(0), current.getMaxSourceBytes()),
-        "smali", true, List.of("static_initializer_disassembly"));
+        TextUtilKt.truncateToBytes(java, current.getMaxSourceBytes()), "java", false, markers);
+  }
+
+  private JadxSession.SmartCode exactMethodSmali(JadxSession current, ClassSnapshot cls,
+      MethodSnapshot selected, List<String> markers) {
+    JadxSession.MethodSmali smali = current.getMethodSmali(cls.node(), selected.rawName());
+    String block = selectMethodSmali(smali, selected.shortId());
+    return new JadxSession.SmartCode(
+        TextUtilKt.truncateToBytes(block, current.getMaxSourceBytes()), "smali", true, markers);
+  }
+
+  private static String selectMethodSmali(JadxSession.MethodSmali smali, String shortId) {
+    if (!smali.getFound())
+      throw new IllegalArgumentException("JADX could not disassemble " + shortId
+          + (smali.getNote() == null ? "" : ": " + smali.getNote()));
+    // Upstream returns every same-name overload, and can return an unterminated
+    // trailing block. Match the full original identity before applying output limits.
+    List<String> matches = smali.getBlocks().stream().filter(block -> {
+      String header = block.lines().findFirst().orElse("").strip();
+      return header.startsWith(".method ") && header.endsWith(" " + shortId);
+    }).toList();
+    if (matches.size() != 1)
+      throw new IllegalArgumentException("Disassembly could not establish one selected method: " + shortId);
+    String block = matches.get(0);
+    if (!block.stripTrailing().endsWith("\n.end method"))
+      throw new IllegalArgumentException("Selected method disassembly is incomplete: " + shortId);
+    return block;
   }
 
   // The local transport admits 8 MiB frames. Reserve 1 KiB for the fixed
@@ -203,7 +404,8 @@ class ReaJadxBridge {
     return new DirectInvokes(sites, exhaustive[0]);
   }
 
-  private CallToolResult handle(String name, JsonObject args) {
+  private CallToolResult handle(String name, kotlinx.serialization.json.JsonObject input) {
+    JsonObject args = JsonParser.parseString(input.toString()).getAsJsonObject();
     if (name.equals("rea_jvm_status")) return json(Map.of(
         "engine_reported_version", BuildInfo.VERSION,
         "max_heap_bytes", Runtime.getRuntime().maxMemory(),
@@ -211,14 +413,30 @@ class ReaJadxBridge {
         "java_version", System.getProperty("java.version"),
         "metadata_scope", "parsed_members"));
     JadxSession current = session();
+    if (name.equals("get_app_info")) return json(manifest(current).summary());
+    if (name.equals("get_android_manifest"))
+      return text(TextUtilKt.truncateToBytes(manifest(current).text(), current.getMaxSourceBytes()), false);
     if (name.equals("list_classes")) {
       List<String> names = current.getClassFqns();
+      String query = args.has("query") ? string(args, "query") : "";
       int offset = integer(args, "offset", 0);
       int limit = integer(args, "limit", 200);
       int start = Math.min(offset, names.size());
-      int end = (int) Math.min((long) start + limit, names.size());
-      return json(Map.of("total", names.size(), "offset", offset, "limit", limit,
-          "items", names.subList(start, end)));
+      int end = start;
+      List<String> items;
+      if (query.isEmpty()) {
+        end = (int) Math.min((long) start + limit, names.size());
+        items = names.subList(start, end);
+      } else {
+        items = new ArrayList<>();
+        while (end < names.size() && items.size() < limit) {
+          String nameInInventory = names.get(end++);
+          if (nameInInventory.contains(query)) items.add(nameInInventory);
+        }
+      }
+      return json(Map.of("total_classes", names.size(), "next_offset", end,
+          "scope", "top_level", "offset", offset, "limit", limit,
+          "items", items));
     }
     ClassSnapshot cls = selected(args);
     if (name.equals("get_class_summary")) return json(cls.summary());
@@ -230,6 +448,7 @@ class ReaJadxBridge {
       JadxSession.SmartCode code = methodBody(current, cls, overloads.get(index), method);
       return json(Map.of("class_name", cls.summary().get("full_name"),
           "method_name", overloads.get(index).name(), "full_name", method.getFullName(),
+          "dex_descriptor", overloads.get(index).summary().get("dex_descriptor"),
           "overload_index", index, "overload_count", overloads.size(),
           "mode", code.getKind(), "fell_back", code.getFellBack(),
           "markers", code.getMarkers(), "body", code.getText()));
@@ -241,7 +460,7 @@ class ReaJadxBridge {
     if (name.equals("get_xrefs_to_class")) uses = cls.node().getUseIn();
     else {
       List<MethodSnapshot> overloads = methods(cls, args);
-      if (overloads.size() != 1 && !args.containsKey("overload_index"))
+      if (overloads.size() != 1 && !args.has("overload_index"))
         throw new InputFailure("overload_index", "Select one of " + overloads.size() + " method overloads");
       overloadIndex = integer(args, "overload_index", 0);
       if (overloadIndex >= overloads.size())
@@ -284,30 +503,52 @@ class ReaJadxBridge {
     return json(result);
   }
 
+  private static void registerBridgeTool(Server server, ReaJadxBridge bridge, String name,
+      ToolSchema schema, String description) {
+    Tool tool = new Tool(name, schema, description, null, null, null, null, null, null);
+    server.addTool(tool, (connection, request, continuation) -> {
+      try { return bridge.handle(name, request.getArguments()); }
+      catch (InputFailure error) { return text(JSON.toJson(Map.of(
+          "kind", "invalid-input", "field", error.field, "message", error.getMessage())), true); }
+      catch (IllegalArgumentException error) { return text(error.getMessage(), true); }
+    });
+  }
+
   public static void main(String[] args) throws Exception {
     PrintStream protocolOut = System.out;
     System.setOut(System.err);
-    SessionHolder holder = new SessionHolder(new SessionConfig(
-        1048576, 0, 90000L, 1, List.of(), List.of(), 64, ResourceMode.FULL));
+    SessionConfig defaults = new SessionConfig(
+        1048576, 0, 90000L, 1, List.of(), List.of(), 64, ResourceMode.FULL);
+    SessionHolder holder = new SessionHolder(defaults);
     ServerCapabilities capabilities = new ServerCapabilities(
         new ServerCapabilities.Tools(null), null, null, null, null, null, null, Map.of());
-    Server server = new Server(new Implementation("rea-jadx-bridge", "1", null, null, null),
+    Server server = new Server(new Implementation("rea-jadx-bridge", "5", null, null, null),
         new ServerOptions(capabilities), "", ignored -> Unit.INSTANCE);
     SessionToolsKt.registerSessionTools(server, holder);
-    ManifestToolsKt.registerManifestTools(server, holder);
     ClassToolsKt.registerClassTools(server, holder);
     MethodToolsKt.registerMethodTools(server, holder);
     XrefToolsKt.registerXrefTools(server, holder);
     ReaJadxBridge bridge = new ReaJadxBridge(holder);
-    Set<String> consumed = Set.of("load_apk", "get_app_info", "get_android_manifest",
+    Tool loadTool = server.getTools().get("load_apk").getTool();
+    server.removeTool("load_apk");
+    server.addTool(loadTool, (connection, request, continuation) ->
+        bridge.load(request.getArguments(), defaults, continuation));
+    Set<String> consumed = Set.of("load_apk",
         "list_classes", "get_class_summary", "get_method_body",
         "get_xrefs_to_class", "get_xrefs_to_method");
     for (String name : new ArrayList<>(server.getTools().keySet()))
       if (!consumed.contains(name)) server.removeTool(name);
     Tool diagnostics = new Tool("rea_jvm_status",
-        new ToolSchema(null, new JsonObject(Map.of()), List.of(), null),
+        new ToolSchema(null, new kotlinx.serialization.json.JsonObject(Map.of()), List.of(), null),
         "Observed worker JVM and metadata scope", null, null, null, null, null, null);
     server.addTool(diagnostics, (connection, request, continuation) -> bridge.handle("rea_jvm_status", request.getArguments()));
+    for (String name : List.of("get_app_info", "get_android_manifest")) {
+      String description = name.equals("get_app_info")
+          ? "Namespace-aware summary of the exact root AndroidManifest.xml resource."
+          : "Exact decoded root AndroidManifest.xml text, bounded by the session source byte limit.";
+      registerBridgeTool(server, bridge, name,
+          new ToolSchema(null, new kotlinx.serialization.json.JsonObject(Map.of()), List.of(), null), description);
+    }
     for (String name : List.of("list_classes", "get_class_summary",
         "get_method_body", "get_xrefs_to_class", "get_xrefs_to_method")) {
       Tool tool = server.getTools().get(name).getTool();
@@ -320,20 +561,18 @@ class ReaJadxBridge {
       };
       Map<String, kotlinx.serialization.json.JsonElement> properties = new LinkedHashMap<>();
       for (String field : fields) properties.put(field, tool.getInputSchema().getProperties().get(field));
+      if (name.equals("list_classes")) properties.put("query",
+          new kotlinx.serialization.json.JsonObject(Map.of(
+              "type", kotlinx.serialization.json.JsonElementKt.JsonPrimitive("string"))));
       String description = switch (name) {
-        case "list_classes" -> "Indexed pages of all top-level class names, with offset and limit.";
+        case "list_classes" -> "Pages of case-sensitive matching top-level class names; offset and next_offset are inventory positions, limit bounds matching rows. Omitted or empty query inventories all names.";
         case "get_class_summary" -> "Parsed member metadata for one exact class name; no source generation.";
         case "get_method_body" -> "Source for an exact class and stable same-name overload index; static initializers use smali to preserve folded field initialization.";
         default -> "Complete static using-node inventory for an exact class or selected method overload; one node may contain multiple invoke instructions, and source lines are unresolved.";
       };
-      Tool narrowed = new Tool(name, new ToolSchema(null, new JsonObject(properties),
-          tool.getInputSchema().getRequired(), null), description, null, null, null, null, null, null);
-      server.addTool(narrowed, (connection, request, continuation) -> {
-        try { return bridge.handle(name, request.getArguments()); }
-        catch (InputFailure error) { return text(JSON.toJson(Map.of(
-            "kind", "invalid-input", "field", error.field, "message", error.getMessage())), true); }
-        catch (IllegalArgumentException error) { return text(error.getMessage(), true); }
-      });
+      registerBridgeTool(server, bridge, name,
+          new ToolSchema(null, new kotlinx.serialization.json.JsonObject(properties),
+              tool.getInputSchema().getRequired(), null), description);
     }
     CountDownLatch done = new CountDownLatch(1);
     server.onClose(() -> { done.countDown(); return Unit.INSTANCE; });
@@ -359,6 +598,7 @@ class ReaJadxBridge {
           BuildersKt.runBlocking(EmptyCoroutineContext.INSTANCE,
               (scope, continuation) -> server.close(continuation));
         } finally {
+          bridge.clearManifest();
           BuildersKt.runBlocking(EmptyCoroutineContext.INSTANCE,
               (scope, continuation) -> holder.unload(continuation));
         }

bridge/cutter/rea_bridge.py

@@ -11,6 +11,7 @@ import json
 import hashlib
 import hmac
 import os
+import re
 import secrets
 import socket
 import stat
@@ -28,14 +29,156 @@ if os.name == "nt":
 import cutter
 
 try:
-    from PySide6.QtCore import QObject, Signal
+    from PySide6.QtCore import QCoreApplication, QObject, Signal
 except ImportError:
-    from PySide2.QtCore import QObject, Signal
+    from PySide2.QtCore import QCoreApplication, QObject, Signal
+
+
+# Request and response frame budget in bytes, excluding the trailing LF.
+# REA_BRIDGE_MAX_REQUEST_BYTES selects it. Unset stays 16777216.
+# 7464 is the authenticated frame of the largest bounded symbol name.
+# The 64-receipt recovery status is smaller with current_file null. A path
+# that fits is kept. A path that does not fit is null, not the descriptor path.
+# 536870887 is Node's 536870888-character string ceiling minus the framing LF.
+# The ceiling is the encoded payload; JSON expansion is measured before it.
+# A fitting response is signed from bounded JSON pieces and written in 64 KiB
+# chunks, so this maximum does not retain several copies of the frame.
+_BRIDGE_MAX_REQUEST_BYTES_SETTING = "REA_BRIDGE_MAX_REQUEST_BYTES"
+DEFAULT_BRIDGE_MAX_REQUEST_BYTES = 16 * 1024 * 1024
+MINIMUM_BRIDGE_MAX_REQUEST_BYTES = 7464
+MAXIMUM_BRIDGE_MAX_REQUEST_BYTES = 536870887
+# The outer authenticated frame around the escaped JSON payload, excluding LF.
+_AUTHENTICATED_PREFIX = b'{"response":"'
+_AUTHENTICATED_PROOF_MARK = b'","proof":"'
+_PROOF_HEX_LENGTH = 64
+_AUTHENTICATED_FRAME_OVERHEAD = (
+    len(_AUTHENTICATED_PREFIX)
+    + len(_AUTHENTICATED_PROOF_MARK)
+    + _PROOF_HEX_LENGTH
+    + len(b'"}')
+)
+_AUTHENTICATED_WRITE_BUFFER_BYTES = 64 * 1024
+# Raw JSON is produced in pieces no larger than this many characters.
+# One character expands to at most six JSON bytes (\u00XX) before the outer
+# string escape, so a piece stays far below the frame ceiling.
+_JSON_CHUNK_CHARS = 64 * 1024
+# Fixed admission cap. It does not follow REA_BRIDGE_MAX_REQUEST_BYTES.
+# A further accept is closed. Starting a worker must not stop the listener.
+MAX_CONNECTION_WORKERS = 8
+
+
+def _close_socket(connection: socket.socket) -> None:
+    try:
+        connection.close()
+    except OSError:
+        pass
+
+
+def _bridge_max_request_bytes_constraint():
+    return (
+        "REA_BRIDGE_MAX_REQUEST_BYTES must be a positive decimal integer "
+        "from %d to %d bytes; the default is %d bytes (16 MiB). "
+        "A value below %d bytes cannot carry the largest admitted "
+        "symbol-name command, and a larger value is refused so the encoded "
+        "payload plus its framing LF stays one string under Node's "
+        "single-string ceiling."
+        % (
+            MINIMUM_BRIDGE_MAX_REQUEST_BYTES,
+            MAXIMUM_BRIDGE_MAX_REQUEST_BYTES,
+            DEFAULT_BRIDGE_MAX_REQUEST_BYTES,
+            MINIMUM_BRIDGE_MAX_REQUEST_BYTES,
+        )
+    )
+
+
+def _parse_bridge_max_request_bytes(raw):
+    """Refuse a non-decimal or out-of-range ceiling. Never fall back."""
+    constraint = _bridge_max_request_bytes_constraint()
+    if isinstance(raw, bool):
+        raise RuntimeError(constraint)
+    if isinstance(raw, int):
+        value = raw
+    elif isinstance(raw, str):
+        # ASCII digits only. str.isdigit() accepts numerals such as U+0668
+        # that the TypeScript setting rejects.
+        if re.fullmatch(r"[1-9][0-9]*", raw) is None:
+            raise RuntimeError(constraint)
+        value = int(raw)
+    else:
+        raise RuntimeError(constraint)
+    if (
+        value < MINIMUM_BRIDGE_MAX_REQUEST_BYTES
+        or value > MAXIMUM_BRIDGE_MAX_REQUEST_BYTES
+    ):
+        raise RuntimeError(constraint)
+    return value
+
+
+def _resolve_bridge_max_request_bytes():
+    """Prefer an injected global, then the process environment, then 16 MiB."""
+    injected = globals().get(_BRIDGE_MAX_REQUEST_BYTES_SETTING)
+    if injected is not None:
+        return _parse_bridge_max_request_bytes(injected), "setting"
+    raw = os.environ.get(_BRIDGE_MAX_REQUEST_BYTES_SETTING)
+    if raw is None:
+        return DEFAULT_BRIDGE_MAX_REQUEST_BYTES, "default"
+    return _parse_bridge_max_request_bytes(raw), "setting"
+
+
+MAX_FRAME_BYTES, BRIDGE_MAX_REQUEST_BYTES_SOURCE = (
+    _resolve_bridge_max_request_bytes()
+)
+
+
+def _bridge_max_request_bytes_recovery():
+    return (
+        "The effective REA_BRIDGE_MAX_REQUEST_BYTES value is %d bytes "
+        "(source %s; default %d, floor %d, maximum %d). "
+        "Set REA_BRIDGE_MAX_REQUEST_BYTES in the environment that starts REA "
+        "to a decimal integer from %d to %d, then restart REA. "
+        "Restart Hopper or Cutter from that same environment so the bridge "
+        "uses this ceiling."
+        % (
+            MAX_FRAME_BYTES,
+            BRIDGE_MAX_REQUEST_BYTES_SOURCE,
+            DEFAULT_BRIDGE_MAX_REQUEST_BYTES,
+            MINIMUM_BRIDGE_MAX_REQUEST_BYTES,
+            MAXIMUM_BRIDGE_MAX_REQUEST_BYTES,
+            MINIMUM_BRIDGE_MAX_REQUEST_BYTES,
+            MAXIMUM_BRIDGE_MAX_REQUEST_BYTES,
+        )
+    )
+
+
+def _frame_limit_exceeded():
+    return ValueError(
+        "frame-limit-exceeded: Bridge request exceeds the %d-byte payload limit. %s"
+        % (MAX_FRAME_BYTES, _bridge_max_request_bytes_recovery())
+    )
 
 
-MAX_FRAME_BYTES = 16 * 1024 * 1024
 SOCKET_TIMEOUT_SECONDS = 20
 AUTH_FRAME_TIMEOUT_SECONDS = 1
+# The client request deadline. The authenticated frame read may use up to
+# this long; it does not grow past it when the byte ceiling is raised.
+REQUEST_DEADLINE_SECONDS = 25
+
+
+def _authenticated_frame_timeout_seconds():
+    """Time to hash and transfer one maximum-sized authenticated frame.
+
+    The hello frame stays at one second. The following command frame is
+    built only after that challenge, so its read budget follows the
+    configured payload size, at one second per mebibyte, and never exceeds
+    the 25 second client deadline.
+    """
+    mebibytes = (MAX_FRAME_BYTES + (1024 * 1024) - 1) // (1024 * 1024)
+    return min(
+        REQUEST_DEADLINE_SECONDS,
+        max(AUTH_FRAME_TIMEOUT_SECONDS, mebibytes),
+    )
+
+
 MAX_DESCRIPTOR_BYTES = 64 * 1024
 MAX_COMPLETION_RECEIPTS = 64
 
@@ -326,15 +469,16 @@ class _RequestDispatcher(QObject):
         self._receipt_lock = threading.Lock()
         self.current_file = self._read_current_file()
         try:
-            self.cutter_version = str(cutter.core().getVersionInformation())
+            value = QCoreApplication.applicationVersion()
+            self.cutter_version = value if isinstance(value, str) and value else None
         except Exception:
             self.cutter_version = None
 
     @staticmethod
     def _read_current_file() -> str | None:
         try:
-            value = cutter.core().currentlyOpenFile()
-            return str(value) if value else None
+            value = cutter.core().getConfig("file.path")
+            return value if isinstance(value, str) and value else None
         except Exception:
             return None
 
@@ -479,6 +623,17 @@ class _BridgeServer(threading.Thread):
         self.token = secrets.token_urlsafe(32)
         self.poisoned = False
         self.connection_ordinal = 0
+        # The client proof is inside the authenticated frame, so that read
+        # cannot start after proof. It runs off the accept loop. One
+        # authenticated command still runs at a time. completed_connection
+        # is the longest prefix of these ordinals whose handlers have
+        # terminated, not "this status ordinal minus one".
+        self._command_lock = threading.Lock()
+        self._connections_lock = threading.Lock()
+        self._connections: set[socket.socket] = set()
+        self._workers: list[threading.Thread] = []
+        self._finished_through = 0
+        self._finished_ordinals: set[int] = set()
         self.directory = _bridge_directory()
         self.descriptor = self.directory / f"cutter-{os.getpid()}-{self.session_id}.json"
 
@@ -499,6 +654,7 @@ class _BridgeServer(threading.Thread):
             "current_file": self.dispatcher.current_file,
             "cutter_version": self.dispatcher.cutter_version,
             "identity_status": "partial",
+            "max_request_bytes": MAX_FRAME_BYTES,
         }
         if os.name == "nt":
             _write_windows_descriptor(self.directory, self.descriptor, json.dumps(payload).encode("utf-8"))
@@ -540,8 +696,46 @@ class _BridgeServer(threading.Thread):
                 continue
             except OSError:
                 break
+            with self._connections_lock:
+                self._workers = [
+                    worker for worker in self._workers if worker.is_alive()
+                ]
+                at_capacity = len(self._workers) >= MAX_CONNECTION_WORKERS
+            if at_capacity:
+                _close_socket(connection)
+                continue
             self.connection_ordinal += 1
             connection_ordinal = self.connection_ordinal
+            worker = threading.Thread(
+                target=self._serve_connection,
+                args=(connection, connection_ordinal),
+                name="rea-cutter-bridge-connection",
+                daemon=True,
+            )
+            with self._connections_lock:
+                self._connections.add(connection)
+                self._workers.append(worker)
+            try:
+                worker.start()
+            except RuntimeError:
+                with self._connections_lock:
+                    self._connections.discard(connection)
+                    if worker in self._workers:
+                        self._workers.remove(worker)
+                self._note_connection_finished(connection_ordinal)
+                _close_socket(connection)
+        with self._connections_lock:
+            workers = list(self._workers)
+        for worker in workers:
+            worker.join(timeout=1)
+        self.server.close()
+        try:
+            self.descriptor.unlink(missing_ok=True)
+        except OSError:
+            pass
+
+    def _serve_connection(self, connection: socket.socket, connection_ordinal: int) -> None:
+        try:
             with connection:
                 connection.settimeout(AUTH_FRAME_TIMEOUT_SECONDS)
                 authenticated = False
@@ -549,60 +743,73 @@ class _BridgeServer(threading.Thread):
                     hello = self._receive_frame(connection)
                     if not isinstance(hello, dict) or hello.get("kind") != "rea-auth-v2" or not isinstance(hello.get("nonce"), str) or len(hello["nonce"]) != 64:
                         self._send(connection, {"ok": False, "error": "authentication-failed"})
-                        continue
+                        return
                     client_nonce = hello["nonce"]
                     server_nonce = f"{connection_ordinal:016x}{secrets.token_hex(24)}"
                     server_proof = hmac.new(self.token.encode(), f"rea-cutter-server-v2:{self.session_id}:{client_nonce}:{server_nonce}".encode(), hashlib.sha256).hexdigest()
                     self._send(connection, {"session_id": self.session_id, "nonce": server_nonce, "proof": server_proof})
-                    authenticated_frame = self._receive_frame(connection)
+                    authenticated_frame = self._receive_frame(
+                        connection, _authenticated_frame_timeout_seconds()
+                    )
                     request_bytes = authenticated_frame.get("request") if isinstance(authenticated_frame, dict) else None
-                    expected_proof = hmac.new(self.token.encode(), f"rea-cutter-client-v2:{self.session_id}:{client_nonce}:{server_nonce}:{request_bytes}".encode(), hashlib.sha256).hexdigest()
-                    if not isinstance(request_bytes, str) or not isinstance(authenticated_frame.get("proof"), str) or not hmac.compare_digest(authenticated_frame["proof"], expected_proof):
+                    presented_proof = authenticated_frame.get("proof") if isinstance(authenticated_frame, dict) else None
+                    if not isinstance(request_bytes, str) or not isinstance(presented_proof, str):
                         self._send(connection, {"ok": False, "error": "authentication-failed"})
-                        continue
+                        return
+                    # Update the prefix and the request separately. One joined
+                    # preimage would retain another copy of a maximum-sized command.
+                    expected_proof = hmac.new(self.token.encode(), digestmod=hashlib.sha256)
+                    expected_proof.update(
+                        f"rea-cutter-client-v2:{self.session_id}:{client_nonce}:{server_nonce}:".encode()
+                    )
+                    expected_proof.update(request_bytes.encode())
+                    if not hmac.compare_digest(presented_proof, expected_proof.hexdigest()):
+                        self._send(connection, {"ok": False, "error": "authentication-failed"})
+                        return
                     request = json.loads(request_bytes)
                     if not isinstance(request, dict) or request.get("session_id") != self.session_id:
                         self._send(connection, {"ok": False, "error": "authentication-failed"})
-                        continue
+                        return
                     authenticated = True
-                    connection.settimeout(SOCKET_TIMEOUT_SECONDS)
-                    if self.poisoned and request.get("kind") not in ("status", "ack"):
-                        self._send_authenticated(connection, {
-                            "ok": False,
-                            "error": "session-poisoned-after-ui-timeout",
-                            "execution_state": "not_started",
-                        }, client_nonce, server_nonce)
-                        continue
-                    if (
-                        request.get("kind") == "command"
-                        and not self.dispatcher.can_accept_command()
-                    ):
-                        self._send_authenticated(connection, {
-                            "ok": False,
-                            "error": "completion-receipt-capacity-exceeded",
-                            "execution_state": "not_started",
-                        }, client_nonce, server_nonce)
-                        continue
-                    request["session_id"] = self.session_id
-                    event = threading.Event()
-                    cancelled = threading.Event()
-                    reply: dict[str, Any] = {}
-                    request["_event"] = event
-                    request["_cancelled"] = cancelled
-                    self.dispatcher.dispatch.emit(request, reply)
-                    if not event.wait(SOCKET_TIMEOUT_SECONDS):
-                        cancelled.set()
-                        self.poisoned = True
-                        self._send_authenticated(connection, {
-                            "ok": False,
-                            "error": "ui-thread-request-timeout",
-                            "execution_state": "unknown",
-                            "message": "The UI operation may still complete after this timeout; restart Cutter before issuing further commands.",
-                        }, client_nonce, server_nonce)
-                    else:
-                        if request.get("kind") == "status" and reply.get("ok") is True:
-                            reply["completed_connection"] = f"{connection_ordinal - 1:016x}"
-                        self._send_authenticated(connection, reply, client_nonce, server_nonce)
+                    with self._command_lock:
+                        connection.settimeout(SOCKET_TIMEOUT_SECONDS)
+                        if self.poisoned and request.get("kind") not in ("status", "ack"):
+                            self._send_authenticated(connection, {
+                                "ok": False,
+                                "error": "session-poisoned-after-ui-timeout",
+                                "execution_state": "not_started",
+                            }, client_nonce, server_nonce)
+                            return
+                        if (
+                            request.get("kind") == "command"
+                            and not self.dispatcher.can_accept_command()
+                        ):
+                            self._send_authenticated(connection, {
+                                "ok": False,
+                                "error": "completion-receipt-capacity-exceeded",
+                                "execution_state": "not_started",
+                            }, client_nonce, server_nonce)
+                            return
+                        request["session_id"] = self.session_id
+                        event = threading.Event()
+                        cancelled = threading.Event()
+                        reply: dict[str, Any] = {}
+                        request["_event"] = event
+                        request["_cancelled"] = cancelled
+                        self.dispatcher.dispatch.emit(request, reply)
+                        if not event.wait(SOCKET_TIMEOUT_SECONDS):
+                            cancelled.set()
+                            self.poisoned = True
+                            self._send_authenticated(connection, {
+                                "ok": False,
+                                "error": "ui-thread-request-timeout",
+                                "execution_state": "unknown",
+                                "message": "The UI operation may still complete after this timeout; restart Cutter before issuing further commands.",
+                            }, client_nonce, server_nonce)
+                        else:
+                            if request.get("kind") == "status" and reply.get("ok") is True:
+                                reply["completed_connection"] = self._completion_watermark()
+                            self._send_authenticated(connection, reply, client_nonce, server_nonce)
                 except socket.timeout:
                     if authenticated:
                         self._send_authenticated(connection, {
@@ -613,77 +820,217 @@ class _BridgeServer(threading.Thread):
                         }, client_nonce, server_nonce)
                 except Exception as error:
                     self._send(connection, {"ok": False, "error": str(error)})
-        self.server.close()
-        try:
-            self.descriptor.unlink(missing_ok=True)
-        except OSError:
-            pass
+        finally:
+            self._note_connection_finished(connection_ordinal)
+            with self._connections_lock:
+                self._connections.discard(connection)
+
+    def _note_connection_finished(self, ordinal: int) -> None:
+        """Record one accepted connection as terminated.
+
+        The watermark is the longest prefix 1..W that has terminated. A later
+        status cannot report an earlier connection that is still reading or
+        still waiting to run.
+        """
+        with self._connections_lock:
+            if ordinal <= self._finished_through or ordinal in self._finished_ordinals:
+                return
+            self._finished_ordinals.add(ordinal)
+            while self._finished_through + 1 in self._finished_ordinals:
+                self._finished_through += 1
+                self._finished_ordinals.discard(self._finished_through)
+
+    def _completion_watermark(self) -> str:
+        with self._connections_lock:
+            return f"{self._finished_through:016x}"
 
     @staticmethod
-    def _receive_frame(connection: socket.socket) -> dict[str, Any]:
+    def _receive_frame(
+        connection: socket.socket,
+        timeout_seconds: float = AUTH_FRAME_TIMEOUT_SECONDS,
+    ) -> dict[str, Any]:
         data = bytearray()
-        deadline = time.monotonic() + AUTH_FRAME_TIMEOUT_SECONDS
-        while b"\n" not in data:
-            if len(data) >= MAX_FRAME_BYTES:
-                raise ValueError("frame-limit-exceeded")
+        deadline = time.monotonic() + timeout_seconds
+        newline = -1
+        while newline < 0:
+            if len(data) > MAX_FRAME_BYTES:
+                raise _frame_limit_exceeded()
             remaining = deadline - time.monotonic()
             if remaining <= 0:
                 raise socket.timeout("authentication frame timed out")
             connection.settimeout(remaining)
+            # MAX_FRAME_BYTES limits the JSON payload; the terminating LF is
+            # framing and may be the one byte beyond that payload budget.
             part = connection.recv(min(65536, MAX_FRAME_BYTES + 1 - len(data)))
             if not part:
                 raise ValueError("incomplete authentication frame")
+            scanned = len(data)
             data.extend(part)
-        line, remainder = bytes(data).split(b"\n", 1)
-        if len(line) > MAX_FRAME_BYTES or remainder:
+            # The delimiter is one byte, so it cannot straddle a chunk.
+            # Search only the bytes just received.
+            newline = data.find(b"\n", scanned)
+        if newline > MAX_FRAME_BYTES:
+            raise _frame_limit_exceeded()
+        if newline != len(data) - 1:
             raise ValueError("invalid framed request")
-        value = json.loads(line)
+        value = json.loads(bytes(data[:newline]))
         return value if isinstance(value, dict) else {}
 
     def _send_authenticated(self, connection: socket.socket, value: dict[str, Any], client_nonce: str, server_nonce: str) -> None:
-        encoded = self._encode_authenticated_response(value, client_nonce, server_nonce)
-        if encoded is None:
-            value = self._truncated_response(value)
-            encoded = self._encode_authenticated_response(value, client_nonce, server_nonce)
-        if encoded is None:
-            # This fixed-size fallback is checked through the same signed framing path.
-            value = {
-                "ok": False,
-                "error": "response-limit-exceeded",
-                "execution_state": value.get("execution_state", "unknown"),
-                "output": None,
-                "output_truncated": True,
-                "message": "The result exceeded the bridge frame limit; command completion state is preserved.",
-            }
-            encoded = self._encode_authenticated_response(value, client_nonce, server_nonce)
-        if encoded is None:
+        if self._deliver_authenticated(connection, value, client_nonce, server_nonce):
             return
-        try:
-            connection.sendall(encoded)
-        except OSError:
-            pass
+        # Keep the recovery payload. Include the observed path when it
+        # fits. An explicit null path means unknown, so the client does
+        # not pair this generation with the descriptor's older path.
+        bounded = self._bounded_status_response(value)
+        if bounded is not None:
+            observed_path = value.get("current_file")
+            if isinstance(observed_path, str) and observed_path:
+                with_path = dict(bounded)
+                with_path["current_file"] = observed_path
+                with_path.pop("message", None)
+                if self._deliver_authenticated(
+                    connection, with_path, client_nonce, server_nonce
+                ):
+                    return
+            if self._deliver_authenticated(
+                connection, bounded, client_nonce, server_nonce
+            ):
+                return
+        command = self._bounded_command_response(value)
+        if command is not None:
+            observed_path = value.get("current_file")
+            if isinstance(observed_path, str) and observed_path:
+                with_path = dict(command)
+                with_path["current_file"] = observed_path
+                with_path.pop("message", None)
+                if self._deliver_authenticated(
+                    connection, with_path, client_nonce, server_nonce
+                ):
+                    return
+            if self._deliver_authenticated(
+                connection, command, client_nonce, server_nonce
+            ):
+                return
+        value = self._truncated_response(value)
+        if self._deliver_authenticated(connection, value, client_nonce, server_nonce):
+            return
+        # This fixed-size fallback is checked through the same signed framing path.
+        self._deliver_authenticated(connection, {
+            "ok": False,
+            "error": "response-limit-exceeded",
+            "execution_state": value.get("execution_state", "unknown"),
+            "output": None,
+            "output_truncated": True,
+            "message": "The result exceeded the bridge frame limit; command completion state is preserved.",
+        }, client_nonce, server_nonce)
 
-    def _encode_authenticated_response(self, value: dict[str, Any], client_nonce: str, server_nonce: str) -> bytes | None:
+    def _deliver_authenticated(
+        self, connection: socket.socket, value: dict[str, Any], client_nonce: str, server_nonce: str
+    ) -> bool:
+        proof = self._authenticated_proof(value, client_nonce, server_nonce)
+        if proof is None:
+            return False
+        try:
+            self._write_authenticated_frame(connection, value, proof)
+        except OSError:
+            # A disconnected caller must not fall through to another payload.
+            pass
+        return True
+
+    def _authenticated_proof(
+        self, value: dict[str, Any], client_nonce: str, server_nonce: str
+    ) -> str | None:
         if not _within_response_limit(value):
             return None
-        prefix = b'{"response":"'
-        suffix = b'","proof":"' + (b"0" * 64) + b'"}\n'
-        frame_size = len(prefix) + len(suffix)
-        response_parts: list[bytes] = []
-        escaped_parts: list[bytes] = []
-        for fragment in json.JSONEncoder(ensure_ascii=False, separators=(",", ":")).iterencode(value):
-            raw = fragment.encode("utf-8")
-            escaped = json.dumps(fragment, ensure_ascii=False)[1:-1].encode("utf-8")
-            frame_size += len(escaped)
+        # The payload budget excludes the terminating LF. Fragments are
+        # measured and signed here, then discarded. The write pass encodes
+        # them again so a maximum-sized reply is not held in several copies.
+        frame_size = _AUTHENTICATED_FRAME_OVERHEAD
+        proof = hmac.new(self.token.encode(), digestmod=hashlib.sha256)
+        proof.update(
+            f"rea-cutter-response-v2:{self.session_id}:{client_nonce}:{server_nonce}:".encode()
+        )
+        for fragment in _response_fragments(value):
+            frame_size += len(_escape_json_fragment(fragment))
             if frame_size > MAX_FRAME_BYTES:
                 return None
-            response_parts.append(raw)
-            escaped_parts.append(escaped)
-        proof = hmac.new(self.token.encode(), digestmod=hashlib.sha256)
-        proof.update(f"rea-cutter-response-v2:{self.session_id}:{client_nonce}:{server_nonce}:".encode())
-        for part in response_parts:
-            proof.update(part)
-        return prefix + b"".join(escaped_parts) + b'","proof":"' + proof.hexdigest().encode() + b'"}\n'
+            proof.update(fragment.encode("utf-8"))
+        return proof.hexdigest()
+
+    def _encode_authenticated_response(
+        self, value: dict[str, Any], client_nonce: str, server_nonce: str
+    ) -> bytes | None:
+        """Return one exact frame. The socket path writes the same pieces."""
+        proof = self._authenticated_proof(value, client_nonce, server_nonce)
+        if proof is None:
+            return None
+        return b"".join(self._iter_authenticated_frame(value, proof))
+
+    def _iter_authenticated_frame(self, value: dict[str, Any], proof: str):
+        yield _AUTHENTICATED_PREFIX
+        for fragment in _response_fragments(value):
+            yield _escape_json_fragment(fragment)
+        yield _AUTHENTICATED_PROOF_MARK + proof.encode("ascii") + b'"}\n'
+
+    def _write_authenticated_frame(
+        self, connection: socket.socket, value: dict[str, Any], proof: str
+    ) -> None:
+        pending = bytearray()
+        for chunk in self._iter_authenticated_frame(value, proof):
+            pending.extend(chunk)
+            if len(pending) >= _AUTHENTICATED_WRITE_BUFFER_BYTES:
+                connection.sendall(pending)
+                pending.clear()
+        if pending:
+            connection.sendall(pending)
+
+    @staticmethod
+    def _bounded_status_response(value: dict[str, Any]) -> dict[str, Any] | None:
+        if "completed_commands" not in value and "completed_connection" not in value:
+            return None
+        bounded: dict[str, Any] = {
+            "ok": value.get("ok") is True,
+            "session_id": value.get("session_id"),
+            "document_generation": value.get("document_generation"),
+            "identity_status": value.get("identity_status", "partial"),
+            "completed_commands": value.get("completed_commands"),
+            "current_file": None,
+            "message": (
+                "current_file is unknown because the status frame exceeded "
+                "REA_BRIDGE_MAX_REQUEST_BYTES. Do not reuse an older descriptor "
+                "path with this document_generation."
+            ),
+        }
+        if "completed_connection" in value:
+            bounded["completed_connection"] = value["completed_connection"]
+        return bounded
+
+    @staticmethod
+    def _bounded_command_response(value: dict[str, Any]) -> dict[str, Any] | None:
+        if "execution_state" not in value and "document_generation" not in value:
+            return None
+        bounded: dict[str, Any] = {
+            "ok": value.get("ok") is True,
+            "output": None,
+            "execution_state": value.get("execution_state", "unknown"),
+            "output_truncated": True,
+            "current_file": None,
+            "message": (
+                "current_file is unknown because the result frame exceeded "
+                "REA_BRIDGE_MAX_REQUEST_BYTES. The command completion state and "
+                "document_generation are preserved. Do not reuse an older "
+                "descriptor path with this document_generation."
+            ),
+        }
+        generation = value.get("document_generation")
+        if isinstance(generation, int) and not isinstance(generation, bool):
+            bounded["document_generation"] = generation
+        if value.get("identity_status") == "partial":
+            bounded["identity_status"] = "partial"
+        if value.get("ok") is not True and isinstance(value.get("error"), str):
+            bounded["error"] = value["error"]
+        return bounded
 
     @staticmethod
     def _truncated_response(value: dict[str, Any]) -> dict[str, Any]:
@@ -741,6 +1088,17 @@ class _BridgeServer(threading.Thread):
             self.server.close()
         except OSError:
             pass
+        with self._connections_lock:
+            connections = list(self._connections)
+        for connection in connections:
+            try:
+                connection.shutdown(socket.SHUT_RDWR)
+            except OSError:
+                pass
+            try:
+                connection.close()
+            except OSError:
+                pass
 
 
 class ReaCutterPlugin(cutter.CutterPlugin):
@@ -774,6 +1132,128 @@ def create_cutter_plugin() -> ReaCutterPlugin:
     return ReaCutterPlugin()
 
 
+def _json_object_key(key: Any) -> str:
+    """Match CPython's JSON object-key text before quoting."""
+    if isinstance(key, str):
+        return key
+    if key is True:
+        return "true"
+    if key is False:
+        return "false"
+    if key is None:
+        return "null"
+    if isinstance(key, int):
+        return str(key)
+    if isinstance(key, float):
+        return json.dumps(key, separators=(",", ":"))
+    raise TypeError("keys must be str, int, float, bool or None")
+
+
+def _response_fragments(value: Any):
+    """Yield compact JSON pieces whose concatenation is the encoded value.
+
+    String contents are sliced before encoding so a large output never becomes
+    one frame-sized fragment. Each piece is at most ``_JSON_CHUNK_CHARS``
+    characters, or the six-byte ``\\u00XX`` expansion of one such slice.
+    """
+    if isinstance(value, str):
+        yield '"'
+        for start in range(0, len(value), _JSON_CHUNK_CHARS):
+            yield json.dumps(
+                value[start:start + _JSON_CHUNK_CHARS], ensure_ascii=False
+            )[1:-1]
+        yield '"'
+        return
+    if isinstance(value, dict):
+        yield "{"
+        first = True
+        for key, item in value.items():
+            if not first:
+                yield ","
+            first = False
+            yield from _response_fragments(_json_object_key(key))
+            yield ":"
+            yield from _response_fragments(item)
+        yield "}"
+        return
+    if isinstance(value, (list, tuple)):
+        yield from _list_fragments(value)
+        return
+    token = _scalar_token(value)
+    if token is None:
+        raise TypeError("response value is not JSON")
+    yield token
+
+
+def _scalar_token(value: Any) -> str | None:
+    """Compact JSON for one scalar. None means the value needs structure."""
+    if type(value) is int:
+        return str(value)
+    if value is None:
+        return "null"
+    if value is True:
+        return "true"
+    if value is False:
+        return "false"
+    if isinstance(value, float):
+        return json.dumps(value, separators=(",", ":"))
+    return None
+
+
+def _list_fragments(value: list[Any] | tuple[Any, ...]):
+    """Encode a list, batching scalars into bounded pieces.
+
+    One ``json.dumps`` per number makes a dense array miss the client
+    deadline. A batch is at most ``_JSON_CHUNK_CHARS`` of compact JSON.
+    """
+    yield "["
+    first = True
+    batch: list[str] = []
+    batch_len = 0
+
+    def flush_batch() -> str | None:
+        nonlocal first, batch, batch_len
+        if not batch:
+            return None
+        piece = ",".join(batch)
+        batch = []
+        batch_len = 0
+        if first:
+            first = False
+            return piece
+        return "," + piece
+
+    for item in value:
+        token = _scalar_token(item)
+        if token is None:
+            emitted = flush_batch()
+            if emitted is not None:
+                yield emitted
+            if first:
+                first = False
+            else:
+                yield ","
+            yield from _response_fragments(item)
+            continue
+        extra = len(token) if not batch else len(token) + 1
+        if batch and batch_len + extra > _JSON_CHUNK_CHARS:
+            emitted = flush_batch()
+            if emitted is not None:
+                yield emitted
+            extra = len(token)
+        batch.append(token)
+        batch_len += extra
+    emitted = flush_batch()
+    if emitted is not None:
+        yield emitted
+    yield "]"
+
+
+def _escape_json_fragment(fragment: str) -> bytes:
+    """Escape one raw JSON piece for the outer response string."""
+    return json.dumps(fragment, ensure_ascii=False)[1:-1].encode("utf-8")
+
+
 def _within_response_limit(value: Any) -> bool:
     pending = [value]
     seen: set[int] = set()
@@ -782,20 +1262,37 @@ def _within_response_limit(value: Any) -> bool:
         current = pending.pop()
         if isinstance(current, str):
             size += 2
+            if len(current) > MAX_FRAME_BYTES - size:
+                return False
             if current.isascii():
                 size += len(current)
-                size += current.count('"') + current.count("\\")
-                size += sum(5 for character in current if ord(character) < 0x20)
             else:
-                size += 4 * len(current)
-                size += current.count('"') + current.count("\\")
-                size += sum(5 for character in current if ord(character) < 0x20)
+                # Count UTF-8 bytes without allocating a copy of the whole output.
+                try:
+                    size += sum(
+                        len(current[start:start + 65536].encode("utf-8"))
+                        for start in range(0, len(current), 65536)
+                    )
+                except UnicodeEncodeError:
+                    # Unencodable output also needs the signed omission fallback.
+                    return False
+            size += current.count('"') + current.count("\\")
+            # A short escape is two JSON bytes and was already counted as one.
+            # Other controls expand to a six-byte \\u00XX sequence.
+            size += sum(
+                1 if ord(character) in (0x08, 0x09, 0x0A, 0x0C, 0x0D) else 5
+                for character in current
+                if ord(character) < 0x20
+            )
         elif isinstance(current, dict):
             identity = id(current)
             if identity in seen:
                 return False
             seen.add(identity)
-            size += 2 + 4 * len(current)
+            count = len(current)
+            # Braces, one colon per entry, and commas between entries.
+            # Keys and values are counted when they are visited.
+            size += 2 + count + max(0, count - 1)
             if size > MAX_FRAME_BYTES:
                 return False
             pending.extend(current.keys())
@@ -805,12 +1302,29 @@ def _within_response_limit(value: Any) -> bool:
             if identity in seen:
                 return False
             seen.add(identity)
-            size += 2 + 2 * len(current)
+            count = len(current)
+            # Brackets and commas between items. Each item is counted below.
+            size += 2 + max(0, count - 1)
             if size > MAX_FRAME_BYTES:
                 return False
             pending.extend(current)
+        elif current is None:
+            size += 4
+        elif current is True:
+            size += 4
+        elif current is False:
+            size += 5
+        elif isinstance(current, int):
+            size += len(str(current))
+        elif isinstance(current, float):
+            size += len(json.dumps(current, separators=(",", ":")))
         else:
-            size += 16
+            try:
+                size += len(
+                    json.dumps(current, ensure_ascii=False, separators=(",", ":"))
+                )
+            except (TypeError, ValueError):
+                return False
         if size > MAX_FRAME_BYTES:
             return False
     return True

bridge/ghidra/ReaGhidraAnalysisOptions.java

@@ -1,38 +1,91 @@
+import java.util.HashSet;
 import java.util.Map;
+import java.util.Set;
+
+import com.google.gson.JsonArray;
+import com.google.gson.JsonObject;
 
 import ghidra.app.plugin.core.analysis.AutoAnalysisManager;
+import ghidra.app.services.Analyzer;
 import ghidra.app.util.headless.HeadlessScript;
+import ghidra.program.model.listing.Program;
+import ghidra.util.classfinder.ClassSearcher;
 
-/** Disable the named auto-analyzers before default auto-analysis. */
+/**
+ * Disable the named auto-analyzers before default auto-analysis.
+ *
+ * REA_GHIDRA_DISABLED_ANALYZERS is process-wide, but Ghidra registers only the
+ * installed analyzers whose canAnalyze(program) accepts this program. A name an
+ * installed analyzer provides without being registered here is not applicable
+ * and is recorded instead of rejected; a name no installed analyzer provides
+ * aborts the import. The outcome is stored as a program property that the
+ * bridge reports during its handshake.
+ */
 public final class ReaGhidraAnalysisOptions extends HeadlessScript {
+    static final String REPORT_OPTION = "REA disabled analyzers";
+    private static final String SETTING = "REA_GHIDRA_DISABLED_ANALYZERS";
+
     @Override
     public void run() throws Exception {
         String[] arguments = getScriptArgs();
         if (arguments.length != 1 || arguments[0].isBlank()) {
             throw new IllegalArgumentException(
-                "ReaGhidraAnalysisOptions requires one comma-separated analyzer-name argument");
+                "ReaGhidraAnalysisOptions requires one comma-separated " + SETTING + " argument");
         }
         String[] names = arguments[0].split(",");
+        boolean[] registered = new boolean[names.length];
         AutoAnalysisManager analysis = AutoAnalysisManager.getAnalysisManager(currentProgram);
         Map<String, String> options = getCurrentAnalysisOptionsAndValues(currentProgram);
-        for (String requestedName : names) {
+        Set<String> installed = null;
+        for (int index = 0; index < names.length; index++) {
             monitor.checkCancelled();
-            String name = requestedName.trim();
-            if (analysis.getAnalyzer(name) == null || !options.containsKey(name)) {
-                throw new IllegalArgumentException("Unknown Ghidra analyzer for this program: " + name);
+            String name = names[index].trim();
+            names[index] = name;
+            if (analysis.getAnalyzer(name) != null) {
+                if (!options.containsKey(name)) {
+                    throw new IllegalStateException(SETTING + " names " + name +
+                        ", which Ghidra registered for this program without an enablement option");
+                }
+                registered[index] = true;
+                continue;
+            }
+            if (installed == null) installed = installedAnalyzerNames();
+            if (!installed.contains(name)) {
+                throw new IllegalArgumentException(SETTING + " names " + name +
+                    ", which is not an analyzer in this Ghidra installation");
             }
         }
-        for (String requestedName : names) {
+        for (int index = 0; index < names.length; index++) {
             monitor.checkCancelled();
-            setAnalysisOption(currentProgram, requestedName.trim(), "false");
+            if (registered[index]) setAnalysisOption(currentProgram, names[index], "false");
         }
         options = getCurrentAnalysisOptionsAndValues(currentProgram);
-        for (String requestedName : names) {
+        JsonArray analyzers = new JsonArray();
+        for (int index = 0; index < names.length; index++) {
             monitor.checkCancelled();
-            String name = requestedName.trim();
-            if (!"false".equals(options.get(name))) {
-                throw new IllegalStateException("Ghidra analyzer remained enabled: " + name);
+            String name = names[index];
+            if (registered[index] && !"false".equals(options.get(name))) {
+                throw new IllegalStateException(SETTING + " names " + name +
+                    ", which remained enabled for this program");
             }
+            JsonObject analyzer = new JsonObject();
+            analyzer.addProperty("name", name);
+            analyzer.addProperty("status", registered[index] ? "disabled" : "not_applicable");
+            analyzers.add(analyzer);
         }
+        JsonObject report = new JsonObject();
+        report.addProperty("format", "rea-ghidra-disabled-analyzers-v1");
+        report.add("analyzers", analyzers);
+        currentProgram.getOptions(Program.PROGRAM_INFO).setString(REPORT_OPTION, report.toString());
+        println("REA disabled analyzers: " + report);
+    }
+
+    /** The same analyzer universe AutoAnalysisManager filters by canAnalyze(program). */
+    private static Set<String> installedAnalyzerNames() {
+        Set<String> names = new HashSet<>();
+        for (Analyzer analyzer : ClassSearcher.getInstances(Analyzer.class)) {
+            names.add(analyzer.getName());
+        }
+        return names;
     }
 }

bridge/ghidra/ReaGhidraBridge.java

@@ -140,6 +140,7 @@ public final class ReaGhidraBridge extends HeadlessScript {
         "params"
     );
     private static final String[] CAPABILITIES = {
+        "read_binary_inventory_summary",
         "annotate_native_function",
         "set_address_name",
         "set_addresses_names",
@@ -427,6 +428,7 @@ public final class ReaGhidraBridge extends HeadlessScript {
             );
         }
         return switch (request.method) {
+            case "read_binary_inventory_summary" -> binaryInventorySummary(request.params);
             case "inspect_native_load_image" -> inspectNativeLoadImage(request.params);
             case "read_bytes" -> readMemoryBytes(request.params);
             case "search_bytes" -> searchMemoryBytes(request.params, descriptor);
@@ -456,6 +458,7 @@ public final class ReaGhidraBridge extends HeadlessScript {
                 FunctionAnalysisObservation observation = analyzeFunctionObservation(request.params);
                 JsonObject result = functionObservation(observation.dossier(), observation.limitations());
                 result.add("relationship_coverage", observation.relationshipCoverage());
+                if (observation.decompilerFailed()) result.addProperty("cache_policy", "live");
                 yield result;
             }
             case "set_address_name", "set_addresses_names" -> {
@@ -520,6 +523,11 @@ public final class ReaGhidraBridge extends HeadlessScript {
         String seeds = currentProgram.getOptions(ghidra.program.model.listing.Program.PROGRAM_INFO)
             .getString("REA analysis seeds", null);
         if (seeds != null) result.add("analysis_seeds", JsonParser.parseString(seeds));
+        // Written by ReaGhidraAnalysisOptions before auto-analysis; REA checks it
+        // against the committed profile.
+        String disabledAnalyzers = currentProgram.getOptions(ghidra.program.model.listing.Program.PROGRAM_INFO)
+            .getString("REA disabled analyzers", null);
+        if (disabledAnalyzers != null) result.add("disabled_analyzers", JsonParser.parseString(disabledAnalyzers));
         // Written by ReaGhidraMipsSoftFloat after analysis; REA requires it
         // exactly when the committed profile declares soft-float storage.
         String softFloat = currentProgram.getOptions(ghidra.program.model.listing.Program.PROGRAM_INFO)
@@ -1260,9 +1268,11 @@ public final class ReaGhidraBridge extends HeadlessScript {
         result.add("procedure", function == null ? JsonNull.INSTANCE : GSON.toJsonTree(canonicalAddress(function.getEntryPoint())));
         result.addProperty("architecture", currentProgram.getLanguageID().getIdAsString());
         result.addProperty("mode", currentProgram.getLanguage().getLanguageDescription().getVariant());
-        if (currentProgram.getLanguageID().getIdAsString().startsWith("MIPS:")) {
+        if (currentProgram.getLanguageID().getIdAsString().startsWith("MIPS:") ||
+            currentProgram.getLanguageID().getIdAsString().startsWith("ARM:")) {
             JsonArray decoderContext = new JsonArray();
-            var modeRegister = currentProgram.getRegister("ISA_MODE");
+            String regName = currentProgram.getLanguageID().getIdAsString().startsWith("MIPS:") ? "ISA_MODE" : "TMode";
+            var modeRegister = currentProgram.getRegister(regName);
             if (modeRegister != null) {
                 var modeValue = currentProgram.getProgramContext().getValue(modeRegister, address, false);
                 JsonObject context = new JsonObject();
@@ -1538,12 +1548,23 @@ public final class ReaGhidraBridge extends HeadlessScript {
         requireKeys(params, Set.of("procedure"));
         Function function = resolveProcedure(requireString(params, "procedure"));
         InstructionScan scan = scanInstructions(function, Integer.MAX_VALUE);
-        DecompileResults decompilation = decompile(function);
+        DecompileResults decompilation;
+        String decompilerFailure = null;
+        try {
+            decompilation = decompile(function);
+        } catch (RequestFailure failure) {
+            // The listing, reference and CFG facts below do not depend on the
+            // native decompiler. Cancellation still ends the request.
+            if (!failure.code.equals("decompile_failed")) throw failure;
+            decompilation = null;
+            decompilerFailure = failure.getMessage().strip();
+        }
         String pseudocode = decompiledText(decompilation);
         JsonObject availability = new JsonObject();
         availability.add("pseudocode", facetAvailability(
             pseudocode != null,
-            function.isExternal() ? "Ghidra does not decompile external functions."
+            decompilerFailure != null ? decompilerFailure
+                : function.isExternal() ? "Ghidra does not decompile external functions."
                 : function.getBody().isEmpty() ? "Ghidra has no function body to decompile."
                 : "Ghidra completed decompilation without returning pseudocode."
         ));
@@ -1592,9 +1613,16 @@ public final class ReaGhidraBridge extends HeadlessScript {
         result.add("basic_blocks", blocks);
         result.add(
             "native_api",
-            nativeApiBoundary(function, decompilation)
+            decompilerFailure == null
+                ? nativeApiBoundary(function, decompilation)
+                : unavailableNativeApiBoundary(decompilerFailure)
+        );
+        result.add(
+            "native_value_flow",
+            decompilerFailure == null
+                ? nativeValueFlow(decompilation)
+                : unavailableNativeValueFlow(decompilerFailure)
         );
-        result.add("native_value_flow", nativeValueFlow(decompilation));
         // Successful full queries distinguish collected empty sets from missing facts.
         // Availability does not claim that Ghidra resolved every runtime relationship.
         for (String facet : List.of("comments", "callers", "callees", "references", "strings_names", "cfg")) {
@@ -1641,6 +1669,12 @@ public final class ReaGhidraBridge extends HeadlessScript {
             !decompilation.getErrorMessage().isBlank()) {
             limitations.add("Ghidra decompiler diagnostic: " + decompilation.getErrorMessage());
         }
+        if (decompilerFailure != null) {
+            limitations.add(
+                decompilerFailure + (decompilerFailure.endsWith(".") ? " " : ". ") +
+                "Pseudocode, the native API boundary and p-code value flow are unavailable for this procedure; assembly, comments, callers, callees, references, strings, names and basic blocks come from the Ghidra program listing and were still collected."
+            );
+        }
         result.add("limitations", limitations);
         monitor.checkCancelled();
         int missingCallers = 0;
@@ -1649,25 +1683,16 @@ public final class ReaGhidraBridge extends HeadlessScript {
             if (unrepresentedCaller(reference, callerFunctions)) missingCallers++;
         }
         return new FunctionAnalysisObservation(result, calls.limitations(),
-            callRelationshipCoverage(calls, missingCallers));
+            callRelationshipCoverage(calls, missingCallers), decompilerFailure != null);
     }
 
     private JsonObject nativeValueFlow(DecompileResults decompilation) throws Exception {
         HighFunction highFunction =
             decompilation == null ? null : decompilation.getHighFunction();
         if (highFunction == null) {
-            JsonObject unavailable = new JsonObject();
-            unavailable.addProperty("available", false);
-            unavailable.addProperty(
-                "reason",
+            return unavailableNativeValueFlow(
                 "Ghidra has no HighFunction p-code model for this procedure"
             );
-            JsonArray limitations = new JsonArray();
-            limitations.add(
-                "No p-code operations or def-use relationships were recovered for this procedure."
-            );
-            unavailable.add("limitations", limitations);
-            return unavailable;
         }
 
         List<PcodeOpAST> selected = new ArrayList<>();
@@ -1974,6 +1999,18 @@ public final class ReaGhidraBridge extends HeadlessScript {
         return result;
     }
 
+    private JsonObject binaryInventorySummary(JsonObject params) throws Exception {
+        requireKeys(params, Set.of());
+        JsonObject result = new JsonObject();
+        result.addProperty("document", currentProgram.getName());
+        result.add("segments", segments());
+        // Count the owning inventories, preserving their exact discovery rules
+        // without serializing function identities or string values for counts.
+        result.addProperty("procedure_count", procedures().size());
+        result.addProperty("string_count", strings().size());
+        return result;
+    }
+
     private JsonArray listNames(JsonObject params) throws Exception {
         requireKeys(params, Set.of("document", "address"));
         requireDocument(params);
@@ -2000,6 +2037,10 @@ public final class ReaGhidraBridge extends HeadlessScript {
     private JsonArray listSegments(JsonObject params) {
         requireKeys(params, Set.of("document"));
         requireDocument(params);
+        return segments();
+    }
+
+    private JsonArray segments() {
         JsonArray result = new JsonArray();
         String imageBase = canonicalAddress(currentProgram.getImageBase());
         MemoryBlock[] blocks = currentProgram.getMemory().getBlocks();
@@ -2339,6 +2380,18 @@ public final class ReaGhidraBridge extends HeadlessScript {
         return result;
     }
 
+    private static JsonObject unavailableNativeValueFlow(String reason) {
+        JsonObject unavailable = new JsonObject();
+        unavailable.addProperty("available", false);
+        unavailable.addProperty("reason", reason);
+        JsonArray limitations = new JsonArray();
+        limitations.add(
+            "No p-code operations or def-use relationships were recovered for this procedure."
+        );
+        unavailable.add("limitations", limitations);
+        return unavailable;
+    }
+
     private static JsonObject unavailableNativeApiBoundary(String reason) {
         JsonObject unavailable = new JsonObject();
         unavailable.addProperty("available", false);
@@ -4857,7 +4910,7 @@ public final class ReaGhidraBridge extends HeadlessScript {
     private record InventoryItem(Address address, String value, JsonObject facts) {}
     private record AddressNameRequest(String inputAddress, Address address, String name) {}
     private record FunctionEntry(Function function, InventoryItem item) {}
-    private record FunctionAnalysisObservation(JsonObject dossier, List<String> limitations, JsonObject relationshipCoverage) {}
+    private record FunctionAnalysisObservation(JsonObject dossier, List<String> limitations, JsonObject relationshipCoverage, boolean decompilerFailed) {}
     private record CallBoundaryObservation(List<String> limitations, Integer unrepresentedCalleeCalls, JsonArray unresolvedCalls) {}
     private record SwiftStringLiteralDecode(JsonArray literals, JsonArray omissions) {}
     private record SwiftLiteralLayout(

bridge/hopper_bridge.py

@@ -8,12 +8,100 @@ API access on this thread: moving dispatch to a worker can deadlock Hopper.
 import json
 import hmac
 import os
+import re
 import socket
 from typing import Any, Optional, Protocol, Sequence
 
 BAD_ADDRESSES = (-1, 0xFFFFFFFFFFFFFFFF, None)
 # The native label setter truncates NSString names at this UTF-16 extent.
 HOPPER_SYMBOL_NAME_UTF16_UNITS = 1024
+# Request JSON payload budget in bytes, excluding the trailing LF.
+# REA_BRIDGE_MAX_REQUEST_BYTES selects it. Unset stays 16777216.
+# 7464 is the authenticated Cutter frame of the largest bounded symbol name.
+# Cutter's 64-receipt recovery status is smaller with a null path. A path
+# that does not fit is null, not an older descriptor path.
+# 536870887 is Node's 536870888-character string ceiling minus the framing LF.
+# The ceiling is the encoded payload; JSON expansion is measured before it.
+_BRIDGE_MAX_REQUEST_BYTES_SETTING = "REA_BRIDGE_MAX_REQUEST_BYTES"
+DEFAULT_BRIDGE_MAX_REQUEST_BYTES = 16 * 1024 * 1024
+MINIMUM_BRIDGE_MAX_REQUEST_BYTES = 7464
+MAXIMUM_BRIDGE_MAX_REQUEST_BYTES = 536870887
+
+
+def _bridge_max_request_bytes_constraint():
+    return (
+        "REA_BRIDGE_MAX_REQUEST_BYTES must be a positive decimal integer "
+        "from %d to %d bytes; the default is %d bytes (16 MiB). "
+        "A value below %d bytes cannot carry the largest admitted "
+        "symbol-name command, and a larger value is refused so the encoded "
+        "payload plus its framing LF stays one string under Node's "
+        "single-string ceiling."
+        % (
+            MINIMUM_BRIDGE_MAX_REQUEST_BYTES,
+            MAXIMUM_BRIDGE_MAX_REQUEST_BYTES,
+            DEFAULT_BRIDGE_MAX_REQUEST_BYTES,
+            MINIMUM_BRIDGE_MAX_REQUEST_BYTES,
+        )
+    )
+
+
+def _parse_bridge_max_request_bytes(raw):
+    """Refuse a non-decimal or out-of-range ceiling. Never fall back."""
+    constraint = _bridge_max_request_bytes_constraint()
+    if isinstance(raw, bool):
+        raise RuntimeError(constraint)
+    if isinstance(raw, int):
+        value = raw
+    elif isinstance(raw, str):
+        # ASCII digits only. str.isdigit() accepts numerals such as U+0668
+        # that the TypeScript setting rejects.
+        if re.fullmatch(r"[1-9][0-9]*", raw) is None:
+            raise RuntimeError(constraint)
+        value = int(raw)
+    else:
+        raise RuntimeError(constraint)
+    if (
+        value < MINIMUM_BRIDGE_MAX_REQUEST_BYTES
+        or value > MAXIMUM_BRIDGE_MAX_REQUEST_BYTES
+    ):
+        raise RuntimeError(constraint)
+    return value
+
+
+def _resolve_bridge_max_request_bytes():
+    """Prefer the bootstrap injection, then the process environment, then 16 MiB."""
+    injected = globals().get(_BRIDGE_MAX_REQUEST_BYTES_SETTING)
+    if injected is not None:
+        return _parse_bridge_max_request_bytes(injected), "setting"
+    raw = os.environ.get(_BRIDGE_MAX_REQUEST_BYTES_SETTING)
+    if raw is None:
+        return DEFAULT_BRIDGE_MAX_REQUEST_BYTES, "default"
+    return _parse_bridge_max_request_bytes(raw), "setting"
+
+
+MAX_REQUEST_BYTES, BRIDGE_MAX_REQUEST_BYTES_SOURCE = (
+    _resolve_bridge_max_request_bytes()
+)
+
+
+def _bridge_max_request_bytes_recovery():
+    return (
+        "The effective REA_BRIDGE_MAX_REQUEST_BYTES value is %d bytes "
+        "(source %s; default %d, floor %d, maximum %d). "
+        "Set REA_BRIDGE_MAX_REQUEST_BYTES in the environment that starts REA "
+        "to a decimal integer from %d to %d, then restart REA. "
+        "Restart Hopper or Cutter from that same environment so the bridge "
+        "uses this ceiling."
+        % (
+            MAX_REQUEST_BYTES,
+            BRIDGE_MAX_REQUEST_BYTES_SOURCE,
+            DEFAULT_BRIDGE_MAX_REQUEST_BYTES,
+            MINIMUM_BRIDGE_MAX_REQUEST_BYTES,
+            MAXIMUM_BRIDGE_MAX_REQUEST_BYTES,
+            MINIMUM_BRIDGE_MAX_REQUEST_BYTES,
+            MAXIMUM_BRIDGE_MAX_REQUEST_BYTES,
+        )
+    )
 _rea_session_document = None
 _search_inventory_cache = {}
 _pseudocode_cache = {}
@@ -236,6 +324,15 @@ def _name_matches(document, address, name):
     return observed == name or (not name and observed is None)
 
 
+def _bookmark_name(document, address):
+    getter = getattr(document, "getBookmarkName", None)
+    if not callable(getter):
+        raise CapabilityUnavailableError(
+            "Hopper's public Python API does not expose getBookmarkName"
+        )
+    return getter(address)
+
+
 def _address(document, value=None):
     """Keep explicit hexadecimal coordinates distinct from symbol names."""
     if value is None:
@@ -431,10 +528,6 @@ def _procedure_map(document):
     return result
 
 
-def _strings(document):
-    return {address: record["value"] for address, record in _string_inventory(document).items()}
-
-
 def _string_record(segment, address, display):
     """Read one native typed object, retaining its shortened display separately."""
     length = segment.getObjectLength(address)
@@ -521,7 +614,7 @@ def _search_inventory(document, kind):
         if kind == "procedure":
             values = _procedure_map(document)
         elif kind == "string":
-            values = _strings(document)
+            values = _string_inventory(document)
         elif kind == "name":
             values = {_hex(address): value for address, value in _name_map(document).items()}
         else:
@@ -548,23 +641,17 @@ def _search_results(document, kind, params):
     else:
         raise CapabilityUnavailableError("Regex matching requires REA's supervised worker; the native bridge only executes literal searches")
 
-    selected = []
-    for item in _search_inventory(document, kind):
-        if not matches(item[1]):
-            continue
-        selected.append(item)
-    if not selected:
-        return []
     if kind == "string":
-        records = _string_inventory(document)
-        return [{"address": address, **records[address]} for address, _ in selected]
-    return [
-            {
-                "address": address,
-                "value": value,
-            }
-            for address, value in selected
+        return [
+            {"address": address, **record}
+            for address, record in _search_inventory(document, kind)
+            if matches(record["value"])
         ]
+    return [
+        {"address": address, "value": value}
+        for address, value in _search_inventory(document, kind)
+        if matches(value)
+    ]
 
 
 def _unavailable(reason):
@@ -741,6 +828,7 @@ def _dispatch(method, params):
             "version": "1.0.0",
             "run_id": REA_RUN_ID,
             "process_id": os.getpid(),
+            "max_request_bytes": MAX_REQUEST_BYTES,
         }
     if method in ("shutdown", "shutdown_document"):
         document = _session_document()
@@ -1004,15 +1092,46 @@ def _dispatch(method, params):
         observed = getter(address)
         return observed == comment or (comment == "" and observed is None)
     if method == "list_bookmarks":
-        return [{"address": _hex(item), "name": document.getBookmarkName(item)} for item in document.getBookmarks()]
+        return [{"address": _hex(item), "name": _bookmark_name(document, item)} for item in document.getBookmarks()]
     if method == "set_bookmark":
         address = _address(document, params.get("address"))
         _segment(document, address)
         name = params.get("name")
         if name is not None:
             _api_text(name, "Bookmark name")
-        document.setBookmarkAtAddress(address, name)
-        return document.hasBookmarkAtAddress(address)
+        has_bookmark = getattr(document, "hasBookmarkAtAddress", None)
+        if not callable(has_bookmark):
+            raise CapabilityUnavailableError(
+                "Hopper's public Python API does not expose hasBookmarkAtAddress"
+            )
+        exists = has_bookmark(address)
+        if name is None and exists:
+            # Omission means preserve the existing bookmark and its name.
+            return True
+        if not callable(getattr(document, "getBookmarkName", None)):
+            raise CapabilityUnavailableError(
+                "Hopper's public Python API does not expose getBookmarkName"
+            )
+        if name == "" and exists:
+            rename = getattr(document, "renameBookmarkAtAddress", None)
+            if not callable(rename):
+                raise CapabilityUnavailableError(
+                    "Hopper's public Python API does not expose renameBookmarkAtAddress"
+                )
+            rename(address, name)
+        else:
+            setter = getattr(document, "setBookmarkAtAddress", None)
+            if not callable(setter):
+                raise CapabilityUnavailableError(
+                    "Hopper's public Python API does not expose setBookmarkAtAddress"
+                )
+            setter(address, name)
+        if not has_bookmark(address):
+            return False
+        observed = _bookmark_name(document, address)
+        if name is None:
+            return observed is None or observed == ""
+        return observed == name or (name == "" and observed is None)
     if method == "unset_bookmark":
         address = _address(document, params.get("address"))
         # Permit removing a legacy orphan bookmark, but do not report success
@@ -1030,9 +1149,23 @@ def _serve_connection(connection):
     should_stop = False
     try:
         while True:
-            line = file.readline()
+            # Bound the read itself. The payload budget excludes LF, so permit
+            # one additional byte for the delimiter while still refusing an
+            # oversized unterminated peer without waiting for EOF.
+            line = file.readline(MAX_REQUEST_BYTES + 1)
             if not line:
                 break
+            if not line.endswith(b"\n"):
+                message = (
+                    "Bridge request exceeds the %d-byte payload limit. %s"
+                    % (MAX_REQUEST_BYTES, _bridge_max_request_bytes_recovery())
+                    if len(line) > MAX_REQUEST_BYTES
+                    else "Incomplete bridge request frame"
+                )
+                _write_message(file, {"id": 0, "error": _safe_diagnostic(
+                    InvalidRequestError(message)
+                )})
+                break
             request_id = None
             authenticated = False
             should_stop = False
@@ -1057,6 +1190,10 @@ def _serve_connection(connection):
                 ):
                     raise PermissionError("Invalid bridge capability")
                 authenticated = True
+                # The connection grace period only bounds unauthenticated idle
+                # clients. Native analysis and a connected owner's idle time
+                # remain unbounded after capability validation.
+                connection.settimeout(None)
                 _write_message(file, {"id": request_id, "event": {
                     "type": "progress",
                     "phase": "hopper_bridge",
@@ -1148,13 +1285,17 @@ def _run():
     server.bind(REA_SOCKET)
     os.chmod(REA_SOCKET, 0o600)
     server.listen(1)
+    # A bridge can start after its caller's startup/cleanup deadline. Bound
+    # the first accept as well as reconnects so an orphan cannot pin Hopper's
+    # dedicated Python thread and prevent later documents from starting REA.
+    server.settimeout(REA_RECONNECT_TIMEOUT_SECONDS)
     try:
         while True:
             try:
                 connection, _ = server.accept()
             except socket.timeout:
                 break
-            connection.settimeout(None)
+            connection.settimeout(REA_RECONNECT_TIMEOUT_SECONDS)
             if _serve_connection(connection):
                 break
             # Match REA's cleanup handshake budget without pinning Hopper's

bridge/native/rea_lldb_tracer.py

@@ -555,6 +555,38 @@ def _failed_launch_description(process):
     return None
 
 
+def _retire_failed_launch(process, config, listener, stdout, stderr):
+    """Reconcile an SBProcess returned alongside a failed Launch error."""
+    if not process.IsValid():
+        stdout.close()
+        stderr.close()
+        return False
+
+    pid = process.GetProcessID()
+    if pid > 0:
+        # Even an already-exited process needs its PID published so the host
+        # can distinguish a completed launch failure from an unknown target.
+        with open(config["pid_path"], "w") as handle:
+            handle.write(str(pid))
+
+    if process.GetState() != lldb.eStateExited:
+        stdout.mark_launched()
+        stderr.mark_launched()
+        if pid > 0:
+            # Hold a surviving target until the host has captured its start
+            # identity. If that proof is unavailable, Kill still addresses
+            # this exact SBProcess created by the current launch request.
+            _wait_for_identity_ack(config)
+        process.Kill()
+        _wait_for_exit(listener, process, 5)
+
+    exit_observed = process.GetState() == lldb.eStateExited
+    expected_exit = exit_observed
+    stdout.close(expected_exit=expected_exit)
+    stderr.close(expected_exit=expected_exit)
+    return exit_observed
+
+
 def trace(debugger, config):
     global _TRACE_CONTEXT
     debugger.SetAsync(True)
@@ -630,8 +662,9 @@ def trace(debugger, config):
             )
         process = target.Launch(_launch_info(target, config, listener), error)
         if not error.Success() or not process.IsValid():
-            stdout.close()
-            stderr.close()
+            exit_observed = _retire_failed_launch(
+                process, config, listener, stdout, stderr
+            )
             return {
                 "status": "launch-error",
                 "error": error.GetCString(),
@@ -640,15 +673,26 @@ def trace(debugger, config):
                     if process.IsValid()
                     else {}
                 ),
+                **(
+                    {"target_process_exit_observed": True}
+                    if exit_observed
+                    else {}
+                ),
             }
         launch_failure = _failed_launch_description(process)
         if launch_failure is not None:
-            stdout.close(expected_exit=True)
-            stderr.close(expected_exit=True)
+            exit_observed = _retire_failed_launch(
+                process, config, listener, stdout, stderr
+            )
             return {
                 "status": "launch-error",
                 "error": launch_failure,
                 "target_process_created": True,
+                **(
+                    {"target_process_exit_observed": True}
+                    if exit_observed
+                    else {}
+                ),
             }
         stdout.mark_launched()
         stderr.mark_launched()

bridge/native/test_rea_lldb_tracer.py

@@ -15,6 +15,8 @@ fake_lldb = sys.modules.setdefault("lldb", types.ModuleType("lldb"))
 fake_lldb.LLDB_INVALID_ADDRESS = -1
 fake_lldb.eLaunchFlagDebug = 2
 fake_lldb.eLaunchFlagStopAtEntry = 4
+fake_lldb.eStateExited = 10
+fake_lldb.SBEvent = lambda: object()
 spec = importlib.util.spec_from_file_location(
     "rea_lldb_tracer", os.path.join(os.path.dirname(__file__), "rea_lldb_tracer.py")
 )
@@ -23,6 +25,136 @@ spec.loader.exec_module(tracer)
 
 
 class NativeTraceBudgetTests(unittest.TestCase):
+    def test_failed_launch_does_not_infer_exit_from_a_failed_kill(self):
+        class Process:
+            def IsValid(self):
+                return True
+
+            def GetProcessID(self):
+                return 12345
+
+            def GetState(self):
+                return 6
+
+            def Kill(self):
+                return None
+
+        class Sink:
+            close_exit = None
+
+            def mark_launched(self):
+                pass
+
+            def close(self, expected_exit=False):
+                self.close_exit = expected_exit
+
+        with tempfile.TemporaryDirectory() as directory:
+            stdout, stderr = Sink(), Sink()
+            with patch.object(tracer, "_wait_for_identity_ack", return_value="unavailable"), patch.object(
+                tracer, "_wait_for_exit", return_value=None
+            ):
+                exited = tracer._retire_failed_launch(
+                    Process(), {"pid_path": os.path.join(directory, "pid")}, object(), stdout, stderr
+                )
+        self.assertFalse(exited)
+        self.assertFalse(stdout.close_exit)
+        self.assertFalse(stderr.close_exit)
+
+    def test_failed_launch_publishes_pid_and_retires_live_process(self):
+        class Process:
+            state = 6
+            killed = False
+
+            def IsValid(self):
+                return True
+
+            def GetProcessID(self):
+                return 12345
+
+            def GetState(self):
+                return self.state
+
+            def Kill(self):
+                self.killed = True
+                self.state = tracer.lldb.eStateExited
+
+        class Sink:
+            launched = False
+            close_exit = None
+
+            def mark_launched(self):
+                self.launched = True
+
+            def close(self, expected_exit=False):
+                self.close_exit = expected_exit
+
+        with tempfile.TemporaryDirectory() as directory:
+            pid_path = os.path.join(directory, "pid")
+            process = Process()
+            stdout, stderr = Sink(), Sink()
+
+            def acknowledge(config):
+                with open(config["pid_path"], "r") as pid_file:
+                    self.assertEqual(pid_file.read(), "12345")
+                return "readable"
+
+            with patch.object(tracer, "_wait_for_identity_ack", acknowledge):
+                exited = tracer._retire_failed_launch(
+                    process,
+                    {"pid_path": pid_path},
+                    object(),
+                    stdout,
+                    stderr,
+                )
+
+        self.assertTrue(exited)
+        self.assertTrue(process.killed)
+        self.assertTrue(stdout.launched)
+        self.assertTrue(stderr.launched)
+        self.assertTrue(stdout.close_exit)
+        self.assertTrue(stderr.close_exit)
+
+    def test_failed_launch_records_when_valid_process_is_already_exited(self):
+        class Process:
+            def IsValid(self):
+                return True
+
+            def GetProcessID(self):
+                return 12345
+
+            def GetState(self):
+                return tracer.lldb.eStateExited
+
+            def Kill(self):
+                raise AssertionError("an exited process must not be killed")
+
+        class Sink:
+            close_exit = None
+
+            def mark_launched(self):
+                raise AssertionError("an exited process has no output to drain")
+
+            def close(self, expected_exit=False):
+                self.close_exit = expected_exit
+
+        with tempfile.TemporaryDirectory() as directory:
+            pid_path = os.path.join(directory, "pid")
+            stdout, stderr = Sink(), Sink()
+            with patch.object(
+                tracer,
+                "_wait_for_identity_ack",
+                side_effect=AssertionError("exit is already observed"),
+            ):
+                exited = tracer._retire_failed_launch(
+                    Process(), {"pid_path": pid_path}, object(), stdout, stderr
+                )
+            with open(pid_path, "r") as pid_file:
+                self.assertEqual(pid_file.read(), "12345")
+
+        self.assertTrue(exited)
+        self.assertTrue(stdout.close_exit)
+        self.assertTrue(stderr.close_exit)
+
     def test_pre_target_integrity_failure_reports_no_process_created(self):
         original_debugger = getattr(tracer.lldb, "SBDebugger", None)
 

bridge/process/ProcessRunTokenReader.swift

@@ -12,16 +12,29 @@ struct ProcessIdentityObservation: Encodable {
   let pid: Int32
   let state: String
   let identity: String?
+  let effective_uid: UInt32?
+  let lifecycle: String?
+  let relation: ProcessRelationObservation?
   let reason: String?
 }
 
-private struct ProcessMetadata {
+struct ProcessRelationObservation: Encodable {
+  let state: String
+  let parent_pid: Int32
+  let process_group_id: Int32
+}
+
+struct ProcessMetadata {
   let identity: String
+  let effectiveUid: UInt32
+  let lifecycle: String
+  let parentPid: Int32
+  let processGroupId: Int32
   let pointerSize: Int
 }
 
 struct ProcessSnapshot {
-  fileprivate let metadata: [Int32: ProcessMetadata]
+  let metadata: [Int32: ProcessMetadata]
   let failure: String?
 }
 
@@ -75,6 +88,10 @@ func readProcessSnapshot(
           process.kp_proc.p_pid,
           ProcessMetadata(
             identity: "\(start.tv_sec):\(start.tv_usec)",
+            effectiveUid: process.kp_eproc.e_ucred.cr_uid,
+            lifecycle: process.kp_proc.p_stat == SZOMB ? "zombie" : "non-zombie",
+            parentPid: process.kp_eproc.e_ppid,
+            processGroupId: process.kp_eproc.e_pgid,
             pointerSize: process.kp_proc.p_flag & P_LP64 != 0 ? 8 : 4
           )
         )
@@ -87,20 +104,61 @@ func readProcessSnapshot(
 }
 
 func readProcessIdentities(pids: [Int32]) -> [ProcessIdentityObservation] {
-  let snapshot = readProcessSnapshot()
+  readProcessIdentities(pids: pids, snapshot: readProcessSnapshot())
+}
+
+func readProcessIdentities(
+  pids: [Int32],
+  snapshot: ProcessSnapshot
+) -> [ProcessIdentityObservation] {
   return pids.map { pid in
+    guard pid > 0 else {
+      return unavailableIdentity(pid: pid, reason: "invalid_process_id")
+    }
     guard let metadata = snapshot.metadata[pid] else {
-      return unavailableIdentity(
+      if let failure = snapshot.failure {
+        return unavailableIdentity(pid: pid, reason: failure)
+      }
+      // This records absence from this successful KERN_PROC_ALL snapshot only.
+      // Unlike the Linux procfs path, it has no PID-specific kill(0)
+      // corroboration; callers must not treat it as universal or future exit
+      // proof.
+      return ProcessIdentityObservation(
         pid: pid,
-        reason: snapshot.failure ?? "process_not_in_snapshot"
+        state: "absent",
+        identity: nil,
+        effective_uid: nil,
+        lifecycle: nil,
+        relation: nil,
+        reason: "not-in-snapshot"
       )
     }
-    return ProcessIdentityObservation(pid: pid, state: "readable", identity: metadata.identity, reason: nil)
+    return ProcessIdentityObservation(
+      pid: pid,
+      state: "readable",
+      identity: metadata.identity,
+      effective_uid: metadata.effectiveUid,
+      lifecycle: metadata.lifecycle,
+      relation: ProcessRelationObservation(
+        state: "observed",
+        parent_pid: metadata.parentPid,
+        process_group_id: metadata.processGroupId
+      ),
+      reason: nil
+    )
   }
 }
 
 private func unavailableIdentity(pid: Int32, reason: String) -> ProcessIdentityObservation {
-  ProcessIdentityObservation(pid: pid, state: "unavailable", identity: nil, reason: reason)
+  ProcessIdentityObservation(
+    pid: pid,
+    state: "unavailable",
+    identity: nil,
+    effective_uid: nil,
+    lifecycle: nil,
+    relation: nil,
+    reason: reason
+  )
 }
 
 enum RunTokenReadError: Error {

bridge/pwndbg/core_context.py

@@ -20,8 +20,9 @@ def inspect(request):
     root = Path(os.environ["REA_CORE_CONTEXT_REQUEST"]).parent
     empty = root / "no-libraries"
     empty.mkdir(mode=0o700)
-    for command in ["set auto-load off", "set debuginfod enabled off", "set auto-solib-add off", "set history save off", "set pagination off",
-                    "set sysroot ./no-libraries", "set solib-search-path ./no-libraries"]:
+    for command in ["set auto-load off", "set debuginfod enabled off", "set auto-solib-add off", "set history save off", "set pagination off", "set target-file-system-kind unix",
+                    "set sysroot ./no-libraries", "set solib-search-path ./no-libraries",
+                    "set debug-file-directory ./no-libraries"]:
         gdb.execute(command, to_string=True)
     entry = request["gdbinit"]
     try:
@@ -38,7 +39,19 @@ def inspect(request):
     # relative filename in the owned cwd so no selected path becomes GDB syntax.
     if Path(request["snapshot_path"]) != root.parent / "object.snapshot":
         raise ContextFailure("decoder", "Core context request did not select the shared owned snapshot.")
+    # GDB 16 searches AT_EXECFN/NT_FILE outside sysroot when no executable BFD
+    # is selected. Occupy that slot with the already running, owned debugger;
+    # exec-file loads no symbols and never runs it. Clear it before any query.
+    # /proc/self/exe identifies this process, not the core's historical PID.
+    architecture = {"x86_64-little-endian": "i386:x86-64", "aarch64-little-endian": "aarch64"}.get(request["architecture"])
+    if architecture is None:
+        raise ContextFailure("unsupported", "Unsupported recorded core architecture.")
+    gdb.execute("exec-file /proc/self/exe", to_string=True)
+    # The placeholder must not select the core's architecture. Reuse the
+    # decoder's validated profile, including for cross-architecture GDB builds.
+    gdb.execute("set architecture " + architecture, to_string=True)
     gdb.execute("core-file ../object.snapshot", to_string=True)
+    gdb.execute("file", to_string=True)
     inferior = gdb.selected_inferior()
     if inferior.connection is None or inferior.connection.type != "core" or gdb.current_progspace().filename is not None:
         raise ContextFailure("decoder", "Requested core-only boundary was not established; no mapping query performed.")
@@ -82,7 +95,7 @@ def main():
     try:
         encoded = json.dumps(reply, ensure_ascii=True, allow_nan=False).encode("utf-8")
         if len(encoded) > OUTPUT_BYTES:
-            encoded = json.dumps({"ok": False, "reason": "output-limit", "message": "Complete debugger context exceeds64 MiB reply budget."}).encode("utf-8")
+            encoded = json.dumps({"ok": False, "reason": "output-limit", "message": "Complete debugger context exceeds the 64 MiB reply budget."}).encode("utf-8")
         descriptor = os.open(request["reply_path"], os.O_WRONLY | os.O_CREAT | os.O_EXCL, 0o600)
         with os.fdopen(descriptor, "wb") as handle:
             handle.write(encoded)

bridge/pwndbg/test_core_context.py

@@ -0,0 +1,126 @@
+"""Producer regressions with controlled GDB state, not native debugger qualification."""
+import json
+import os
+from pathlib import Path
+import runpy
+import sys
+import tempfile
+import types
+import unittest
+from unittest.mock import patch
+
+
+BRIDGE = Path(__file__).with_name("core_context.py")
+
+
+class CoreContextTests(unittest.TestCase):
+    def inspect(self, failure=None, connection="core", symbols=None, architecture="x86_64-little-endian", objfile="recorded-executable"):
+        commands, reads, queries = [], [], []
+        state = {"executable": None, "symbols": None, "connection": None}
+        gdb = types.ModuleType("gdb")
+        gdb.VERSION = "16.3"
+        gdb.selected_inferior = lambda: types.SimpleNamespace(
+            connection=state["connection"], pid=123)
+        gdb.current_progspace = lambda: types.SimpleNamespace(filename=state["symbols"])
+
+        def execute(command, to_string):
+            self.assertTrue(to_string)
+            commands.append(command)
+            if command == failure:
+                raise RuntimeError("controlled command failure")
+            if command == "exec-file /proc/self/exe":
+                state["executable"] = "/proc/self/exe"
+            elif command == "core-file ../object.snapshot":
+                if state["executable"] is None:
+                    reads.append("recorded-executable")
+                    state["executable"] = state["symbols"] = "recorded-executable"
+                state["connection"] = types.SimpleNamespace(type=connection)
+            elif command == "file":
+                state["executable"] = None
+                state["symbols"] = symbols
+
+        gdb.execute = execute
+        pwndbg = types.ModuleType("pwndbg")
+        pwndbg.aglib = types.ModuleType("pwndbg.aglib")
+        pwndbg.aglib.vmmap = types.ModuleType("pwndbg.aglib.vmmap")
+        pwndbg.lib = types.ModuleType("pwndbg.lib")
+        pwndbg.lib.version = types.ModuleType("pwndbg.lib.version")
+        pwndbg.lib.version.__version__ = "2026.09.15"
+
+        def maps():
+            self.assertIsNone(state["executable"], "Helper executable must be discarded before mapping queries")
+            queries.append("maps")
+            return [types.SimpleNamespace(start=4096, end=8192, offset=0, flags=5, objfile=objfile)]
+
+        pwndbg.aglib.vmmap.get = maps
+        modules = {module.__name__: module for module in [
+            gdb, pwndbg, pwndbg.aglib, pwndbg.aglib.vmmap, pwndbg.lib, pwndbg.lib.version]}
+        with tempfile.TemporaryDirectory() as directory:
+            root = Path(directory)
+            stage = root / "debugger"
+            stage.mkdir()
+            plugin = root / "gdbinit.py"
+            plugin.write_text("")
+            request = stage / "request.json"
+            reply = stage / "reply.json"
+            request.write_text(json.dumps({"gdbinit": str(plugin), "architecture": architecture, "snapshot_path": str(root / "object.snapshot"), "reply_path": str(reply)}))
+            with patch.dict(sys.modules, modules), patch.dict(os.environ, REA_CORE_CONTEXT_REQUEST=str(request)):
+                runpy.run_path(str(BRIDGE))
+            result = json.loads(reply.read_text())
+        return result, commands, reads, queries
+
+    def test_core_load_never_opens_recorded_executable_and_queries_only_after_discard(self):
+        result, commands, reads, queries = self.inspect()
+        self.assertTrue(result["ok"], result)
+        self.assertEqual(reads, [])
+        self.assertEqual(queries, ["maps"])
+        self.assertEqual(commands[-4:], ["exec-file /proc/self/exe", "set architecture i386:x86-64", "core-file ../object.snapshot", "file"])
+        self.assertLess(commands.index("set debug-file-directory ./no-libraries"), commands.index("core-file ../object.snapshot"))
+        self.assertIsNone(result["value"]["executable"])
+        self.assertEqual(result["value"]["maps"][0]["current_file_identity"], "unknown")
+
+    def test_failed_guard_core_load_or_discard_never_queries_mappings(self):
+        for command in ["exec-file /proc/self/exe", "set architecture i386:x86-64", "core-file ../object.snapshot", "file"]:
+            with self.subTest(command=command):
+                result, commands, reads, queries = self.inspect(failure=command)
+                self.assertFalse(result["ok"])
+                self.assertEqual(reads, [])
+                self.assertEqual(queries, [])
+                if command in ["exec-file /proc/self/exe", "set architecture i386:x86-64"]:
+                    self.assertNotIn("core-file ../object.snapshot", commands)
+
+    def test_cross_architecture_uses_the_validated_decoder_profile(self):
+        result, commands, reads, queries = self.inspect(architecture="aarch64-little-endian")
+        self.assertTrue(result["ok"], result)
+        self.assertEqual(reads, [])
+        self.assertEqual(queries, ["maps"])
+        self.assertLess(commands.index("set architecture aarch64"), commands.index("core-file ../object.snapshot"))
+
+    def test_unknown_architecture_refuses_before_loading_core(self):
+        result, commands, reads, queries = self.inspect(architecture="unknown")
+        self.assertFalse(result["ok"])
+        self.assertEqual(result["reason"], "unsupported")
+        self.assertNotIn("core-file ../object.snapshot", commands)
+        self.assertEqual(reads, [])
+        self.assertEqual(queries, [])
+
+    def test_output_limit_message_separates_the_words_and_names_the_limit(self):
+        result, _, reads, queries = self.inspect(objfile="m" * (64 * 1024 * 1024))
+        self.assertFalse(result["ok"])
+        self.assertEqual(result["reason"], "output-limit")
+        self.assertNotIn("exceeds64", result["message"])
+        self.assertEqual(result["message"], "Complete debugger context exceeds the 64 MiB reply budget.")
+        self.assertEqual(reads, [])
+        self.assertEqual(queries, ["maps"])
+
+    def test_wrong_connection_or_remaining_symbols_never_queries_mappings(self):
+        for options in [{"connection": "native"}, {"symbols": "unexpected-symbols"}]:
+            with self.subTest(options=options):
+                result, _, reads, queries = self.inspect(**options)
+                self.assertFalse(result["ok"])
+                self.assertEqual(reads, [])
+                self.assertEqual(queries, [])
+
+
+if __name__ == "__main__":
+    unittest.main()

docs/.vitepress/config.ts

@@ -29,7 +29,13 @@ export default defineConfig({
         items: [
           { text: "Native binaries", link: "/native-investigation" },
           { text: "Go build metadata", link: "/go-binaries" },
+          { text: "IL2CPP metadata", link: "/il2cpp-metadata" },
+          { text: "Model artifacts", link: "/model-artifacts" },
           { text: "Rust dependencies", link: "/rust-binaries" },
+          { text: "PE certificates", link: "/pe-certificates" },
+          { text: "ELF notes", link: "/elf-notes" },
+          { text: "CPython pyc files", link: "/python-pyc" },
+          { text: "Erlang BEAM", link: "/erlang-beam" },
           {
             text: "JavaScript and Electron",
             link: "/javascript-artifact-reconstruction",
@@ -43,12 +49,18 @@ export default defineConfig({
             text: "Web network captures",
             link: "/web-network-captures",
           },
+          { text: "Agent environments", link: "/agent-environments" },
           { text: "Electron observation", link: "/electron-observation" },
           { text: "Electron fuses", link: "/electron-fuses" },
           { text: "Process capture", link: "/process-capture" },
           { text: ".NET assemblies", link: "/managed-code-analysis" },
+          {
+            text: "Python packaged applications",
+            link: "/python-packaged-applications",
+          },
           { text: "Android applications", link: "/android-analysis" },
           { text: "Apple applications", link: "/apple-application-analysis" },
+          { text: "Desktop installations", link: "/desktop-applications" },
           { text: "Firmware", link: "/firmware-analysis" },
           { text: "SQLite databases", link: "/sqlite-databases" },
           { text: "Packet captures", link: "/packet-captures" },
@@ -75,6 +87,10 @@ export default defineConfig({
         items: [
           { text: "MCP contracts", link: "/mcp-contracts" },
           { text: "Tool design", link: "/tool-design" },
+          {
+            text: "Admission limits and recovery",
+            link: "/design/limits-and-recovery",
+          },
           { text: "Testing", link: "/testing" },
           { text: "Client verification", link: "/client-verification" },
           {

docs/adb-device-analysis.md

@@ -90,7 +90,7 @@ existing analysis families directly, for example
 | `read_adb_logcat`       | read               | Bounded `logcat -d -t` dump for one buffer, optional pid filter                                                                                                                                                                       |
 | `list_adb_packages`     | read               | `pm list packages -f` with scope; base APK device paths                                                                                                                                                                               |
 | `resolve_adb_packages`  | read               | Case-insensitive package-name substring search with exact-match flag                                                                                                                                                                  |
-| `inspect_adb_package`   | read               | `dumpsys package` projection: versions, install times, installer, flags, permission count                                                                                                                                             |
+| `inspect_adb_package`   | read               | `dumpsys package` projection: versions, install times, installer, flags, and count of listed requested-permission rows (including Unicode names)                                                                                      |
 | `list_adb_directory`    | read               | NUL-framed immediate children, including hidden entries: kinds, permissions, owner, group, sizes, symlink targets                                                                                                                     |
 | `pull_adb_package`      | read               | Complete split APK set from `pm path`, each file digested                                                                                                                                                                             |
 | `pull_adb_file`         | read               | One absolute device path pulled and digested                                                                                                                                                                                          |
@@ -98,16 +98,42 @@ existing analysis families directly, for example
 | `collect_adb_bugreport` | read               | Device-generated bugreport zip, digested                                                                                                                                                                                              |
 | `push_adb_file`         | **mutates device** | One local file to a device path (e.g. pushing tools into `/data/local/tmp`); checks for existing targets and refuses them without explicit overwrite (the check is not atomic); compares digests when device `sha256sum` is available |
 | `install_adb_package`   | **mutates device** | Local APK via `adb install [-r]`; preserves its Success/Failure verdict, records `replace_requested`, and leaves actual replacement `null` because prior package state is not observed                                                |
-| `uninstall_adb_package` | **mutates device** | Removes a package and its data for user 0                                                                                                                                                                                             |
+| `uninstall_adb_package` | **mutates device** | Requests package removal for user 0 via `adb shell pm uninstall --user 0`, without data retention                                                                                                                                     |
 | `start_adb_app`         | **mutates device** | Resolves the package's launcher activity (`cmd package resolve-activity --brief`) and starts it (`am start -n`)                                                                                                                       |
 | `start_adb_activity`    | **mutates device** | Starts an intent (`am start -a` with optional URI, component, and up to 16 typed extras); preserves the activity manager's verdict                                                                                                    |
 | `stop_adb_app`          | **mutates device** | Force-stops every process of a package, ending a dynamic-analysis run                                                                                                                                                                 |
 
+When `push_adb_file` receives an existing directory, adb places the local file
+inside it under the local basename; Evidence reports that effective file path
+while retaining the caller's directory path in the input parameters. The
+private read-only staging copy preserves the source file's executable bits.
+
+For `inspect_adb_window`, `focus_contexts` retains each display header, its exact
+display ID when recognized, and every focus declaration with its original line.
+Declarations without a display header are explicitly `unscoped`. A missing
+declaration differs from a declaration whose `value` is JSON `null`.
+`source_section` identifies the selected dumpsys section; `observed_lines`
+preserves its focus declarations and display headers in source order.
+
+`focus_summary` is `single_context` only for one context with no repeated focus
+field. The singular `focused_window` and `focused_app` summarize that context;
+they do not establish device-wide focus. Multiple contexts or repeated fields
+produce `ambiguous` with null singular fields and all declarations retained.
+When no focus field was declared, the status is `unavailable`. REA reads the
+displays section only when the windows section declares neither focus field.
+
 For `install_adb_package`, `replace_requested` records whether `-r` was sent.
 The `replaced` result is `null`: the successful install response does not say
 whether the package was installed fresh or replaced, and REA does not query the
 device first. The install verdict and host APK digest remain available.
 
+Push and install consume a private snapshot of the selected local input. The
+reported size and SHA-256 identify those consumed bytes; `local_path` and
+`apk_path` retain the original input path, whose contents may change later.
+Snapshot creation needs temporary disk space proportional to the input size.
+Failed snapshot cleanup remains owned and is retried before later operations
+and during shutdown.
+
 ## Interpret the result
 
 - **Directory names retain whitespace and newlines.** Symlink targets do too,
@@ -115,8 +141,10 @@ device first. The install verdict and host APK digest remain available.
   labelled hex in `unparsed_lines` and make directory coverage partial.
 - **Inferences carry their basis.** Device `kind` reports
   `kind_basis` (serial prefix or USB transport); APK `role` reports
-  `role_basis: file_name`; `emulator_observed` is the build's own
-  `ro.kernel.qemu` value, `null` when absent.
+  `role_basis: file_name`; `emulator_observed` is derived from the build's
+  `ro.kernel.qemu` property: `true` only when its value is `"1"`, `false` for
+  any other reported value, and `null` when the property is absent. The raw
+  property value remains available in `properties`.
 - **Device paths are not host paths.** `base_apk_device_path` values live on
   the device filesystem. Only pulls and captures produce host-local paths,
   each paired with byte size and SHA-256 digest. A pulled file's host name is
@@ -141,14 +169,19 @@ device first. The install verdict and host APK digest remain available.
   APKs keep other names report the `unknown` role rather than guessing.
 - **Partial results keep facts.** Failed individual pulls return successful
   artifacts with `coverage: "partial"` and per-file `failures`. Listings
-  that contain unrecognized lines report them as `unparsed_lines` and mark
-  partial coverage instead of dropping rows.
+  and package searches that contain unrecognized lines report the first ten
+  as `unparsed_lines`, state the exact total in a limitation when more were
+  seen, and mark partial coverage instead of dropping rows.
 - **Mutations are explicit.** `push_adb_file`, `install_adb_package`, and
   `uninstall_adb_package` declare `mutatesTarget` and never run implicitly:
   pushes check for existing device files before transfer and refuse them
-  unless `overwrite` is set. The check is not atomic, so a different process
-  can create the path before `adb push` writes it. When available, the device's
-  `sha256sum` verifies the transfer; installs preserve the
+  unless `overwrite` is set. A failed or unexpected existence probe stops
+  transfer even with `overwrite`; only a clean `test -e` exit of 1 with no
+  output is interpreted as absence. The check is not atomic, so a different
+  process can create the path before `adb push` writes it. When available, the device's
+  `sha256sum` verifies the transfer. A missing command or output without a
+  recognizable digest reports `device_digest_source: "unavailable"` with an
+  unverified-transfer limitation; installs preserve the
   device's `Failure [...]` reason verbatim.
 - **Producer identity stays attached.** Every result embeds the producing
   adb client (`binary_path`, `path_source`, `version`, `installed_path`).
@@ -160,14 +193,21 @@ device first. The install verdict and host APK digest remain available.
   The app/activity lifecycle tools use `am start -W` and `am force-stop`.
 - Per-invocation budgets: client/devices 10-20 s, shell reads 20-60 s with
   64 KiB-8 MiB capture caps (logcat 4 MiB, screen 32 MiB), transfers and
-  bugreports 600 s. Exceeded budgets report `resource_constraint`; stopped
-  commands report `provider_timeout`.
+  bugreports 600 s. Output capture overflow reports `resource_constraint`
+  and names the byte allowance. Deadline expiration, including the
+  client-version probe, reports `provider_timeout` and names the effective
+  milliseconds. This release has no runtime override for those deadlines or
+  capture caps. Command cancellation preserves the caller's exact reason;
+  a later abort does not replace an already completed command failure.
 - Pull and capture outputs are additive: the destination must not already
   exist. Re-running a completed pull refuses the existing directory instead
   of overwriting evidence.
 - Cancellation during a package pull preserves caller-selected output. The
   error includes partial observations identifying completed artifacts and the
   interrupted local path; remaining APKs were not attempted.
+- Screen capture checks cancellation before publishing its exclusive output.
+  If cancellation arrives after publication, the error retains the committed
+  file's path, digest and dimensions; the file remains available as evidence.
 - Bugreports collect into isolated staging and publish with exclusive creation;
   filename collisions select the next numeric suffix. If publication fails, the
   completed archive remains in staging. The error's partial observation gives

docs/agent-environments.md

@@ -0,0 +1,851 @@
+# Agent environments from recorded captures
+
+Use `rea environment-serve` to serve an environment manifest as an agent
+environment: an MCP server whose tools answer from what one retained web
+network capture observed, plus the environment card the agentenv framework
+reads to discover it. By default the server derives the
+[state the capture observed](#the-state-a-capture-observed) at startup and
+answers read tools from it, and the write tools whose effect the capture
+confirmed by changing it; `--replay-only` replays the capture's records
+instead. Nothing is fetched from the recorded application; every answer comes
+from retained Evidence. `rea environment-export` writes the same pair as a
+bundle folder `agent-env run` builds and runs; see
+[Export a bundle](#export-a-bundle).
+
+```sh
+rea inspect-web-network-capture /absolute/session.har har --format json > capture.json
+rea inspect-analysis-view view.json --format json > manifest.json
+rea environment-serve manifest.json --evidence capture.json --port 8765
+```
+
+The first command inspects the capture; the second projects the
+[environment manifest](web-network-captures.md#environment-manifest) view from
+it, where `view.json` wraps the capture Evidence as the inline source:
+
+```json
+{
+  "source": { "kind": "inline", "evidence": { "evidence_id": "ev_…" } },
+  "view": {
+    "kind": "environment-manifest",
+    "template_rules": ["after-uid-literal", "opaque-token", "sibling-variance"]
+  }
+}
+```
+
+The third command serves that manifest from the state it derives from the
+capture; see [Serve from the state](#serve-from-the-state). Both served files
+are complete REA Evidence JSON as the commands write it with `--format json`;
+the `evidence` member of `view.json` is the whole of `capture.json`.
+
+## From a browser scenario
+
+The capture can also come from REA itself: a
+[browser scenario](browser-scenario-contract.md) records a scripted session
+in a browser REA launches, and its Evidence is a manifest parent like a HAR
+inspection. Select the `network` event family and response bodies. A replayed
+reply carries up to 256 KiB of a recorded body, but the scenario keeps at most
+`max_network_body_bytes` of each body (64 KiB by default), and only within its
+allocation budgets: base64 is charged about 11 bytes per body byte, and
+redaction more the shorter a declared secret is (about 35 bytes per body byte
+for a 12-character `admin_password`). A body over
+`max_network_body_allocation_bytes` (1 MiB by default) is kept as `unavailable`
+with that reason; one over `max_event_allocation_bytes` (4 MiB) or
+`max_total_event_allocation_bytes` (16 MiB) loses its whole content event,
+headers included, which the capture lists among its omissions. Either way its
+tool replays no body. The example raises all four so that 256 KiB bodies are
+kept when the declared secret has at least 12 characters:
+
+```json
+{
+  "browser": { "mode": "launch", "executable_path": "/usr/bin/chromium" },
+  "start_url": { "url": "http://127.0.0.1:5230/auth", "query": [] },
+  "secrets": [
+    { "secret_id": "admin_password", "environment_variable": "APP_PASSWORD" }
+  ],
+  "actions": [
+    {
+      "step_id": "user",
+      "action": "fill",
+      "locator": { "kind": "css", "selector": "input[name=username]" },
+      "value": { "source": "literal", "value": "admin" }
+    },
+    {
+      "step_id": "password",
+      "action": "fill",
+      "locator": { "kind": "css", "selector": "input[type=password]" },
+      "value": { "source": "secret", "secret_id": "admin_password" }
+    },
+    {
+      "step_id": "sign-in",
+      "action": "press",
+      "locator": { "kind": "css", "selector": "input[type=password]" },
+      "key": "Enter"
+    },
+    {
+      "step_id": "first-memo",
+      "action": "goto",
+      "destination": { "url": "http://127.0.0.1:5230/m/1", "query": [] },
+      "wait_until": "load"
+    }
+  ],
+  "capture": {
+    "events": ["network"],
+    "network": {
+      "request_body": true,
+      "response_body": true,
+      "header_values": false
+    },
+    "limits": {
+      "max_network_body_bytes": 262144,
+      "max_network_body_allocation_bytes": 12582912,
+      "max_event_allocation_bytes": 12582912,
+      "max_total_event_allocation_bytes": 67108864
+    }
+  }
+}
+```
+
+```sh
+APP_PASSWORD=… rea capture-browser-scenario scenario.json --format json > capture.json
+rea inspect-analysis-view view.json --format json > manifest.json   # view.json wraps capture.json as before
+rea environment-serve manifest.json --evidence capture.json --port 8765
+```
+
+Each network transaction becomes one record, in request order, and the
+record's ordinal is its zero-based position; a transaction whose URL carried a
+redacted secret is left unrepresented rather than templated, and one the
+capture cannot attribute to a single request and response is left
+unrepresented too. Tools are chosen by response media type, so a scenario that
+selected neither response bodies nor header values is refused; one that
+selected header values but no response bodies yields a manifest but no state,
+so it is served by replay. The browser
+scenario redacts credential headers and the secrets it declares; values the
+application returns in response bodies, such as session tokens, are kept, and
+the replayed body is the bytes the browser exposes for each response, with any
+declared secret replaced by its redaction marker (a body that redaction had to
+empty replays no body).
+
+## What a manifest is
+
+An environment manifest is a projection of one application's recorded request
+lines into a tool surface: one tool per method and path template on the
+primary origin, with inferred path parameters, observed query parameter
+names, value-free body shapes and pointers to the parent records each tool
+cites. It carries no bodies and no headers. The manifest's
+`parent_evidence_id` names the capture Evidence it was projected from; the
+serve command refuses a manifest whose parent is not the supplied `--evidence`
+file with an `invalid_request` error naming both identifiers, so a manifest can
+never be served from records it did not cite.
+
+The MCP catalog and environment card use a 1 KiB UTF-8 budget for each
+captured display sample. Observation, body-key and cited-record lists show
+whole values with omission counts; long method, path and query-name displays
+show a marked prefix. Complete values remain in manifest Evidence. Argument
+names longer than 64 characters use collision-safe `path_<n>` or `query_<n>` aliases; replay maps
+those aliases to the complete captured names.
+
+Discovery is admitted as a complete catalog under a 1 MiB serialization
+reservation before the listener binds. Reservation charges up to six bytes per
+UTF-16 code unit for JSON escaping/UTF-8 expansion, plus schema overhead;
+a catalog may therefore be refused even when its compact JSON is smaller.
+This was calibrated with 1500 query fields and eight concurrent SDK discovery
+requests (183,798-byte card, 507.27 MiB sampled process-family RSS). Over-capacity
+catalogs return `resource_constraint`, retain complete input Evidence, and require
+a smaller capture or manifest. Fields and tools are never silently omitted.
+Served from the state, each description is longer and the card declares the
+data plane, so a catalog replay admits can exceed the reservation; the state
+is then unavailable and the server serves replay instead (see
+[Serve from the state](#serve-from-the-state)).
+
+## What replay means and does not mean
+
+Replay answers every tool with `--replay-only` or when the state is unavailable,
+and, served from the state, each read tool whose view the state does not back.
+Each such MCP tool call selects one of the records its manifest tool cites and
+returns that record's recorded response. Selection compares the supplied path
+parameters against the record's path segments at the parameter indexes, then
+the supplied query values against the record's query string, and keeps the
+record with the most matches; ties keep manifest order. A query argument accepts
+one string or an array of strings for repeated values; the array is matched
+against every occurrence in captured order. The raw recorded query remains the
+replay evidence. The result reports
+where the answer came from:
+
+```json
+{
+  "status": 200,
+  "content_type": "application/json",
+  "body": { "id": 2, "login": "probe" },
+  "replay": {
+    "record_ordinal": 5,
+    "evidence_id": "ev_…",
+    "exact_match": true,
+    "recorded_path": "/api/users/2",
+    "recorded_query": null,
+    "recorded_line_truncated": false,
+    "body_recorded": true,
+    "body_encoding": "utf-8",
+    "truncated": false
+  }
+}
+```
+
+`exact_match` is true when every supplied path parameter and query occurrence
+matches the recorded one; it says nothing about parameters the caller omitted.
+Comparison probes retain the raw query spelling and its decoded parameter pairs
+in request order, including the interleaving of repeated values across names.
+Standalone tool calls express repeated values as per-name arrays; when matching
+records differ only in cross-name order, `exact_match` is false because those arguments cannot
+identify that order. Recorded query parsing runs only when supplied query
+arguments affect selection.
+A path value matches a recorded segment as written or after percent-decoding;
+when both kinds of match are available, the record whose segment equals the
+value as written is chosen. `recorded_query` keeps the `?` of an explicit empty
+query, and a recorded path or query longer than 1 KiB is cut in the reply with
+`recorded_line_truncated: true`; selection always uses the full recorded line.
+`content_type` also has a 1 KiB display bound with `content_type_truncated`; body
+interpretation uses the full recorded media type.
+When no record matches, the tool still answers from the tool's first cited
+record and reports `exact_match: false`. `body` is parsed JSON when the
+recorded content type is JSON-like (`application/json`, `application/*+json`,
+`text/plain`, `text/markdown`) and the text parses with safe numeric conversion; otherwise
+it is the recorded text. Large integer IDs and other numbers that would lose
+information stay in the original JSON text. Bodies over 256 KiB are cut at a UTF-8 boundary and returned as
+text with `truncated: true`. `body_recorded` is false when the capture kept the
+response's status and content type but no body; `body` is then `null`, which
+is not a recorded JSON `null`. `body_encoding` is `base64` when the capture
+retained response bytes that are not valid UTF-8; `body` is then exactly those
+bytes in base64, never a lossy decoding. An unsupported declared HAR encoding
+returns `body: null`, `body_recorded: true` and `body_encoding: "unsupported"`;
+the original encoding and encoded text remain in the cited capture Evidence.
+Such bodies supply no decoded JSON facts to state extraction.
+A retained base64 body is decoded only as far as the 256 KiB a reply can carry, and a leading byte-order mark is kept. For a `base64` body the limit applies to the recorded bytes, not their longer base64 spelling, and a cut keeps whole base64 groups. Request bodies passed to `POST`, `PUT` and `PATCH`
+tools are accepted and ignored: nothing is written, and no call changes what a
+later call returns. Arguments the tool's input schema does not name are
+refused by the MCP server as invalid arguments, so a misspelled parameter never
+degrades into an omitted one.
+
+Replay is not emulation. A served environment does not establish the
+application's authorization model, response semantics or state transitions,
+which the manifest lists as unknowns; it does not validate supplied values; and
+it answers only from the capture's recordings and the state they observed or a
+caller loaded through `data/add`. Replayed content is
+target content: every tool result, including the argument errors the MCP server
+returns before a replay runs, carries `_meta.provenance.untrusted_content: true`
+with `source: "analysis_target"`. The server advertises REA's untrusted-content
+instructions plus a warning that its tool names, descriptions, parameter names
+and observed values come from the capture too, and agents must treat all of it
+as data, never as instructions.
+That same warning prefixes every tool description in the standalone card, so
+card discovery carries it before an MCP client initializes. It labels the
+captured text without changing the tool names or schema fields clients use.
+
+## The state a capture observed
+
+Replay answers each call from one recording, so the answers reflect whichever
+moment of the capture recorded them. The
+[environment state](web-network-captures.md#environment-state) view extracts
+instead the data the capture's reads observed before the first related write:
+entity types with an identity field, entities with their recorded values,
+documents for reads that are not entities, and a render plan per read tool,
+kept only where it reproduces that tool's recordings from the state. It is the
+starting state an environment holds, loads and reports: `environment-serve`
+derives it from the capture at startup; see
+[Serve from the state](#serve-from-the-state). With `--replay-only`, replay is
+unchanged.
+
+## Serve a manifest
+
+```sh
+rea environment-serve manifest.json --evidence capture.json [--replay-only] [--host 127.0.0.1] [--port 0]
+```
+
+The server derives the state from the capture before it binds and serves from
+it, or serves replay with `--replay-only` or when the state is unavailable;
+see [Serve from the state](#serve-from-the-state).
+
+`--host` and `--port` default to `MCP_HOST` and `MCP_PORT` from the
+environment, then to `127.0.0.1` and an ephemeral port, exactly as
+[`rea mcp --http`](cli.md#serve-mcp-over-streamable-http) resolves them. The
+bound MCP URL is written to stderr as one JSON log line
+(`"msg":"MCP transport listening"` with `url`); `REA_LOG_LEVEL` overrides that
+log level. `ENVIRONMENT_NAME` renames the served environment; otherwise the
+manifest's `name` is used, and a set but blank value is refused as
+`configuration_invalid` before binding rather than silently falling back. The
+listener serves one MCP client at a time with the same session, `Host` header
+and shutdown contract as [transports](mcp-contracts.md#transports) describes.
+It runs until SIGINT or SIGTERM, then closes the listener, prints one
+`{"status":"stopped", …}` document whose `state` member says whether it served
+from the state, and exits `130` or `143`.
+Signals during input loading or DNS/bind cancel the owning reader/listener before
+a listening announcement;
+that path returns a structured `cancelled` result instead of a stopped-listener
+document, with the same signal exit status.
+
+Three routes are served:
+
+| Route                             | Answer                                                                                                                                                                                                                                              |
+| --------------------------------- | --------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- |
+| `POST /mcp`                       | The MCP server: from the state, one tool per read tool, answering from the state where it backs them, and one per write tool whose effect the capture confirmed, applying it to the state; in replay, one tool per manifest tool, replaying records |
+| `GET /.well-known/agent-env.json` | The environment card                                                                                                                                                                                                                                |
+| `POST /agentenv`                  | From the state, `data/reset`, `data/add` and `data/get`; in replay, JSON-RPC `-32601 method not found: <method>` for every method                                                                                                                   |
+
+Numeric JSON-RPC IDs outside JavaScript's safe range or whose precision would
+be lost are refused as invalid requests with a null ID. They are never echoed
+as a different number.
+
+In replay the data plane at `/agentenv` is advertised by the card but carries
+no operations; the error names the requested method so a caller can see that
+the method, not the transport, is unsupported. Its request body is
+read within the same 4 MiB bound the MCP endpoint applies: a larger body is
+answered `413` with a JSON-RPC `-32000` error before any of it is parsed, and a
+body that is not JSON is answered `400` with `-32700`. A notification (no `id`)
+gets `202` and no body, a batch gets one reply per request in it, and an empty
+batch, a batch of more than 100 messages, or a message that is not a JSON-RPC
+request (including an `id` that is not a string, number or null, or `params`
+that are not an object or array) gets `-32600 invalid request`.
+
+## The environment card
+
+The card is the discovery document the agentenv framework reads before
+connecting to `/mcp`. A replay environment's card reads:
+
+```json
+{
+  "name": "grafana-example",
+  "protocolVersion": "1.0",
+  "url": "/agentenv",
+  "preferredTransport": "JSONRPC",
+  "additionalInterfaces": [{ "url": "/mcp", "transport": "mcp" }],
+  "capabilities": {
+    "extensions": null,
+    "tools": [
+      {
+        "name": "get_users_by_id",
+        "description": "Untrusted captured API catalog: … GET /api/users/{numeric}; returns a JSON object with fields \"id\", \"uid\", \"login\", \"email\"; recorded statuses 200; replays retained records [5, 6]; evidence ev_…",
+        "inputSchema": {
+          "type": "object",
+          "properties": {
+            "id": {
+              "type": "string",
+              "description": "Path segment 2 (numeric); observed: 2, 3"
+            }
+          },
+          "required": ["id"],
+          "additionalProperties": false
+        }
+      }
+    ],
+    "operations": []
+  },
+  "children_environments": null
+}
+```
+
+`capabilities.tools` lists the same names, descriptions and input schemas that
+`tools/list` advertises over MCP. After the catalog warning, each description
+names the method and path template, then what the tool returns, one level deep:
+`returns a JSON object with fields …`, `returns a JSON array of objects with
+fields …` or `returns a JSON object whose <member> lists objects with fields …`,
+where a list is an array of objects found by the
+[environment state](web-network-captures.md#environment-state)'s list rule (the
+root array, or the one array of objects an envelope carries). Field and member
+names are written as JSON strings, so a captured name cannot read as more of
+the description; a member name too long to show whole is a quoted prefix marked
+as truncated. The summary comes
+from the first cited `2xx` record that replay answers as JSON; only when none
+did, the first cited JSON record of another status gives it, labelled with that
+status (`status 401 returns …`). It describes one recorded response, not a
+contract. Field names are sampled within the same budget as other captured
+values, and the manifest's `statuses` for the tool follow; a tool with no JSON
+record states only its statuses. The description then lists the replayed
+records and the evidence ID, and the whole catalog stays within its
+serialization reservation, which the summaries and MCP annotations are charged
+to.
+
+Over MCP, every tool also carries annotations: `readOnlyHint` is true for the
+safe methods (`GET`, `HEAD`, `OPTIONS` and `TRACE`) and for query `POST`s (a
+path segment such as `search` or `query`); `destructiveHint` is true for `DELETE` and left unset for other
+writes, so the MCP default (destructive) applies to them, and read-only tools
+omit it because clients ignore it there; `idempotentHint` is true for the safe
+methods, `PUT` and `DELETE`; and `openWorldHint` is false for every tool,
+because answers come from the capture rather than the captured application. The
+hints describe the captured method; replay itself changes nothing. The card
+keeps the `name`, `description` and `inputSchema` of the echo example's tool
+shape and carries no annotations.
+
+Path parameters are required strings whose
+description names the segment index, the templating rule and the observed
+values (sampled within the list budget above; the manifest keeps them all);
+query parameter names accept an optional string or array of strings (for
+repeated values in captured order); write methods accept an
+optional `body` object described by the recorded top-level key types. Each
+input keeps its own argument: a query name equal to a path parameter's name is
+advertised as `query_<name>`, a query name longer than 64 characters as
+`query_<n>` (its position) with the captured name sampled in its description,
+and the write body as `request_body` when a query parameter is named `body`. Every
+schema closes with `additionalProperties: false`, a key the echo example's
+schema does not carry.
+In replay, `extensions`, `operations` and `children_environments` are fixed to
+`null`, `[]` and `null`: a replay environment declares no extensions, no
+data-plane operations and no children. Served
+[from the state](#serve-from-the-state), the card declares the three data-plane
+operations and one intake extension instead, and the catalog holds the read
+tools and only the write tools whose effect the capture confirmed: a read tool
+is described as answering from the state where the state backs it, and a
+served write tool by the effect it applies to the state and the quoted body members it
+accepts.
+
+## Serve from the state
+
+```sh
+rea environment-serve manifest.json --evidence capture.json
+```
+
+Before it binds, the server derives the
+[environment state](web-network-captures.md#environment-state) the capture
+observed: it runs the `environment-state` view on the capture in-process with
+the manifest's `template_rules` in the same order, exactly as
+`inspect-analysis-view` projects it from the same capture, so the state
+Evidence id it reports is the one that projection produces. The view's own
+budget and refusals apply, and the result is checked like the view's Evidence
+for this manifest: extracted from the served capture, with the manifest's
+rules, naming only manifest tools. The server keeps one live state, shared by
+every MCP session and the data plane, starting from the derived state.
+
+A state that is unavailable does not stop the server: it logs a warning naming
+the error `code` and message before it binds, then serves replay. The state is
+unavailable when it cannot be derived, such as from a
+`capture_browser_scenario` capture that selected header values but no response
+bodies or over the view's budget, when `data/get` would write it past the 32 MiB
+`data/add` intake limit, and when its catalog under the served name
+exceeds the [discovery reservation](#what-a-manifest-is). That check runs
+first, so no state is derived for it, and replay is then admitted exactly as
+with `--replay-only`. `--replay-only`
+serves replay without deriving anything; the card, tools and `/agentenv`
+answers are then exactly those of replay. The stopped document's `state`
+member reports which applied:
+
+- `{"served": true, "evidence_id": "ev_…"}`: served from the derived state.
+- `{"served": false, "reason": "replay-only"}`: `--replay-only` was given.
+- `{"served": false, "reason": "unavailable", "code": "unsupported_target", "message": "…"}`:
+  the state is unavailable; `code` and `message` are those of the refusal
+  (`resource_constraint` for a catalog over the reservation), as the warning
+  logged them.
+
+Each read tool answers according to its view in the state:
+
+| View       | Answer                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                    |
+| ---------- | --------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- |
+| `item`     | The entity the path arguments name, rendered into the recorded envelope of `envelope_ordinal`. Arguments match as replay matches them: each collection segment equals the held collection's segment as written or after percent-decoding, and the identity is the last argument percent-decoded, then as written; values are never re-split on `/`. Any entity the state holds answers, including one `data/add` loaded; query arguments only decide `replay.exact_match`. An entity the state does not hold answers the tool's latest recorded 400, 404 or 410 response with `state.view` `missing`, or is a tool error when none was recorded; one lacking a field every recording carried is a tool error, unless a read showed the application leaves that field out; a body is never invented.                                                                                                                                                                                                                                                                                                                                                                                                                                                       |
+| `list`     | When the arguments name one of the lines the view checked exactly (its latest recordings before the first related write; every path and query value equal to the recording's, and every recorded query parameter supplied), that recorded envelope with its array rendered from the state: every entity of the collection in state order for `membership: "all"` or a line `all_record_ordinals` names, or the recorded members still held. A call with no query argument answers from such a line of the collection its path arguments name that only pages its list, with `replay.exact_match` false. On a line the view's `query` answers, and for a call whose every query argument has a role there (an argument-less call aside), the answer is the query evaluated over the state: filtered, searched, sorted and paged, with counters at the number listed and `query.totals` at the number matched, and `state.membership: "query"`; such a call answers on a query or whole line of its collection with `replay.exact_match` false. A value an `inert` parameter never recorded, or a parameter without a role, replays its recording. A member lacking a field every recording carried makes the call a tool error rather than a partial list. |
+| `document` | When the arguments name the document's checked line exactly, the state's value for it; a tool error when the state holds none.                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                            |
+| `recorded` | The recording, as replay answers it.                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                      |
+
+A read with repeated query values uses replay, because the state query model
+does not infer repeated-parameter semantics. A repeated-query write is refused
+before it can change state for the same reason.
+
+An item or list view that always spells a clock field otherwise than the
+entity's stored value names that spelling in `clock_formats`, and a value it
+never recorded, such as a created entity's clock, is rendered in it: a list that
+writes `+00:00` where item reads write `Z` lists a new entity the same way.
+
+A list or document call that names no checked line, including a line recorded
+after a write or with other query values, answers from the recording. Every
+answer has the replay shape plus `state`: `{"source": "state", "view": …}`
+naming the entity, collection or document path, or `{"source": "recording",
+"reason": …}` with the view's reason, `no-view`, `unrecorded-request`, `page-exceeded` (a call naming a page of a whole list that the collection has outgrown) or `projection` (a call to an item tool whose recordings asked for some fields only, naming such a parameter for an entity the state holds; the answer is a recording, so it does not show served writes). A
+state answer carries the base record's status, content type and `replay`
+provenance with `truncated: false`; one whose JSON would exceed 256 KiB is
+refused as a `resource_constraint` tool error rather than cut. Extraction does
+not state-back a document whose rendered JSON exceeds 256 KiB; that read stays
+recorded (`rendered-oversize`) and replays the captured body. Query `POST`s
+stay served and answer from their recordings with the reason `query-body`.
+
+A line establishes whole-collection membership only from the beginning:
+`offset`, `start`, `_start` and `page` must be zero when present. Positive or
+malformed positions do not by themselves establish whole membership. A separate
+query model may describe supported paging; an empty later page cannot start
+returning earlier entities merely because it was empty in the capture.
+
+`POST`, `PUT`, `PATCH` and `DELETE` tools are served only when the state's
+[write plan](web-network-captures.md#environment-state) for them is served:
+the capture's later reads confirmed the effect the plan models. Other write
+tools are not served, so no call claims a change it does not make; each one's
+plan states why in its `reason`. A served write changes the live state, which
+every MCP session and `data/get` then see:
+
+| Effect     | Call                                                                                                                                                                                                                                 |
+| ---------- | ------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------ |
+| `create`   | Adds an entity to the collection the path names, built from the observed entity with each ruled field taken from the call or generated, at the plan's `insert` position. A plan tied to its exemplar's collection writes only there. |
+| `update`   | Sets the sent members on the entity the path names, constants for members accepted only with their recorded value, and the clock, increment, slug, unique and renewed changes the plan models.                                       |
+| `delete`   | Removes the entity the path names.                                                                                                                                                                                                   |
+| `document` | Merges the body into the plan's document object, or replaces it for `PUT`.                                                                                                                                                           |
+
+Its answer has the status and content type of the plan's recorded response,
+with the written entity or its identity rendered in, no body, or the recorded
+body. A call whose body is not an object, sends a member the capture did not
+show stored, sends a fixed member with another value or a member of another
+JSON type, or lacks a required member is a tool error naming the accepted
+members; so is a rendered answer missing a field the recorded response
+carried, and an update or delete of an entity the state does not hold answers
+the recorded 400, 404 or 410 of the tool or of an item view of the type, if
+any. A refused call leaves the state unchanged. Values a write generates are
+deterministic: the clock starts at `write_clock` and advances one second per
+served write, sequences take one past both the largest held value and the
+largest they generated, and UUIDs and tokens derive from the state Evidence
+id, tool, field and write count, skipping held values. `data/reset` and
+`data/add` restart the clock, counter and sequences, so the same calls after a
+reset give the same state. Answers carry `state: {"source": "state", "view":
+"create" | "update" | "delete", …}` or `"document-write"`.
+
+The data plane serves three JSON-RPC methods:
+
+| Method       | Result                                                                                                    |
+| ------------ | --------------------------------------------------------------------------------------------------------- |
+| `data/reset` | `{}` after restoring the state the Evidence holds                                                         |
+| `data/add`   | `{}` after replacing the whole state with one valid `rea-state/v1` document                               |
+| `data/get`   | `{"parts": [{"kind": "data", "data": <rea-state/v1 document>}]}` with `state_evidence_id` and the `types` |
+
+Calls run one at a time in arrival order across requests. Every message of a
+batch is queued when the batch arrives, so another request's call never lands
+between them and each message sees the changes of those before it; a reset
+sent while a download is loading applies after that load. `data/add`
+takes exactly one part: a `data` part whose `data` is the document, or a
+`file` part whose `file` carries `bytes` (base64 of UTF-8 JSON) or a `uri`,
+either a `file://` URL of a local regular file, whose host is empty or
+`localhost`, or an `https://` URL fetched without following redirects and
+within 30 seconds, including reading the body; `mimeType`, when given, is
+`application/json`. A `file://` URL naming another host, or a UNC path such as
+`file:////server/share/state.json`, is refused before anything is opened, so a
+load never reaches a network share. A loaded document is at most 32
+MiB, and so is the state as `data/get` writes it back with the members a
+document may omit (`types`, `state_evidence_id`, record citations), so a
+reported state always loads again; an inline part also stays within the 4 MiB
+request body bound, so larger states arrive by `uri`. Anything else, including `{"parts": []}`, is `-32602
+invalid params`; a download that fails or answers a non-2xx status is
+`-32000`. Errors never echo the uri or the content; a file or download failure
+names its category in `error.data.reason`: `not-found`, `permission`,
+`not-regular-file`, `changed` or `read-failed` for files, `timeout`,
+`http-status` or `network` (connection, TLS or a refused redirect) for
+downloads, and `cancelled` for a download the server's stop or its caller's
+departure ended.
+
+Served creates also bound additional collection history since the last reset
+or `data/add` to 1 MiB of deterministic accounting: UTF-8 bytes for each new
+collection spelling and its possible sequence keys, plus a 64-byte allowance
+per collection and potential sequence-key record. The score bounds identifier
+payload and a fixed record allowance; it is not a measurement or guarantee of
+process memory. Collections
+already present in the Evidence or loaded state are part of the state itself;
+the store indexes their unique spellings so they remain known after deletion,
+but they do not consume this additional-history budget. A create that would
+exceed the budget is refused before the state changes. `data/reset` or `data/add`
+clears generated history; if the same calls would hit the limit again, use
+fewer distinct collection paths. The 1 MiB score ceiling is 1/32 of the
+separate 32 MiB state-intake ceiling, reserving a small, explicit allowance
+for identifiers retained beyond the loaded state. A loaded-state path that is
+later deleted and then introduced again may consume the additional-history
+budget when the store must retain it beyond the source entity.
+
+Stopping the server cancels every pending call, and a caller that leaves before
+its reply cancels its own: a download in progress ends, and calls not yet run
+are answered `-32000` with `error.data.reason` `cancelled` without running. A
+cancelled call leaves the state unchanged, and shutdown never waits for a
+download.
+
+One `/agentenv` response carries at most 64 MiB of replies, twice the 32 MiB
+a loaded document may hold: room for a `data/get` of a loaded state, while a
+batch of reads cannot multiply a large state into one body. Replies are
+measured in order as they are produced, in bounded parts and before they are
+assembled, and room is kept for every later reply's budget error and the
+batch's brackets and commas, so one that would take the response past the
+budget is never built whole and is answered instead with `-32000` whose `error.data` gives
+`max_response_bytes` and the `reply_bytes` it would have taken. That call still
+ran, so repeat such a read in another request. A reply no larger than that
+error, such as a write's `{}`, is always kept.
+
+```json
+{
+  "rea_state": {
+    "version": 1,
+    "entities": [
+      {
+        "type": "/api/users",
+        "collection": "/api/users",
+        "id": "2",
+        "fields": { "id": 2, "login": "probe" }
+      }
+    ],
+    "documents": []
+  }
+}
+```
+
+Each entity's `type` is a state type's `collection_template`, its `collection`
+matches that template segment by segment (a template parameter matches any one
+segment), its identity field's value is its `id`, and its type, collection and
+id are unique; `view_fields` is optional and names item or list views of its
+type. Views render their values over the entity's fields, so a view's values
+must not set the type's identity field to a value whose text differs from the
+entity's `id`. Spell a collection as requests spell it, as `data/get` reports it: list
+calls find entities under the checked line's recorded path. State collection
+identity preserves repeated separators inside captured paths: `/api/books`
+and `/api//books` are distinct. A single trailing slash continues to name the
+same collection. Each document
+names a `document` view once, at its initial document's `path`. The `version`,
+and the `state_evidence_id` and `types` that `data/get` reports, must match
+the served state, so a `data/get` result loads back unchanged. A number without
+an exact native value, or whose serialized spelling would not be read as exact
+(an integer from 2^53 below 1e21, such as `1e20`, in any spelling), or a
+`__proto__` member is refused rather than altered.
+An invalid document is `-32602` naming the first problem by JSON pointer, as in
+`/rea_state/entities/3/collection` (also in `error.data.path`); a problem
+inside caller data (`fields`, `value`, `types`) is located at that member, so
+the pointer never names the caller's keys. A refused document changes nothing;
+a valid one replaces the whole state at once. The order of `entities` is the
+state order lists use.
+
+Record citations stay Evidence provenance. An entity's `record_ordinals` may
+be omitted or empty and a document's `record_ordinal` omitted; otherwise they
+must be the served Evidence's citations for that entity or document, or the
+document is refused. After loading, an entity or document whose values equal the Evidence's cites
+the Evidence's records, and one whose values were edited or added cites none
+(`record_ordinals: []`, no `record_ordinal`), so `data/get` never presents a
+loaded value as recorded.
+
+The card declares the operations and what `data/add` accepts:
+
+```json
+{
+  "extensions": [
+    {
+      "uri": "urn:agentenv:intake/v1",
+      "description": "data/add replaces the whole environment state with one rea-state/v1 document; data/get returns the current state in the same format.",
+      "params": {
+        "add": [
+          { "part": "data", "format": "rea-state/v1", "load": "replace" },
+          {
+            "part": "file",
+            "format": "rea-state/v1",
+            "mimeTypes": ["application/json"],
+            "load": "replace"
+          }
+        ],
+        "get": [{ "part": "data", "format": "rea-state/v1" }]
+      }
+    }
+  ],
+  "operations": ["data/reset", "data/add", "data/get"]
+}
+```
+
+The state starts as a projection of recorded reads and changes only through
+served write tools and the data plane. List and document calls answer from the
+state only for the lines their views checked, and every recorded view answers
+from recordings; item calls answer for any held entity but render only the
+envelope one recording carried. Values copied outside entity objects, derived
+values and counters other than those that count a list stay as recorded, and a
+served write changes only the one entity or document it writes: side effects
+on counters, relations and copies of entities in other collections, reordering
+after an update and referential rules on delete are not modelled. An exported
+bundle carries no state file: its server derives the state at startup, as the
+export does to predict what it serves.
+
+## Export a bundle
+
+`rea environment-export` writes the same pair as a bundle folder the agentenv
+framework runs directly, shaped like its own echo example:
+
+```sh
+rea environment-export manifest.json --evidence capture.json ./grafana-example --json
+agent-env run ./grafana-example
+```
+
+```text
+grafana-example/
+  README.md                                 provenance, what it serves, run command, rights note
+  envs/grafana-example/Dockerfile           pinned image: node:24.22.1-bookworm-slim + rea-agents@<this version>
+  envs/grafana-example/env.toml             environment_name = "grafana-example"
+  envs/grafana-example/manifest.json        byte-exact copy of manifest.json
+  envs/grafana-example/evidence.json        byte-exact copy of capture.json
+  artifacts/check_grafana-example/verify.py the verifier
+  tasks/smoke.json                          deploy_env (local sandbox), then env_outcome_verifier
+```
+
+The command reads each file once, after checking that reading and parsing the
+whole file, including sealing traversal, fits the process's live heap headroom (a file that does not is
+refused as a `resource_constraint` before it is read), then authenticates and cross-checks the
+pair as `environment-serve` does, so a manifest projected from another capture
+is refused with the same `invalid_request` before anything is written, and the
+bytes it copies are the bytes it authenticated. The bundle directory must be
+absent or empty; an occupied directory is refused and nothing in it changes.
+After checking the bundle name, the export prepares the state from the capture
+exactly as the bundle's server will at startup under that name, to predict what
+the bundle serves; an unavailable state, including one whose catalog exceeds
+the discovery reservation, does not stop the export, and the bundle then
+serves replay. The complete tool catalog the bundle serves must also fit the
+server's advertised catalog budget before any files are written. Privately parsed JSON is sealed once so authentication
+and replay share immutable response trees rather than copying the capture repeatedly.
+`--name` replaces the manifest's portable kebab-case name as the env folder and env id
+(Windows device names such as `con` and `lpt1` are refused);
+`--package-tarball <path>` copies a packed `rea-agents` tarball into the env
+folder and makes the Dockerfile install it instead of the registry package and
+remove it in the same layer, which is how a checkout rather than a release is
+run. The command prints the
+written files with their sizes, the pinned package source, the number of tools
+the bundle serves as `tool_count`, and in `state` whether it serves from the
+state, in the shape of the serve command's stopped document.
+
+Export uses the shared output-tree owner to check directory identities around
+exclusive creates, verify written bytes, and reject unowned entries before
+reporting success. It does not delete an inspectable partial folder on failure.
+Node's portable pathname APIs leave a syscall-boundary race, so the output's
+parent directories must be trusted. On POSIX output files use mode `0600`;
+Windows inherits ACLs and REA does not establish owner-only access there. Choose
+an already private Windows destination when captured content needs it.
+
+The Dockerfile sets `MCP_HOST=0.0.0.0` and `MCP_PORT=18765`, the port the
+framework maps for an MCP server env; the framework's own deploy supplies
+`MCP_HOST`, `MCP_PORT` and `ENVIRONMENT_NAME`, so the served card carries the
+env's name. It installs Debian's `python3` because the framework's local
+sandbox probes a container's readiness with `python3` inside it, which the slim
+Node image lacks. It copies the two inputs owned by the image's non-root `node`
+user, so a private host umask cannot leave them unreadable, runs as that user and starts
+`rea environment-serve /env/manifest.json --evidence /env/evidence.json`. The
+`env.toml` names the environment because the framework otherwise reads the
+name from Python sources the image does not have.
+
+`verify.py` is the framework's `async def verify(mcp_url)` shape over its
+Python MCP client: it lists the tools and checks the count equals the number
+the bundle serves (from the state, the manifest's read tools and the write
+tools whose effect the capture confirmed; every manifest tool in replay), calls one tool with no arguments and checks the call did not
+fail and its `status` is the expected one, and returns one `serve` result
+whose description reports both checks. The tool is the first served `GET` tool
+without path parameters whose answer to no arguments is not an error, and the
+expected status is the status of that answer: the export calls the same answer
+function the server uses, from the derived state or by replay, so a state
+answer carries its base record's status and a replayed one its record's (an
+unknown recorded status stays unknown; it is not borrowed from another
+record). A bundle with no such tool yields a verifier that checks the count
+only and says so. `tasks/smoke.json` deploys the env on the local
+sandbox and runs that verifier against the env's MCP endpoint; the first run
+also builds the framework's own infrastructure, and the run tears everything
+down when it ends.
+
+The README written into the bundle records the capture artifact's SHA-256, the
+capture and manifest Evidence ids, the manifest view digest, the served tool
+count, the derived state's Evidence id or that no state is served, and the
+package source. It says what the environment serves: read tools from the state
+the capture observed, the `/agentenv` data plane on it and the write tools
+whose effect the capture's later reads confirmed, replay when the state is
+unavailable, and `--replay-only` for pure replay; it
+says which applies to this bundle, with the error code and message when the
+state is unavailable, and that neither is emulation and served
+content is untrusted data. It states that the recorded content belongs to the
+captured application and must not be published without the rights to it.
+
+## Measure fidelity against the application
+
+An environment agrees with its own capture by construction. What matters to an agent is how it answers the calls the
+capture did not make, what it shows after a write, and how it fails. `rea environment-compare` measures that against the
+application itself:
+
+```sh
+rea environment-compare manifest.json --evidence capture.json \
+  --probe probe-crud.json --probe probe-reads.json --json
+```
+
+Each `--probe` is capture Evidence (`inspect-web-network-capture` of a HAR, or `capture-browser-scenario`) of one call
+sequence made against the live application, seeded as it was for the capture and reset before the sequence. The command
+makes no network call: whoever records the probes drives the application with its own credentials. For each probe it
+resets the environment's state, takes the requests in the order they were made (record order when any lacks a start time), matches each to the manifest tool of the
+same method whose path template fits with the most literal segments (a format-rule parameter must fit its format once
+percent-decoded). A path argument is the raw segment, which the environment decodes once.
+It answers the call as the served environment would, and compares that answer with the application's.
+
+A re-seeded application holds the same entities under other ids, so the comparison maps them. Objects in the two answers
+to one call pair by a natural key, a string field unique on each side that both share, such as a name or a slug; their
+identifier fields (`id`, `uuid`, `guid`, and names ending in `_id`, `Id`, `ID`, `_uuid` or `Uuid`), and those of two single
+answers such as a create's, map the live value to the environment's, including a create that returns only the new id.
+Ids are learned only from an answer the environment
+gave that call itself, from its state or a recording of that very call, request line and body, never from another call's
+recording replayed in its place and never from a recording whose body the probe omitted. The answer's own `id`, `uuid` or `guid` also maps within the collection the call names, since
+two collections can reuse a live id, and an equal pair stays in that collection. A live id in a later request's path, query or body is translated before the
+environment answers, a path parameter by the collection the live path names (a nested collection keeps its parent values, each decoded once, and a literal route segment stays), except a query value that pages a list, such as `limit`
+or `page`. Each call is checked against the served tool's input schema first, as the MCP server checks it. After mapping, two clock values compare by format only (ISO-8601 strings, and epoch numbers under
+`at`, `time`, `date`, `timestamp`, `created`, `updated` or `modified` field names, including snake-case and camel-case forms), and two different values each side generated on its own compare equal: both UUIDs, or one length of
+eight or more with the same character classes and a digit. The report counts each kind of normalised value.
+ISO clock offsets are accepted only within RFC 3339's `00`–`23` hour and `00`–`59` minute ranges.
+Timestamp ordering retains submillisecond precision when the source records it.
+
+Comparison keeps response representations distinct: JSON values compare structurally, text compares as text, and a
+retained base64 body compares by bytes (including with UTF-8 text carrying those same bytes), never by its base64
+spelling. ID mappings apply at identifier fields, `ids`/`uuids`/`guids` lists, an entity's own identity within its
+collection, and a string field learning confirmed as a state-held ID. ID mappings leave other numeric and string values unchanged. A
+global conflict involving an equal pair is not used to normalize either pair; the report counts it as an ID conflict.
+
+A step agrees when the statuses and the normalised bodies are equal. The report gives agreement overall, per probe and
+per tool, `agreement_ignoring_order` for steps that differ only in item order, and each step's differences, at most 20,
+with a JSON pointer and both values cut short:
+
+| Category        | Meaning                                                                                                      |
+| --------------- | ------------------------------------------------------------------------------------------------------------ |
+| `status`        | The statuses differ                                                                                          |
+| `missing-field` | The application's object has a member the environment's lacks                                                |
+| `extra-field`   | The environment's object has a member the application's lacks                                                |
+| `value`         | Values differ, or one is an object or array and the other is not                                             |
+| `length`        | Arrays differ in length, or not every item pairs                                                             |
+| `order`         | Paired items come in a different order                                                                       |
+| `clock-format`  | Both are clocks, in different formats                                                                        |
+| `refused`       | The environment refused the call                                                                             |
+| `not-served`    | The request's tool is one the environment does not offer, such as a withheld write                           |
+| `no-tool`       | No manifest tool matches the request                                                                         |
+| `unmapped-id`   | The application answered with success and the environment did not, for a path id the comparison never mapped |
+
+What a probe did not observe is not counted as the application's behaviour. A record without a usable request line,
+such as a redacted URL or a transaction the capture cannot attribute, is not replayed, nor is a request to another
+origin than the application's (`other-origin`: the probe's origin with the most JSON-like records, as the manifest
+chooses its own), and a request with no recorded
+answer is made, since it may have changed the application, but not compared: each probe lists them under `unobserved` with the reason, and the summary counts them. A live
+body the probe did not record compares by status only and is counted as `unobserved_bodies`, unless the answer never
+carries one (204, 205, 304, `HEAD`). A browser probe captured without `capture.network.response_body` is refused, as
+is one that makes POST, PUT or PATCH calls without `capture.network.request_body`.
+
+Agreement covers only the calls the probes make, after the mapping the answers allow. Values compared by form, clocks
+and generated values, can still differ in content, and an equal answer after a write shows the state matched for that
+read, not that the write had no other effect.
+
+## Compatibility verification
+
+`npm run verify:agentenv:environment` is the hosted `real-agentenv` lane
+described in [testing](testing.md#exported-environment-bundles). It projects
+the Grafana-shaped source-owned capture through the built CLI, packs the
+checkout, exports a bundle with `--package-tarball`, runs it with the pinned
+`agentenv-framework` release on the runner's Docker and requires the smoke task
+to pass with no container left behind other than the framework's own
+image-registry helper, which the receipt reports. It proves that one exported bundle runs
+on one framework release; it does not claim any hosted platform or other
+framework versions.
+
+## Resource and lifecycle boundaries
+
+The Evidence files are read once at startup through the bounded JSON input
+reader and authenticated as Evidence before the listener binds; a file that is
+not Evidence, a manifest of another view kind, a capture of another format, or
+a manifest projected from a different capture is refused before any port is
+opened. The state is derived once, in-process and before binding, from the
+already authenticated capture: it is projected as retained Evidence, so the
+capture's strings are not charged again as inline view input, and the view's
+96 MiB budget bounds the derivation. Because authentication covers the capture's records, a manifest whose
+parent check passes cites only records the capture retains. The server keeps
+only the cited records' request lines and response text in memory, plus the
+live state when it serves from one: the derived initial state, or one
+validated document whose JSON was at most 32 MiB. Data-plane calls run one at a time, so
+at most one loaded document is being read and parsed at once; `data/get`
+serializes the held state whole. No process is launched and nothing is
+written to disk. No outbound network request is made unless a `data/add` call names an
+`https://` uri, which is fetched once within the limits above; a `file://` uri
+is read once. A listener that fails after binding is logged and ends the command with
+a `resource_constraint` error and exit `1` instead of serving nothing.

docs/android-analysis.md

@@ -103,9 +103,10 @@ that every class or engine operation will load successfully. Discovery limits
 each ZIP metadata read to 8 MiB before allocation; a larger directory reports
 the exhausted read budget and the selected JAR path.
 
-The current metadata bridge is verified on macOS arm64 with OpenJDK 21 and the
-public Appium ApiDemos fixture. The POSIX adapter also supports Linux; the new
-bridge has not yet undergone real verification on Linux/JDK 25.
+The filtered class-page bridge (protocol 2) is verified on Linux x86-64 with
+Temurin 21, public Appium ApiDemos 6.0.18 and NewPipe 0.29.1 APKs. Earlier
+metadata-bridge records covered macOS arm64/OpenJDK 21 and Windows x64/JDK 21;
+this protocol revision has not been rerun on those hosts or Linux/JDK 25.
 Windows x64 uses an owned Job Object with bidirectional stdio, atomic job
 assignment and kill-on-owner-close. Real Windows verification uses JDK 21 and
 the pinned JADX 0.7.1 JAR; see the real verification lane below.
@@ -146,10 +147,20 @@ queries accept the same `overload_index`; the bridge resolves the selected
 method node and the result reports its index and displayed signature. These
 indices are tied to the current artifact and engine, not stable across builds.
 
-The engine's smali fallback also joins same-name overloads; REA rejects that
-fallback for an overloaded method instead of attributing all bodies to one
-selected overload. Native/abstract methods with no decompiled body report
-`body_status: "not_available"` without inferring which implementation flag caused it.
+Smali fallbacks select one complete block using the original method name and DEX
+descriptor, before applying the source byte limit. This preserves the selected
+overload even when Java code generation renames a method. If the selected Java
+body was omitted, including a method merged into a synthetic bridge, REA returns
+its original smali with the `java_body_unavailable` marker. The marker describes
+missing Java; it does not infer why the method was omitted or name a replacement.
+Unresolved or duplicate disassembly identities fail instead of attributing another
+method's instructions to the selection.
+
+Native and abstract methods are identified from access flags
+captured before code generation. They report `body_status: "not_available"`, an
+empty source, and the `native_method` or `abstract_method` marker; their declarations
+are not passed through the engine's source slicer. Other missing decompiled bodies
+do not establish either access flag.
 
 Static initializers (`<clinit>`) return smali with the
 `static_initializer_disassembly` marker. Java decompilation can move their
@@ -160,10 +171,19 @@ the same source byte limit and explicit truncation reporting.
 ## Interpret the result
 
 - Package inspection retains decoded manifest XML, summary declarations,
-  permissions and class/resource counts. Summary fields are extracted by the
-  upstream engine and should be checked against the retained XML when needed.
+  permissions and class/resource counts. REA's bridge selects the exact root
+  `AndroidManifest.xml` resource by original archive name and manifest type,
+  independently of asset order. One namespace-aware JDK XML parse supplies
+  summary attributes and direct permission/SDK declarations; namespace prefixes
+  and XML entity spelling do not change their values. Raw decoded XML is retained
+  exactly and only its displayed text is subject to the existing source byte cap.
+  Summaries use the complete selected XML before that cap. Document types are
+  unsupported; parsing does not load external DTDs, schemas, or XInclude files.
   Signature verification is explicitly `not_performed`.
-- Class search consumes all pages and checks their counts. An empty `query`
+- Class search filters the engine's cached top-level names before transferring
+  pages, then follows contiguous inventory positions without rescanning earlier
+  rows. It checks selection, scope, examined ranges and counts before reporting
+  complete coverage. Complete producer replies stay in Evidence; an empty `query`
   inventories all class-name rows; class names are decompiler representations.
   Repeated names remain repeated in `matches` so the inventory row count and
   producer evidence stay visible. `duplicate_names` identifies matching names
@@ -233,11 +253,19 @@ rea project-android-application-graph '{"inventory_evidence":[<inventory_artifac
 ```
 
 The projection reports exact component paths and hashes, runtime-family hints,
-and path-based bridge hypotheses. It does not decode DEX or claim observed
-runtime calls. Framework hints match complete native library names, such as
+and path-based bridge hypotheses. File and slice occurrences supply component
+endpoints; directory trees and symlinks retain their identities in the source
+inventory without being classified as code files. It does not decode DEX or
+claim observed runtime calls. DEX/JVM runtime hints summarize the classified
+bytecode component lists. Framework hints match complete native library names,
+such as
 `libreactnativejni.so`, `libreactnative.so`, `libhermes.so`,
 `libhermes_executor.so`, `libflutter.so`, and `libunity.so`; unrelated names
 such as `libcommunity.so` and `libreactivity.so` do not identify those families.
+The signing list retains the `META-INF/MANIFEST.MF`, `.SF`, `.RSA`, `.DSA`, and
+`.EC` paths selected by Android's v1 verifier, including nested paths. These are
+inventory metadata candidates, not proof that a v1 signature verifies or that
+the APK is signed; use `inspect-apk-signature` for signature analysis.
 The bridge basis is inferred from the native path and repeated for each managed
 component, so it does not identify a specific JNI binding. Bridge hypotheses form a DEX/JVM-class by native-library
 Cartesian product. The projection computes the product and exact serialized
@@ -299,9 +327,37 @@ settings. Evidence reports the worker's observed `Runtime.maxMemory()` rounded
 down to MiB and retains exact bytes, Java version and visible processors in its
 raw runtime response. The observed allocatable heap can differ from configured
 `-Xmx`, especially with Serial GC. Heap is not a total RSS cap: the JVM also
-uses native memory. Independent REA instances have independent queues. Whole
-operations have a **120-second** execution deadline after reaching the queue's
-front; upstream decompilation is limited to **90 seconds**.
+uses native memory. A typed JVM memory allocation failure during APK loading
+reports `resource_constraint` with the observed maximum heap in bytes. Known
+heap failures name `REA_JADX_HEAP_MIB` in remediation; other memory regions
+remain unknown. A healthy doctor configuration check does not establish that
+any particular APK fits. Failures before the bridge can report a typed result
+and upstream errors from other operations retain their existing diagnostics;
+error text alone does not establish memory exhaustion.
+
+Independent REA instances have independent queues. Whole
+operations have a **120-second** execution deadline by default, starting after
+they reach the queue's front. For a known slow cold load, such as a large
+multidex APK, select another finite deadline:
+
+```sh
+export REA_JADX_OPERATION_TIMEOUT_MS=600000
+```
+
+The value must be a decimal millisecond count from 1 to 2147483647; anything
+else is refused before Java starts. REA reads it from the environment captured
+at process start. A change applies on the next process start; an already
+running server keeps the deadline it started with. The same deadline bounds
+each upstream call. A timeout names `REA_JADX_OPERATION_TIMEOUT_MS`, the
+effective milliseconds, and that range. A passing JADX configuration check in
+`rea doctor` reports the effective `operation_timeout_ms`. Explicit
+cancellation still ends the operation immediately, and partial observations
+and owned-process cleanup are unchanged. Raising the deadline does not extend
+the bridge's 30-second startup handshake or the upstream **90-second**
+decompilation limit. MCP clients enforce their own tool-call timeouts, and
+setup's derived [client tool timeout](client-configuration.md) does not include
+this setting, so also raise the client timeout when selecting a longer
+deadline.
 
 Decoded manifest and method text have a **1 MiB** upstream byte budget and report
 truncation. Class-name pages use indexed slices of up to 4,096 names without changing the
@@ -320,6 +376,14 @@ export REA_ANDROID_TEST_APK="$PWD/_reference/apk-integration/ApiDemos-debug.apk"
 npm run verify:android
 ```
 
+The main lane and Windows Java-selection checks validate the selected JDK and
+engine archive through `validateVerifierJadxConfiguration`. This shared verifier
+boundary supplies the production validator's resource owner and waits for
+retained cleanup before proceeding. Failed closes retain their exact owners for
+retry; prerequisite failures must not be mistaken for completed bridge tests.
+Recovery checks match the typed Java or archive failure and its applicable
+remediation.
+
 The lane also uses the selected full JDK's `jlink` to create a disposable JRE
 without the compiler module. It checks real Java startup failures, missing
 selectors, recovery after failures, and CLI cancellation with owned cleanup.

docs/apk-signature.md

@@ -0,0 +1,268 @@
+# APK signature schemes
+
+REA reads the signing of a selected local Android APK (or ZIP/AAB) statically,
+so an analyst can answer **who signed this, under which schemes, and whether the
+declared content digests match these bytes** before trusting or triaging it. The reader
+is in-process and needs no external engine: it parses the ZIP tail and the APK
+Signing Block, reads certificates with the platform X.509 parser, and
+recomputes the content digest. No Android code executes, and no device or
+`adb` is involved.
+
+CLI and MCP share one workflow and return artifact-bound Evidence inline.
+
+```sh
+rea inspect-apk-signature /path/app.apk
+```
+
+## Selected facts and retained Evidence
+
+MCP defaults to a summary with scheme presence, signer certificate fingerprints,
+row counts, posture and explicit coverage. `detail: "complete"` returns the full
+report. A selected page can answer a narrower question in the acquisition call:
+
+```json
+{
+  "path": "/path/app.apk",
+  "view": {
+    "collection": "signers",
+    "schemes": ["v3"],
+    "fields": ["certificate_sha256"],
+    "limit": 32
+  }
+}
+```
+
+The returned `parent_evidence_id` identifies the retained complete report.
+Subsequent reads use `inspect_analysis_view`, including after the APK is deleted:
+
+```json
+{
+  "source": {
+    "kind": "retained-evidence",
+    "evidence_id": "<returned parent_evidence_id>"
+  },
+  "view": {
+    "kind": "apk-signature",
+    "collection": "digests",
+    "signer_ids": ["<returned signer id>"],
+    "fields": ["signature_algorithm_id", "algorithm", "matches_apk"],
+    "offset": 0,
+    "limit": 32
+  }
+}
+```
+
+Collections are `signers`, `digests` and declared `lineage` nodes. `ids`,
+`signer_ids` and `schemes` filters intersect; omitted filters select all available
+records and `[]` selects none. Omitted `fields` returns every collection column;
+`fields: []` retains identity and relationship links. IDs include the complete
+APK digest, scheme and source ordinal; child records retain `signer_id` and
+`signer_index`, and declared certificates retain their raw-DER SHA-256 identity.
+Lineage pages expose `previous_id` when selected. Certificate fingerprints can
+be composed across versions without treating a declaration as verified identity.
+
+Page coverage describes examined declared records. Exhausting a page does not
+establish certificate authenticity, platform acceptance, v1 entry coverage or a
+verified rotation. Partial decoding and its reason remain visible even when a
+lineage page contains no rows. A weak cache reuses validated immutable report
+facts; each read checks its parent subject, and new file inspections read again.
+
+CLI defaults to complete delivery. A compact one-shot response requires a complete
+Evidence export:
+
+```sh
+rea inspect-apk-signature /path/app.apk --detail summary \
+  --view '{"collection":"signers","fields":["certificate_sha256"]}' \
+  --evidence_output /path/app-signing-evidence.json --json
+```
+
+The export contains the same complete parent used by MCP. Historical CLI pages
+use `inspect-analysis-view` with an `evidence-file` source record. The CLI and MCP
+share the application workflow, projection, retention and publication owners.
+
+## What it reports
+
+- **Schemes present:** v1 (JAR signing, from the central directory), and the
+  APK Signing Block's v2, v3 and v3.1. v1 follows apksig's selection: entries
+  under `META-INF/` at any depth ending in `.SF`, or in `.RSA`/`.DSA`/`.EC`
+  (case-sensitive). v1 is `present` only when a signature block has its
+  matching `.SF` file (`META-INF/x/CERT.RSA` pairs with `META-INF/x/CERT.SF`);
+  `signature_files` and `block_files` list every candidate either way.
+- **Per v1 signer** (each block paired with its `.SF`): REA inflates the block
+  and reads its PKCS#7 `SignedData`. `certificate` is the certificate the
+  block's one `SignerInfo` names by `issuerAndSerialNumber` or
+  `subjectKeyIdentifier` (not by position), with the same facts and raw-DER
+  SHA-256 as a v2/v3 signer, so a v1-only APK can be fingerprinted and
+  clustered with v2/v3-signed ones. apksig resolves only
+  `issuerAndSerialNumber`, so its v1 verification fails for a block whose
+  `SignerInfo` uses `subjectKeyIdentifier`: REA still reports the declared
+  certificate (`identified_by: subject-key-identifier`), adds a limitation,
+  and treats that signer as unknown in `v1_signers_match_block_signers`. `certificates` lists every certificate the
+  block carries (SHA-256 and subject) and `signer_infos` gives each
+  `SignerInfo`'s certificate and digest/signature algorithm OIDs. The PKCS#7
+  `SignedData` envelope is read as BER, including constructed signature OCTET
+  STRINGs; each embedded certificate is still interpreted as DER by the X.509
+  owner.
+  `certificate` is `null` when the block has more than one `SignerInfo` (with
+  nothing verified there is no single signer) or names a certificate it does
+  not carry. A block that cannot be read (malformed ASN.1, an unsupported
+  compression method, Zip64 sizes, a local header that disagrees with the
+  central directory on the name, data-descriptor flag, CRC-32 or a size, as
+  apksig refuses it, or over a bound) keeps its names and carries
+  `decode_error`; it is never reported as an absent signer. apksig ignores
+  the ZIP encryption flag (bit 0), so a block that sets it is read as stored:
+  the PKCS#7 decoding decides (bytes that really are encrypted fail to
+  decode), and a limitation names the block.
+  apksig rejects the v1 signature of an archive whose central directory
+  holds any name more than once ("Duplicate entry"). REA checks every entry:
+  a limitation names the first few duplicates (JSON-quoted; a name over 256
+  bytes is cut there and its stored length given) and counts the rest, and a
+  signer whose block, `.SF` or `META-INF/MANIFEST.MF` name is duplicated
+  (one signer per block name) carries `decode_error`
+  `duplicate entry: <name>`, so a duplicated manifest marks every signer.
+  A duplicate outside those names leaves the signers as declared.
+  Names are compared as apksig compares them, as Java strings decoded from
+  UTF-8: malformed bytes become U+FFFD, so `dup\xff` and `dup\xfe` are one
+  name (apksigner: `Duplicate entry: dup�`). REA reproduces Java's
+  replacement rather than WHATWG `TextDecoder`'s; the two differ for encoded
+  surrogates (`ED A0..BF`), which Java replaces as one unit. This was checked
+  against OpenJDK 25's `new String(bytes, UTF_8)` on about 300,000 malformed
+  and well-formed sequences, including a leading byte-order mark, which Java
+  keeps; other Java runtimes are not checked. The same
+  decoding names v1 entries and compares a local header's name with the
+  central directory's; as in apksig, a local name longer in bytes than the
+  directory's is refused before decoding. apksig takes the record count from the classic
+  end-of-central-directory record's 16-bit field and ignores a Zip64 count,
+  so it reads only the first 65,535 records of a larger directory: the
+  limitation then says so, counts as duplicates for apksig only names whose
+  second copy lies within those records, and names separately those
+  duplicated only past them. REA still reads v1 signing entries, and flags
+  duplicated signing names, from every record, so past record 65,535 it
+  reports signers apksig would not see.
+  Bytes after the block's `ContentInfo` are ignored, as apksig's BER parser
+  ignores them, and a limitation names the block and their count.
+- **Per signer:** the certificate's **SHA-256 fingerprint over the raw embedded
+  DER** (the same value `apksigner --print-certs` prints), plus subject, issuer,
+  serial, ISO 8601 validity and public-key algorithm when the X.509 parser can
+  read them (each is `null` otherwise); raw signature algorithm IDs, declared
+  digest hex and digest algorithms, signature lengths and public-key length; and for
+  v3/v3.1 the signer's declared min/max SDK.
+- **Key rotation (v3/v3.1):** the declared proof-of-rotation lineage, oldest
+  first. Each node carries its certificate and the capability flags **retained
+  by that node's own certificate** after rotation (installed data, shared UID,
+  permission, rollback, auth). A lineage attribute that cannot be decoded is
+  reported with `decode_error`, never silently emptied.
+- **Source stamp:** whether a source-stamp block is present and the stamp
+  certificate it declares (the V2 stamp block is preferred over V1, as apksig
+  does). `certificate` is `null` when no stamp is present, or when the block
+  could not be decoded and `decode_error` says why. Stamp signatures are not
+  verified.
+- **Declared digest consistency:** for SHA-256/SHA-512 digests, REA recomputes
+  the v2/v3 content digest and reports `matches_apk` per digest and
+  `posture.self_consistent`. This only shows that the **declared** digest
+  equals the digest of the current bytes. It is **not a signature
+  verification**: anyone who changes the bytes can recompute the digest and
+  write it into the signed data, and REA still reports `true` while
+  `apksigner verify` rejects the APK. A changed byte with an unchanged
+  declared digest reads `false`.
+- **Posture** (derived from the facts above):
+  - `v1_only`: v1 is present and no v2/v3/v3.1 block is. The v1
+    `X-Android-APK-Signed` marker is not read, so this cannot distinguish an
+    APK signed only with v1 from one whose v2/v3 block was removed.
+  - `schemes_agree_on_signer` (observed): every present block scheme presents
+    the same _set_ of signer certificates. An extra signer in one scheme is
+    disagreement; a signer without a certificate makes it `null` (unknown).
+    Build-tools 33+ rotate through v3.1 by default, so a rotated app usually
+    reads `false` here and `true` below.
+  - `signer_change_declared_by_rotation` (declared): the signers differ, but
+    each appears in a lineage the APK declares. Read the two together: a
+    rotated APK shows `false` / `true`; signers that differ with no declared
+    lineage containing them show `false` / `false`. It is `null` when a
+    signer's identity is unknown, or when the signers differ and a declared
+    lineage could not be decoded.
+  - `v1_signers_match_block_signers` (observed): the v1 signer certificates
+    equal what the lowest block scheme that has signers (v2, else v3, else
+    v3.1) expects v1 to carry. That is its signer set, except that a v3/v3.1
+    signer declaring a rotation lineage expects the lineage's first
+    (original) certificate, as apksig requires. `false` is notable: the JAR
+    signature names a different key than the block signature. A key rotation
+    keeps v1 (and v2) on the original key, so it reads `true` while
+    `schemes_agree_on_signer` reads `false`, with or without v2. `null` when
+    v1 or a block scheme is absent, any signer's identity is unknown (a
+    `decode_error`, a `null` certificate or a `subjectKeyIdentifier`
+    `SignerInfo`), or the lineage's original certificate cannot be
+    identified; a limitation then says why.
+  - `strongest_scheme`, `rotation_present`, `source_stamp_present`,
+    `self_consistent`.
+
+## What it does not claim
+
+This is **declared identity and digest consistency, not trust**. No signature
+is cryptographically verified (including the v1 PKCS#7 signature and the
+`.SF`/`MANIFEST.MF` digests, so a v1 signer is reported even where `keytool`
+or `apksigner` reject the archive), certificate chains and revocation are not
+evaluated, and nothing here says an app is authentic. `self_consistent: true`
+means the declared digest equals the digest of the current bytes; it is not a
+signature verification, and it does not prove who made the bytes. Rotation-lineage node signatures are not verified either, so a lineage is
+what the APK _declares_: an attacker who re-signs an APK can declare any
+lineage, which is why `signer_change_declared_by_rotation` is kept apart from
+the observed agreement. SHA-256 identifies exact bytes, not authenticity.
+
+Not yet read: v4 `.apk.idsig` (a sibling file), verity (`0x0421`/`0x0423`/`0x0425`) digests (reported with
+`matches_apk: null`), and the v1 `X-Android-APK-Signed` marker.
+
+## Bounds and errors
+
+The archive is read once, bounded at 512 MiB, without following symlinks and
+with change detection. Every size and offset in the ZIP tail and signing block
+is checked against its containing section before use, including the complete
+central-directory extent before ZIP end records. A process-wide queue permits
+one active whole-file inspection and eight pending scans across provider
+instances; pending scans retain no APK bytes and can be cancelled with their
+original reason. Further requests receive a capacity error. The active slot
+is held through parsing and cleanup. Signers per scheme, v1 signers and
+lineage nodes are capped at 32 each as resource admission limits, not format
+maxima (apksig separately accepts at most 10 v1 signers). Each v1 block is
+inflated to at most 1 MiB, and its BER is read within 32 levels of nesting and
+65,536 elements, with at most 32 certificates and 32 `SignerInfo`s; BER
+indefinite lengths and constructed OCTET STRING fields are accepted. Embedded
+X.509 certificates retain their DER requirement. The report's size is
+bounded from above by those counts before anything is materialised, and the exact
+serialized size of the report together with its limitations (which repeat
+archive names) is then checked against an 8 MiB output budget (the exact check
+is authoritative); an over-budget report is refused, not truncated.
+
+The whole-file reader retains read chunks and a concatenated buffer; total
+process memory exceeds the selected file's size. It also computes the complete
+file SHA-256, so inspection is not header-only. Content digests are recomputed
+at most once per SHA-256/SHA-512 algorithm during one inspection. Input,
+inflation, decoded-record and serialized-output bounds keep work finite;
+resource refusals are distinct from malformed format errors.
+
+A file that is not a ZIP is refused as an unsupported target. A damaged ZIP
+(local headers but no end-of-central-directory), a central directory that
+cannot be walked or whose declared record count does not account for all of
+its bytes, or a malformed signing block or signer is an input error naming the
+scheme and reason. Signer framing includes signature records, the public-key
+field and matching signed/outer v3 SDK ranges. Suffix bytes beyond known
+fields are counted as uninterpreted, as Android readers may ignore extensions;
+their presence does not establish malformed signing.
+When the central directory is not immediately followed by
+the end-of-central-directory record (a declared-size mismatch or a Zip64
+record), the content digest is not recomputed and a limitation says so. A
+rotation attribute that cannot be decoded still counts as `rotation_present`
+(the attribute exists) and carries its `decode_error`.
+
+Verified by the `verify:apk-signature` real lane. It builds a real minimal APK
+with `aapt2`, signs it with the SDK's `apksigner` (default v2+v3, v1-only with
+RSA and EC keys, v1+v2+v3 on a legacy target, a rotated key lineage with and
+without v1, rotation carried by v3.1, a source stamp, and a copy
+with one changed byte), and cross-checks REA against
+`apksigner verify --print-certs`, `apksigner lineage --print-certs` and
+`keytool -printcert -jarfile` through the CLI and MCP.
+
+The real-toolchain CI lane uses pinned, checksum-verified Build Tools 36.1.0
+and platform 36 oracle files in its job workspace. It requires the existing
+host JDK and runs `npm run verify:apk-signature -- --require-tools`; absent
+oracles fail rather than silently establishing provider coverage. REA itself
+requires none of these external tools.

docs/apktool-resource-analysis.md

@@ -8,7 +8,7 @@ installs Apktool or a Java runtime.
 
 ## Supply tools
 
-Use Apktool 2.7.0 or newer (verified with 2.7.0) and a Java runtime the
+Use Apktool 2.7.0 or newer (verified with 2.7.0 and 3.0.3 on Linux) and a Java runtime the
 launcher can find. Select the launcher through the environment:
 
 ```sh
@@ -24,6 +24,30 @@ missing launcher reports `capability_unavailable` with the
 `binary_session` uses the same host probe for tool availability; it does not
 validate a particular APK or guarantee that the selected input can be decoded.
 
+Client inspection and decode results disclose the following reported-version
+advisory windows without refusing otherwise compatible launchers:
+
+| Advisory                                                                                              | Reported version window                          | Host scope |
+| ----------------------------------------------------------------------------------------------------- | ------------------------------------------------ | ---------- |
+| [CVE-2024-21633](https://github.com/iBotPeaches/Apktool/security/advisories/GHSA-2hqv-2xv4-5h5w)      | Before 2.9.2                                     | All        |
+| [GHSA-vgwr-4w3p-xmjv](https://github.com/iBotPeaches/Apktool/security/advisories/GHSA-vgwr-4w3p-xmjv) | Before 2.9.3                                     | Windows    |
+| [CVE-2026-39973](https://github.com/iBotPeaches/Apktool/security/advisories/GHSA-m8mh-x359-vm8m)      | 3.0.0 release line through versions before 3.0.2 | All        |
+
+The 2026 advisory names 3.0.1 and its 3.0.2 fix. Inclusion of 3.0.0 is
+source-derived: [tagged 3.0.0](https://github.com/iBotPeaches/Apktool/blob/f770b39e425af4fa6ac4ca1e9ea6c4824339c6d3/brut.apktool/apktool-lib/src/main/java/brut/androlib/res/table/ResTypeSpec.java)
+passes type names through unchanged; the [3.0.2 patch](https://github.com/iBotPeaches/Apktool/commit/65dd8480dfcb63068562ffaa527f71bb0a9f772c)
+validates those names at their owner before they reach the output-path producer.
+Prereleases before the fixed release are conservatively disclosed; the exact
+`-dirty` downstream marker is compared as a release while preserving its spelling.
+This is version guidance, not verification of a particular prerelease or decoder.
+
+A launcher version does not prove the decoder's contents, execution platform,
+downstream backports, or overall safety. Windows applicability uses the selected
+local host; an opaque wrapper may run another engine. A missing warning means
+only that none of these version/host windows matched. Verify the selected
+decoder's patches before decoding untrusted APKs; the compatibility floor above
+is not a security recommendation. REA does not fetch advisories at runtime.
+
 On Windows, a `.bat` or `.cmd` launcher (such as Chocolatey's `apktool.bat`)
 runs through an explicit `cmd.exe /d /s /c` call. That call has no AutoRun,
 no delayed expansion, and quotes every argument. A launcher path that
@@ -38,30 +62,75 @@ reports `capability_unavailable`.
 rea inspect-apktool-client
 rea decode-android-resources /targets/Example.apk
 rea decode-android-resources /targets/Example.apk --locale de
+rea decode-android-resources /targets/Example.apk --locales de --locales pt-BR
+rea decode-android-resources /targets/Example.apk --locales de --locales pt-BR --string-tables-only
 rea decode-android-resources /targets/Example.apk --no-strings
+rea decode-android-resources /targets/Example.apk --decode-timeout-ms 900000
 ```
 
 Equivalent MCP methods are `inspect_apktool_client` and
 `decode_android_resources`. The decode request takes `path`, an optional
-`locale` (project `res/values-<locale>/strings.xml` instead of the default
-table), and `include_strings` (default true). Regional locales such as `pt-BR`
-select Android's `values-pt-rBR`; Android qualifiers such as `pt-rBR` and
-`b+zh+Hans+CN` may also be supplied directly.
+primary `locale` (project `res/values-<locale>/strings.xml` instead of the default
+table), `locales` (repeatable array of additional localized string tables to project
+from the same single decode pass), and `include_strings` (default true). Regional locales such as `pt-BR`
+select Android's `values-pt-rBR`. A three-digit region such as `es-419`
+selects `values-b+es+419`; `es-r419` is rejected because Android's legacy
+`-r` form requires a two-letter region. Android qualifiers such as `pt-rBR`
+and `b+zh+Hans+CN` may also be supplied directly.
+
+Set `string_tables_only: true` (CLI `--string-tables-only`) to return string
+entries only inside `string_tables`, including when selecting a single table.
+The top-level `strings` array is then empty by request; it does not establish
+that the primary table is empty. Table status, counts and coverage still describe
+the examined source. This avoids repeating primary entries in multi-locale
+responses and charges the aggregate output budget for one copy per table.
+Omitting the option preserves existing responses. `include_strings: false`
+still suppresses all string projection, including in this mode.
 
 ## Interpret the result
 
 - **An owned APK snapshot is digested and decoded.** `target.sha256` and
   `target.bytes` bind the observation to the exact APK that was decoded.
+  The Evidence subject carries the same snapshot digest and caller-selected
+  path, including when the original file changes during decoding.
 - **Metadata comes from apktool.yml as apktool printed it.** Version codes
   keep their own spelling (`'7'` versus `7`); `package_name` is read from
   the decoded manifest, not inferred.
+  Apktool 3's emitted `version: 3.x.y` header selects its single-line string
+  dialect: literal hashes, single quotes, empty strings, and the text `null`
+  stay part of the value; producer backslashes and Unicode escapes decode
+  once. Older/tagged documents retain YAML scalar semantics; double-quoted
+  values also retain YAML string semantics.
+  This distinction follows the released
+  [3.0.3 metadata writer](https://github.com/iBotPeaches/Apktool/blob/18b5e99cb56ff9451e8aa55b065dcf5bbd616975/brut.apktool/apktool-lib/src/main/java/brut/androlib/meta/VersionInfo.java#L49-L55)
+  and [string escape owner](https://github.com/iBotPeaches/Apktool/blob/18b5e99cb56ff9451e8aa55b065dcf5bbd616975/brut.j.yaml/src/main/java/brut/yaml/YamlStringEscapeUtils.java#L55-L116).
 - **The manifest is the decoded text**, bounded at 512 KiB and reported as
   truncated when larger.
 - **Strings are projected verbatim.** Escaped entities stay escaped, and
   entries containing nested markup (`<b>`, `<xliff:g>`) are counted as
-  unparsed rather than flattened — the limitation names the count. The
-  default table is bounded to 2000 entries with partial coverage when more
-  exist; `locales` lists every `values-<locale>` that has a strings table.
+  unparsed rather than flattened — the limitation names the examined count.
+  Malformed XML and a wrong document root are whole-document projection
+  failures: the result is partial, the failure reason is disclosed, and entry
+  counts remain unknown. They are not counted as one unsupported string.
+  The default table is bounded to 2000 entries with partial coverage when more
+  exist. `locales` lists observed regular `strings.xml` tables whose exact
+  `values-<locale>` directory is selectable by the request's locale grammar.
+  Locale suffixes allow up to 23 ASCII characters, derived from Apktool's
+  language, script, region, and variant fields; for example,
+  `b+sr+Latn+RS+REVISED` selects that exact table.
+  Other table directories, including combined qualifiers such as
+  `values-en-rUS-v21` and `values-mcc310-fr`, are named in limitations with
+  their source spelling. They are not advertised as selectable locales;
+  REA does not infer Android's runtime resource matching. Non-regular
+  strings-table paths are disclosed with partial coverage. Discovery uses
+  the same bounded traversal as the file count; filesystem failures remain
+  failures rather than being interpreted as missing tables.
+  When multiple tables are requested with `locales`, `string_tables` records
+  each projected table with its locale, status (`present`, `missing`,
+  `oversized`, `malformed`, or `skipped`), entry counts, and coverage,
+  bounded at an aggregate budget of 10,000 strings and 32 MiB XML. When `locales`
+  is omitted and `string_tables_only` is not true, legacy single-table payloads are preserved
+  without adding `string_tables`.
 - **Smali is skipped by design** (`apktool d --no-src`): Java-source
   recovery belongs to the JADX family. Resource smali edits and rebuild are
   a follow-up lane with mutation authority.
@@ -73,17 +142,24 @@ select Android's `values-pt-rBR`; Android qualifiers such as `pt-rBR` and
 
 ## Resource and lifecycle limits
 
-- The decode runs with a 600 s deadline and a 4 MiB output capture budget;
-  a stopped decode reports `provider_timeout` and exceeded budgets report
-  `resource_constraint`.
+- The decode wall clock is `decode_timeout_ms` (CLI `--decode-timeout-ms`).
+  Omitted means 600000 ms. The integer range is 1 to 2147483647. A change
+  applies on the next decode. It does not raise the 4194304-byte (4 MiB)
+  output capture budget. That capture budget has no runtime override. A
+  stopped decode reports `provider_timeout` and names the effective
+  milliseconds; an exceeded capture budget reports `resource_constraint`.
+- The version probe waits 20000 ms and retains 65536 bytes. This release has
+  no runtime override for that probe. `decode_timeout_ms` does not raise it.
 - Manifest reads are bounded to 512 KiB before decoding UTF-8. Strings XML
   reads are bounded to 8 MiB; larger documents are skipped with partial
   coverage and an explicit limit, rather than reported as missing.
 - `decoded_file_count` is capped at 100,000 files; the cap is reported in
   the limitations when hit.
-- Apktool's own decode failures (`AndrolibException`, directory traversal
-  refusals) surface as `unsupported_target` with apktool's reason preserved
-  verbatim.
+- An ordinary nonzero decode exit surfaces as `unsupported_target` naming
+  the caller-selected APK, with Apktool's diagnostic and captured output.
+  Launcher startup, ownership preparation, signal termination and incomplete
+  output collection remain provider failures; they do not establish that the
+  selected APK is unsupported.
 
 ## Real verification
 
@@ -94,3 +170,11 @@ locales — launcher version, on-disk digest agreement, apktool.yml metadata
 (1.2.3, SDK 24–34), manifest package agreement, default and `de` string
 tables projected. Parser goldens in the test suite come from the same
 decode. See `docs/testing.md` for the lane summary.
+
+Multi-locale and table-only projection were also verified through real CLI and
+stdio MCP workflows on Linux with Apktool 3.0.3/JDK 21 against NewPipe 0.29.1,
+ApiDemos 6.0.18 and Appium Settings 8.2.1. Independent XML projection checks,
+exact facts/Evidence parity and bounded task costs are recorded in the
+[shared-decode and table-only checkpoint](https://github.com/morluto/rea/blob/42e1a17d6e6c0487f17be0bae248553c4caba7f7/experiments/analysis-efficiency/apktool-string-tables.md).
+This qualifies selected resource facts, not application execution or independent
+binary resource-decoder completeness.

docs/apple-application-analysis.md

@@ -60,6 +60,15 @@ states that absence is unknown. Component arrays remain complete when candidate
 hypotheses are omitted; component arrays still include every component from
 the supplied inventory pages, and overall coverage is `partial` in that case.
 
+Inventoried files whose bytes were not read, such as encrypted archive members,
+are listed in `unavailable_components` with their original path, format,
+encryption flag, limitations, and null content identity. The same application-root
+and AppleDouble attribution rules apply. These files make overall coverage
+`partial`; component absence is unknown even when `inventory_complete` is true.
+Directories have no file bytes and symlinks remain in `symlinks`, so neither is
+an unavailable file component. The additive field participates in `projection_id`,
+including when empty, so Apple projection identifiers change.
+
 ## Application roots
 
 | Inventory root                         | Application root                                    |
@@ -68,6 +77,15 @@ the supplied inventory pages, and overall coverage is `partial` in that case.
 | `.app` directory                       | `.`, the inventoried bundle itself                  |
 | Directory, ZIP, or DMG containing apps | Each outermost `<Name>.app` followed by `Contents/` |
 
+File and slice occurrences supply component endpoints. Directory tree identities
+remain in source inventory and bundle anatomy, and links remain in `symlinks`.
+
+Root and layout discovery use known directories and ancestors implied by deeper
+inventory paths. A leaf file or symlink named `*.app`, `Contents`, or `Versions`
+does not establish that directory. Directory and authenticated-file identity are
+selected before applying anatomy case preferences, so a file cannot shadow a
+folded directory name or a directory shadow a metadata/signature file.
+
 ZIP archives made with `ditto -c -k --keepParent` report `<Name>.app`. A DMG
 reports `<image>.dmg/<Volume>/<Name>.app`; DMG child inventory needs the native
 macOS mounting adapter. When a DMG has only its root identity, the projection is
@@ -162,8 +180,21 @@ rea trace-dylib-resolution MyApp.app --root Contents/MacOS/MyApp --architecture
 
 Load commands are parsed in TypeScript from Mach-O and FAT headers, including
 the `dylib_use_command` encoding newer linkers emit for flags such as delayed
-initialization. Nothing is executed and no Apple tool is required, so the tool
-also works off macOS.
+initialization, and `LC_LAZY_LOAD_DYLIB_INFO` (`0x3a`). That command stores its
+install name in a linkedit payload rather than in the load command itself.
+Its flag and chain-start image offsets must lie in the slice VM range, which
+is dyld's one-page floor extended by every segment except `__LINKEDIT`,
+measured from the `__TEXT` vmaddr. Payload strings must start after the 24-byte
+header. The load-path NUL must occur before the payload's final byte, matching
+dyld's scan. Distinct payload bytes and payload-range lookups stay within the 16 MiB
+limit, and a file range already read is reused. String scanning and UTF-8
+checks count every reference, including repeats and overlaps, and stop at that
+same aggregate. A slice retains at most 1024 of these commands; further copies are
+unsupported instead of becoming one traced edge each. Symbol references,
+including repeats, stop at 65536 per slice. A larger table is unsupported
+instead of caching one string per offset.
+Nothing is executed and no Apple tool is required, so the tool also works off
+macOS.
 
 Resolution follows dyld:
 
@@ -182,30 +213,62 @@ Resolution follows dyld:
 Each edge lists every candidate path tried, with its outcome: `resolved`,
 `absent`, `not-mach-o`, `malformed`, `unsupported` (recognized Mach-O outside
 parser coverage, such as big-endian images), `architecture-missing`,
-`outside-target`, `escapes-target`, or `undetermined`. Slices follow dyld's
-graded architectures, so an `x86_64h` process also loads `x86_64`. The edge's `resolution` is one of:
+`not-loadable`, `platform-mismatch`, `outside-target`, `escapes-target`, or
+`undetermined`. `not-loadable` means a
+matching-architecture Mach-O image was found but it is not a dylib and cannot
+satisfy a dylib dependency. `platform-mismatch` means a matching-architecture
+dylib was found but its observed platform is incompatible with the process.
+These outcomes reject that candidate; dyld search may continue and resolve a
+later candidate. Candidate outcomes describe individual search paths, while
+`resolution.status` describes the final edge. For each selected process-root
+slice, candidate slices are matched by CPU type and dyld's subtype preference
+order, independently of FAT table order. An ARMv7s root slice prefers ARMv7s,
+then ARMv7; other known ARM32 subtypes use their compatibility rows from
+[Apple's dyld source](https://github.com/apple-oss-distributions/dyld/blob/dyld-852.2/src/dyld2.cpp#L2811-L3105).
+An `x86_64h` process also loads generic `x86_64`. Unknown ARM32 subtype policy
+remains undetermined when ARM32 candidate slices exist. This is a static model
+keyed to the selected root slice; it does not observe the target device's host
+CPU or prove which slice the device will execute. The edge's `resolution` is
+one of:
 
-| Status         | Meaning                                                                                                                                                   |
-| -------------- | --------------------------------------------------------------------------------------------------------------------------------------------------------- |
-| `resolved`     | An image inside the analyzed root loads, and no earlier candidate was undeterminable.                                                                     |
-| `conditional`  | An image inside the root resolves, but an earlier candidate outside the root (for example `/usr/lib/swift`) would win if it exists on the running system. |
-| `unresolved`   | Every candidate is definitively absent or unusable inside the root.                                                                                       |
-| `undetermined` | No candidate resolves inside the root, and some candidate lies outside it or depends on the environment.                                                  |
+| Status         | Meaning                                                                                                                                                        |
+| -------------- | -------------------------------------------------------------------------------------------------------------------------------------------------------------- |
+| `resolved`     | An image inside the analyzed root loads, and no earlier candidate was undeterminable.                                                                          |
+| `conditional`  | An in-root image may load, but an earlier outside candidate, unmodeled environment input, or unknown platform compatibility can change which image dyld loads. |
+| `unresolved`   | Every candidate is definitively absent or unusable inside the root.                                                                                            |
+| `undetermined` | No candidate resolves inside the root, and some candidate lies outside it or depends on the environment.                                                       |
 
-A `conditional` image is still traced, because it is the image the analyzed
-root supplies. Every edge below it has `loader_conditional: true`, and a
-request that reuses it by install name is `conditional` too. Findings on those
-edges say that they apply only if the conditional image loads.
+A `conditional` image is still traced, because it is an in-root candidate
+that may load. Every edge below it has `loader_conditional: true`, and a
+request that reuses it by install name is `conditional` too. An in-root dylib
+with no platform load command remains unknown unless the root is an executable
+with observed macOS platform and the candidate’s CPU type is x86_64, i386, or arm64. In that macOS context,
+[Apple's dyld](https://github.com/apple-oss-distributions/dyld/blob/fd8d0c4d52320ebf64db34f3cb280310d905c5ae/mach_o/UnsafeHeader.cpp#L1137-L1145) supplies an implicit macOS platform to legacy dylibs on those CPU
+types. The candidate records `platform_basis: "macos-dyld-implicit"`; the
+reported slice fields `platform: null` and `platforms: []` remain the parsed
+facts. For other process contexts the platform remains unknown, coverage is
+partial, and the candidate is not treated as incompatible. If a later known
+candidate resolves, it remains the traced image with conditional resolution;
+the trace does not claim the unknown candidate's dependency graph. Findings on
+conditional edges say that they apply only if the conditional image loads.
 
-Absolute install names such as `/usr/lib/libSystem.B.dylib` are
-`outside-target`. On macOS 11 and later most of them live in the dyld shared
-cache rather than on disk, so REA does not check them against the host.
+Absolute install names that name a path beneath the analyzed root are checked
+within that root, with symlinks and `..` resolved by the same bounded tree
+resolver. Absolute paths outside the root, such as
+`/usr/lib/libSystem.B.dylib`, are `outside-target`; on macOS 11 and later most
+system libraries live in the dyld shared cache rather than on disk, so REA does
+not check them against the host. Mach-O path strings are not converted from
+`file:` URLs or looked up through unrelated host paths.
 
 `findings` are derived, not observed:
 
 - `required-load-unresolved`, `weak-load-unresolved`, and `lazy-load-unresolved`.
-  `LC_LAZY_LOAD_DYLIB` dependencies are resolved but not traversed, because dyld
-  loads them on first use; an unresolved one is not a launch failure.
+  `LC_LAZY_LOAD_DYLIB` and `LC_LAZY_LOAD_DYLIB_INFO` dependencies are resolved
+  but not traversed, because dyld loads them on first use; an unresolved one is
+  not a launch failure. `LC_LAZY_LOAD_DYLIB_INFO` has no compatibility version,
+  so those fields are null. Its may-be-missing flag is reported as `weak`, and
+  an unresolved dependency with that flag is not a first-use failure: dyld
+  continues if the image is absent.
 - `earlier-rpath-candidate-absent`: a compatible Mach-O placed at an earlier
   modeled search path could take precedence, subject to unmodeled search inputs
   and code-signing library validation. These findings do not establish which

docs/architecture.mermaid

@@ -1,8 +1,8 @@
 graph TD
     subgraph "Entry Points"
         CLI["src/cli.ts + src/cli/<br/>One-shot CLI adapter<br/>setup, diagnostics, analysis, decompilation"]
-        MAIN["src/main.ts + src/main/<br/>MCP stdio server adapter<br/>config, session, process lifetime"]
-        DISPATCH["scripts/rea.mjs<br/>Executable dispatcher<br/>routes mcp / one-shot"]
+        MAIN["src/main.ts + src/main/<br/>MCP server adapter: stdio default, streamable HTTP opt-in<br/>config, session, process lifetime"]
+        DISPATCH["scripts/rea.mjs<br/>Executable dispatcher<br/>routes mcp / mcp --http / one-shot"]
     end
 
     subgraph "Production Composition"
@@ -10,7 +10,7 @@ graph TD
     end
 
     subgraph "Adapters"
-        SERVER["src/server/<br/>MCP request translation<br/>createServer, tool registration, toolResult"]
+        SERVER["src/server/<br/>MCP request translation<br/>createServer, tool registration, toolResult<br/>http/: streamable HTTP listener and node:http bridge"]
         HOPPER_BRIDGE["bridge/hopper_bridge.py<br/>Runs inside Hopper<br/>adapts operations to Hopper Python API"]
         GHIDRA_BRIDGE["bridge/ghidra/ReaGhidraBridge.java<br/>Packaged HeadlessScript<br/>inventory, decompiler, references, CFG"]
         INSTALL["src/cli/ + src/application/Setup*<br/>Host installation adapters<br/>macOS: verified vendor DMG to ~/Applications<br/>Linux: verified vendor package + apt/dnf/pacman"]
@@ -26,14 +26,16 @@ graph TD
     end
 
     subgraph "Provider Layers"
+        PROVIDER_OWNERS["src/providers/<br/>shared deferred client and inspection owners<br/>exact cleanup retry and attributed failures"]
         PWNTOOLS["src/native/pwntools/ + src/native/pwndbg/<br/>BYO offline ELF/core adapters<br/>owned snapshots/Python and optional core-only GDB<br/>unchanged upstream parsers, ABI structures and mapping API"]
         WASM["src/wasm/<br/>caller-supplied WABT; immutable artifact inspection<br/>owned process supervision and retryable workspace cleanup"]
+        APKSIG["src/apkSignature/<br/>in-process APK Signing Block + v1 reader<br/>bounded single read, no device or execution"]
         EVM["src/evm/<br/>unchanged EVMole/WASM in bounded owned worker<br/>explicit carrier parsing and inferred interface adapter"]
         HOPPER["src/hopper/<br/>Hopper launch & Unix-socket protocol<br/>BridgeLauncher, HopperClient, protocol"]
         GHIDRA["src/ghidra/<br/>Ghidra discovery, profile & function analysis<br/>snapshot, Unix/Windows transport, bounded lifecycle"]
         IDA["src/ida/<br/>Upstream MCP contract adaptation<br/>live GUI binding or owned headless database"]
         IDA_MCP["External ida-pro-mcp<br/>GUI proxy or database supervisor<br/>upstream installation and engine lifecycle"]
-        AUX["src/artifacts/ + src/native/<br/>artifact inventory and extraction<br/>format readers and native capabilities"]
+        AUX["src/artifacts/ + src/native/<br/>artifact inventory, bounded directory triage and extraction<br/>format readers and native capabilities"]
         BROWSER["src/browser/<br/>CDP / Playwright capture<br/>captured scripts and native module resolution<br/>execution/: coverage, listener and source attribution"]
         JS_CODEC["src/javascript/sourceMaps/<br/>pinned upstream map codec<br/>owned heap-limited Node process, complete point evidence"]
         INSPECTOR["src/inspector/<br/>passive V8 Inspector discovery and capture<br/>script and target locations"]
@@ -100,6 +102,8 @@ graph TD
     SERVER -->|narrow record ports| RECORDS
     RECORDS --> DOMAIN
     COMPOSITION --> WASM
+    COMPOSITION --> APKSIG
+    APKSIG --> DOMAIN
     WASM --> PROCESS
     WASM --> DOMAIN
     COMPOSITION --> PWNTOOLS
@@ -138,6 +142,11 @@ graph TD
     GHIDRA --> GHIDRA_BRIDGE
     HOPPER --> PROCESS
     GHIDRA --> PROCESS
+    GHIDRA --> PROVIDER_OWNERS
+    APP --> PROVIDER_OWNERS
+    PROVIDER_OWNERS --> PROCESS
+    PROVIDER_OWNERS --> AUX
+    PROVIDER_OWNERS --> DOMAIN
     GHIDRA --> WINDOWS
     PROCESS --> WINDOWS
     WINDOWS --> WINDOWS_NATIVE

docs/binary-diagnostics.md

@@ -147,6 +147,29 @@ received signal alone does not establish its exact cause. Dynamic tags require
 a complete DT_NULL inside PT_DYNAMIC; interpreter names and ranges end at the
 first NUL, with any padding still represented by the original segment range.
 
+## Ghidra process failures
+
+When output from the failing Ghidra startup or operation reports
+`java.lang.OutOfMemoryError`, REA returns `resource_constraint`. Java heap or GC
+overhead exhaustion includes heap-specific recovery advice. Metaspace and array
+size failures do not recommend increasing the heap; native-thread exhaustion is
+reported as a capacity constraint with unknown limits.
+
+`configured_jvm_max_heap` records the explicit `-Xmx` value selected by REA's
+launch builder. It is a configuration fact, not measured memory use or proof of
+the effective JVM limit; JVM environment options may affect that limit.
+When a Windows script wrapper selects the JVM, this value remains unknown.
+Historical stdout and stderr remain available, but classification uses only the
+failing operation's `failure_output`. Retained stdout and stderr share an 8 MiB
+budget, but every drained byte is scanned for `java.lang.OutOfMemoryError`:
+`failure_output.observed_oome` records the latest confirmed token's bounded line per stream from
+the failing operation, with its byte offset, even after the budget is spent. These observations own OOME detection; retained text that ends at a class-name prefix cannot confirm a complete token. `line_complete` distinguishes a known line end from an incomplete trailing context; `invalid_utf8` marks replacement-decoded context.
+Incomplete operation output without an observed OOME keeps the original failure code and explains that a memory cause cannot be ruled out. Incompleteness alone does not confirm a memory constraint or justify heap-specific advice. Token context stays within the failing operation's checkpoint; a class-name prefix in a different identifier is not OOME evidence. A trailing bare token requires an observed stream EOF to establish its boundary.
+Output draining waits at most one second,
+subject to earlier cancellation or the startup deadline. `output_incomplete`
+marks a snapshot taken before stream closure; cleanup ownership remains with
+the session until release succeeds.
+
 ## Portable PE resources
 
 `inspect_pe_resources` / `inspect-pe-resources` inspects an explicit PE32 or
@@ -192,3 +215,157 @@ Large Evidence envelopes may require a caller-configured MCP transport receive
 buffer because envelope copies add overhead. Requests honor cancellation during
 snapshot reading, directory traversal and payload hashing. The operation reads
 regular files only and rejects symbolic links.
+
+## Byte entropy triage
+
+`inspect_byte_entropy` / `inspect-byte-entropy` profiles any local regular file
+without an analysis engine, execution, extraction or network access.
+
+```sh
+rea inspect-byte-entropy ./firmware.bin --block-size 4096 --json
+rea inspect-byte-entropy ./firmware.bin --offset 4096 --length 65536 --json
+```
+
+```json
+{
+  "name": "inspect_byte_entropy",
+  "arguments": { "path": "/artifacts/firmware.bin", "detail": "summary" }
+}
+```
+
+Complete Evidence includes `artifact` (original path, byte size and SHA-256),
+`block_size`, `window`, `file_entropy`, `thresholds`, `blocks`, `runs` and
+`limitations`. For a 4096-byte zero-filled file, the entropy fields are:
+
+```json
+{
+  "block_size": 4096,
+  "window": { "offset": 0, "length": 4096 },
+  "file_entropy": 0,
+  "thresholds": { "high": 7.2, "low": 1 },
+  "blocks": [{ "offset": 0, "length": 4096, "entropy": 0 }],
+  "runs": [{ "offset": 0, "length": 4096, "classification": "low" }]
+}
+```
+
+Offsets and lengths are decimal byte counts. Blocks begin at the requested
+`offset` (default zero); omitted `length` extends to EOF. Explicit windows past
+EOF fail rather than silently clipping. The final block may be shorter.
+`block_size` must be a power of two from 256 through 65536, default 4096. A
+window may produce at most 32,768 block rows; larger requests fail before the
+file scan with a capacity error. Choose a shorter window or larger block size.
+SHA-256 and `file_entropy` always cover the entire file, including when a window
+is selected. Empty files and empty windows are accepted; empty entropy is zero
+by convention. The tool reads the size the file reports. A file that reports size
+0, such as a Linux `/proc` entry, returns no blocks with a limitation that the
+result does not show the file is empty. File identity, size and timestamps are checked across the read;
+detected changes produce an error rather than Evidence.
+
+Shannon entropy is reported in bits per byte from 0 to 8, rounded to four decimal
+places. Adjacent blocks with reported entropy at least 7.2 form `high` runs;
+blocks at most 1.0 form `low` runs. Other blocks interrupt runs. Entropy does not
+prove encryption, and this measure cannot distinguish encrypted from compressed
+data. Small blocks can underestimate entropy; the result is a triage clue.
+
+MCP `detail: "summary"` retains complete Evidence in the session and returns
+file entropy, block/run counts, thresholds, the selected window, and the ten
+longest runs (ties ordered by file offset). Use its `parent_evidence_id` to
+retrieve complete Evidence, or request its summary with `inspect_analysis_view`.
+The CLI returns complete Evidence through the same service. Summary detail
+requires a session capable of retaining Evidence.
+
+Reads use one block-sized buffer and a 256-entry histogram. Complete output
+can become very large because it includes one row per selected block, and
+summary delivery still retains that full list. Use a shorter length or a
+larger block size to reduce it. Full Evidence and its serialization require
+memory proportional to the number of selected blocks. The 32,768-row ceiling bounds the actual retained block objects and
+their serialized representation: a cap-sized run below produced 1,514,877
+serialized bytes and increased V8 heap use by 21.2 MiB on the measured host.
+The cap therefore bounds this output independently of file size; it is not a
+portable process-memory guarantee. Narrow the window or increase block size to
+reduce retained rows further.
+
+Local service measurement (macOS arm64, Node 24.18.0, 100,000,000-byte file,
+4096-byte blocks, file already in the OS cache): first process run took 1.113 s
+wall / 1.148 s CPU with 145.3 MiB peak RSS; a second run in the same process took
+1.015 s / 1.035 s CPU, with cumulative peak RSS 155.8 MiB. Each read 100,000,000
+bytes and produced 24,415 blocks. Measurements include Evidence construction,
+summary projection and JSON serialization, with no child processes. Complete
+Evidence serialized to 1,122,303 bytes and its summary to 2,705 bytes. A
+cap-sized 134,217,728-byte sparse file with 4096-byte blocks produced 32,768
+rows and 1,514,877 serialized bytes; V8 heap use increased by 21.2 MiB and RSS
+by 18.1 MiB in that process. This separate measurement used the same macOS
+arm64 host and Node version. Cold disk I/O, token counts and an agent
+investigation were not measured; there was no previous equivalent tool to
+benchmark. These figures do not establish a portable peak-memory bound,
+constant memory for retained Evidence or faster agent task completion.
+
+## Portable native exports
+
+`inspect_native_exports` / `inspect-native-exports` reads the export table of an
+explicit PE or ELF image on any supported host, without an active target,
+native engine, `objdump` or Windows SDK:
+
+```sh
+rea inspect-native-exports ./runtimes/linux-x64/native/libexample.so --json
+```
+
+```json
+{
+  "name": "inspect_native_exports",
+  "arguments": { "path": "/artifacts/native/example.dll" }
+}
+```
+
+The complete Evidence identifies the stable artifact snapshot by path, byte
+count and SHA-256.
+
+- PE rows reproduce the export directory: each used export-address-table slot
+  with its biased `ordinal` (`OrdinalBase` plus the slot index), one row per
+  name (aliases share an ordinal), `address`
+  as the preferred-base virtual address, or `forwarder` text when the slot
+  forwards to another DLL. Unused zero-address slots are omitted;
+  ordinal-only exports have a null `name`. The name-ordinal table stores an
+  unbiased 16-bit EAT index; reported biased ordinals may exceed 16 bits. The
+  directory's DLL name, ordinal base and timestamp are reported, and an image
+  without an export directory is a complete empty inventory.
+- ELF rows list defined `.dynsym` symbols with global, weak or GNU-unique
+  binding and default or protected visibility. Each row keeps `value`
+  (`st_value` as stored; bit 0 marks Thumb code on 32-bit ARM), `size`, `type`
+  and `binding`. ELF string tables do not declare a text encoding. Names that
+  contain non-ASCII bytes are decoded as UTF-8 when those bytes form valid
+  UTF-8; `name_utf8_decodable` records whether the non-ASCII name bytes are
+  valid UTF-8 and `name_bytes_base64` retains those exact bytes (`null` for
+  ASCII names). When they are not valid UTF-8, `name` uses a byte-preserving
+  Latin-1 mapping and the raw bytes remain available. Older retained results
+  report null for the newly added encoding fields. `version` is null for an
+  unversioned symbol. A versioned
+  symbol reports its GNU index, a name resolved from `.gnu.version_d` or
+  `.gnu.version_r`, the corresponding `name_bytes_base64` and
+  `name_utf8_decodable` facts, and the `hidden` flag from `.gnu.version`.
+  Version names use the same UTF-8-when-valid, otherwise byte-preserving
+  Latin-1 display mapping as symbol names; ELF does not declare either
+  string's text encoding. Older retained results report null for the added
+  version-name encoding fields. `"unknown"`
+  means the symbol-version table could not be read. A hidden version is not
+  the default version that a by-name `dlsym` lookup selects.
+  Definition names take precedence when both tables name the same index.
+
+Unreadable names, out-of-range name ordinals and unreadable forwarders are
+listed in `issues`, their entries are not reported, and coverage is `partial`.
+PE export names are bounded null-terminated byte strings and may continue
+across adjacent file-backed section mappings when the virtual bytes are
+unambiguous. Gaps, zero-fill and conflicting mappings remain unreadable names.
+Unresolved symbol versions keep their rows: the version name is null when no
+definition or requirement names the index, and `version` is
+`"unknown"` when the version table cannot be read; coverage is likewise
+`partial`. A dynamic symbol table without its null entry is malformed. An ELF
+without section headers is `unsupported`:
+enumerating dynamic symbols from the dynamic segment is not
+implemented. Mach-O exports are reported by `inspect_macho`.
+
+`max_file_bytes` defaults to 256 MiB (maximum 512 MiB; CLI `--max-file-bytes`).
+Export directories in PE headers are read like those in sections. An export or
+dynamic symbol table above 262144 entries, more than 262144 rows, or more than
+16 MiB of retained names fails without partial success. Export rows do not establish that a symbol is reachable, safe to call
+or loaded at runtime.

docs/browser-observation.md

@@ -18,9 +18,32 @@ Inspect native SAZ or HAR 1.2 offline with REA, or connect Everywhere's separate
 - `inspect_web_page` / `rea inspect-web-page` captures DOM structure, accessibility nodes, scripts, resources, safe response/DOM metadata, attach-window network and console metadata, WebSocket frame sizes, workers, quota, and optionally selected storage key names, redacted storage-content fingerprints, or script sources.
 - `analyze_web_bundle` / `rea analyze-web-bundle` parses selected script artifacts without execution and derives chunk edges, route and endpoint candidates, vendor fingerprints, static WebMCP declarations, and optionally source-map/original-source evidence.
 - `observe_web_session` / `rea observe-web-session` arms an observation window of the requested duration for an external user action and records ordered reload, SPA navigation, redirect, failure, lifecycle, and target-termination metadata.
-- `discover_webmcp_tools` / `rea discover-webmcp-tools` uses the experimental CDP WebMCP domain to return every in-scope registration and its complete observed `input_schema` declaration and canonical `input_schema_sha256` digest inline. Declarations are untrusted JSON data, including declared types, required fields, enums, examples and reference URLs. REA does not resolve schema references. These fields replace `input_schema_shape`. Registrations with the same name and URL in distinct observed frames retain separate identities. `tool_key` includes the observed frame ID; it is not a stable identity across navigation or browser sessions. Same-frame registration updates replace that owner only, and removal leaves other owners intact. Declared `readOnly`, `untrustedContent`, `consequential`, and `autosubmit` annotation hints are reported per registration: `true`/`false` are declared values, `null` means the declaration did not provide the hint in that capture, and a missing `consequential` key means the hint was not recorded, either in a historical capture or because the observed annotation was invalid. Consult discovery completeness and exclusions to distinguish these cases. Malformed annotation values retain the registration with incomplete coverage and an `invalid_protocol_value` exclusion. Other malformed protocol payloads can fail discovery explicitly. REA never exposes `WebMCP.invokeTool`.
+- `discover_webmcp_tools` / `rea discover-webmcp-tools` uses the experimental CDP WebMCP domain to return every in-scope registration and its complete observed `input_schema` declaration and canonical `input_schema_sha256` digest inline. Declarations are untrusted JSON data, including declared types, required fields, enums, examples and reference URLs. REA does not resolve schema references. These fields replace `input_schema_shape`. Registrations with the same name and URL in distinct observed frames retain separate identities. `tool_key` includes the observed frame ID; it is not a stable identity across navigation or browser sessions. Same-frame registration updates replace that owner only, and removal leaves other owners intact. Declared `readOnly`, `untrustedContent`, `consequential`, and `autosubmit` annotation hints are reported per registration: `true`/`false` are declared values, `null` means the declaration did not provide the hint in that capture, and a missing `consequential` key means the hint was not recorded, either in a historical capture or because the observed annotation was invalid. Consult discovery completeness and exclusions to distinguish these cases. Malformed annotation values or registration stacks retain the registration with incomplete coverage and an `invalid_protocol_value` exclusion. Other malformed protocol payloads can fail discovery explicitly. REA never exposes `WebMCP.invokeTool`.
+
 - `compare_web_captures` / `rea compare-web-captures` compares passive captures or exact-step-aligned browser scenarios. Scenario results commit reproducible literal normalization and expose alignment failures plus artifact-level action, screenshot, DOM, accessibility, URL, history, storage, and event differences. Missing or truncated evidence is never treated as equivalence.
 
+For a compact route query, select `detail: "summary"` and a
+`view: {kind: "javascript-routes", collection: "routes", limit: 20}`. The
+shared [route inventory](javascript-route-inventory.md) preserves source IDs,
+precise locations, partial coverage, and the complete parent Evidence for
+subsequent filtered pages. Static candidates never establish complete application
+coverage or reachability.
+
+For `analyze_web_bundle`, each `webmcp_declarations` entry's
+`schema_property_names` is a source-derived projection of statically known own
+property keys, not schema validation. `schema_property_names_coverage` reports
+how complete that projection is: `complete` means a literal properties object
+was fully enumerated (including an empty literal); `partial` means a literal
+object was available but some computed or spread membership was unresolved,
+and the name array retains keys that were known; `unknown` means the schema or
+properties owner could not be enumerated as an object literal, including
+explicit non-object values; `not_present` means no effective schema or own
+`properties` member was present after the `inputSchema ?? input_schema` alias
+selection. `not_present` describes this projection only: it does not validate a
+schema or rule out properties expressed through other schema constructs. Older
+observations may omit the coverage field; treat omission as `unknown` rather
+than inferring completeness from the array.
+
 Passive DOM comparison checks complete node inventories and complete frame
 parent/child structure independently. Frame IDs and sibling enumeration order
 are capture-local and do not establish a change; duplicate IDs, missing parents,
@@ -31,6 +54,26 @@ from a protocol omission. Rejected DOM facts also mark derived URL and agent-hin
 metadata incomplete, while valid facts from other nodes and documents remain
 available in the capture.
 
+Passive comparison keeps lookup identity, cross-capture correspondence, content
+evidence, and inventory coverage separate:
+
+| Observation                                                                                      | Comparison outcome                                                                      |
+| ------------------------------------------------------------------------------------------------ | --------------------------------------------------------------------------------------- |
+| Renamed WebMCP tool keys/frame IDs with one matching name, owner origin, and frame URL           | Compare the registration content; local IDs do not prove change                         |
+| Duplicate WebMCP owners with the same name, origin, and frame URL                                | Unknown correspondence; complete group count differences still prove change             |
+| Matching script inventory keys with retained source digests                                      | Compare the retained SHA-256 values even without a producer hash                        |
+| One corresponding script source is unavailable, or neither source nor producer hash is available | Source equality remains unknown; inventory changes and other proven differences survive |
+| A partial HTTP response group is a subset of the complete peer                                   | Unknown; omitted repeated responses can explain the difference                          |
+| An observed HTTP response fingerprint is absent from the complete peer group                     | Changed; omissions cannot explain that retained response                                |
+| A WebSocket connection or frame inventory is incomplete                                          | Do not align retained payloads by position                                              |
+
+Storage fingerprints decode supported CDP primitive/JSON values before hashing.
+Debugger descriptions, previews, and remote references do not represent storage
+content. Actual JSON properties with those names remain content. Reference-only
+objects, unsupported values, and opaque cookie partition identities remain
+incomplete. Previously exported fingerprints are not recomputed; recapture both
+sides when comparing captures produced across this decoder change.
+
 - `capture_web_screenshot` / `rea capture-web-screenshot` returns a content-addressed visible-viewport PNG.
 - `compare_web_screenshots` / `rea compare-web-screenshots` performs bounded local PNG pixel comparison without OCR or external services.
 - The CLI comparison commands accept each capture, screenshot artifact or normalization policy as inline JSON or a JSON file path. Use files for screenshot artifacts and scenario captures, whose embedded PNG bytes can exceed the host's command-line length limit.
@@ -66,14 +109,25 @@ Electron `file://` pages use a separate provider and target boundary; see [elect
 
 ## Provider authorities
 
-CDP page capture, V8 Inspector observation, and Playwright scenarios stay
-separate providers because their authorities differ: CDP and V8 attach to an
-already-running target and never launch, drive, evaluate, or mutate it, while
-Playwright providers own and drive the runtime (launch browsers, run actions,
-capture step snapshots). Merging them would mix attach-only and owned-process
-lifecycles in one contract. What they share instead is one exclusion
-vocabulary (`BrowserExclusionReason`): every denial reason in every stack maps
-onto it, even where a wire schema keeps a narrower historical bucket.
+CDP page capture and V8 Inspector observation attach to an already-running
+target without launching it, driving actions, or evaluating target code.
+Playwright scenarios drive actions and capture step snapshots in either a
+provider-owned browser or a caller-selected external browser. Separate providers
+preserve these authority and lifecycle differences. They share one exclusion
+vocabulary (`BrowserExclusionReason`): every denial reason maps onto it, even
+where a wire schema keeps a narrower historical bucket.
+
+CDP attachment is not a guarantee that the browser preserves all state.
+Chromium can clear another client's `navigator.platform` emulation when REA's
+page session disconnects, even if that session sent no emulation commands.
+This affects page inspection, bundle capture, session observation, screenshots,
+WebMCP discovery, and Electron page inspection; target listing alone does not
+attach a page session. REA cannot read the other client's complete
+user-agent/client-hint configuration and does not replay guessed settings.
+Use a disposable browser when preserving externally configured emulation is
+required, or have the owning client reapply its complete configuration afterward.
+See [the reproduction and Chromium ownership limitation](https://github.com/morluto/rea/issues/4089).
+
 Existing static application Evidence and passive web/Electron captures can be
 combined later through
 [JavaScript static/runtime reconciliation](javascript-runtime-reconciliation.md).
@@ -98,14 +152,11 @@ bundle, for example
 `"/Applications/Brave Browser.app/Contents/MacOS/Brave Browser"`. Fedora installs
 Chromium as `chromium-browser`.
 
-REA needs the browser's HTTP discovery endpoint on a loopback port. A browser
-driven over `--remote-debugging-pipe` or a private channel, for example by an
-automation SDK, exposes no such endpoint, so `list_browser_targets` has nothing
-to attach to. Connection failures report `endpoint_unreachable`; a reachable
-HTTP service that returns a non-200 or non-JSON discovery response reports
-`invalid_endpoint_response`.
-Reproduce that flow in a Chromium-family browser started with
-`--remote-debugging-port`.
+REA attaches through the browser's HTTP discovery endpoint on a loopback port,
+or, for some passive tools, through a
+[browser socket](#browsers-without-http-discovery). Connection failures report
+`endpoint_unreachable`; a reachable HTTP service that returns a non-200 or
+non-JSON discovery response reports `invalid_endpoint_response`.
 
 In a minimal Linux container, headed Chromium under Xvfb can keep running while
 its DevTools port never opens. Install the GLES, EGL, and Mesa DRI libraries
@@ -157,7 +208,9 @@ file for you.
   the origins you are investigating.
 
 A browser driven only through `--remote-debugging-pipe` or a vendor's private
-automation channel has neither endpoint, and REA cannot attach to it.
+automation channel has neither endpoint, and REA cannot attach to it. To
+inspect such a page, reproduce it in a Chromium-family browser started with
+`--remote-debugging-port`.
 
 ## Request boundary
 
@@ -190,6 +243,11 @@ rea inspect-web-page http://127.0.0.1:9222 TARGET_ID \
   --json
 ```
 
+Passive observation honors the requested duration using cancellable waits within
+the runtime's timer range. Long windows stay pending across successive waits;
+caller cancellation, target termination or leaving the allowed origins ends
+collection early. `window_elapsed` identifies completion of the requested wait.
+
 Static bundle analysis and a user-driven observation window are separate operations:
 
 ```bash
@@ -272,21 +330,54 @@ Call `list_browser_targets` first because target IDs are browser-instance-specif
 
 REA establishes a main-frame origin boundary before enabling Runtime, Debugger, or Network observation, rechecks it before document capture, and validates it again afterward. If the main frame navigates during final capture, REA discards the mixed result and returns `target_changed`.
 
+The web-session timeline uses CDP monotonic seconds where the event producer
+provides a comparable timestamp; absent or invalid timestamps are `null` rather
+than a synthetic zero. Events buffered while REA samples the window boundary are
+retained only when their timestamp places them after that boundary; older events
+are excluded. A buffered event without a comparable timestamp may remain as
+contextual evidence with `timestamp: null`, accompanied by a limitation stating
+that it does not establish activity in the selected window. Main-frame scope
+enforcement still applies to contextual navigation events. Network redirect
+events retain their paired prior-hop response context when the timestamped
+next-request event falls inside the window. The pre-arm timeline buffer is bounded
+by original CDP message bytes. If clock sampling completes after that budget is
+exhausted, the operation returns `payload_limit` instead of a partial window.
+Cancellation and clock-request failures retain their own outcomes.
+
 ## Data minimization
 
+The Link projection parses repeated CDP header field lines independently and
+retains their declarations in order. Each link keeps its target in `href` and
+its relationship context in `context_url` with a separate `context_scope`;
+absent `anchor` uses the response URL, while a relative `anchor` resolves
+against that URL. Anchored links remain in the projection, but a `link_rel`
+agent hint is emitted only when the resolved context matches the response URL
+before userinfo credentials are removed. Following [RFC 8288](https://www.rfc-editor.org/rfc/rfc8288.html),
+the `rel`, `anchor`, and `type` parameters use their first occurrence, and
+quoted-pair escapes are decoded as specified by [RFC 9110 §5.6.4](https://www.rfc-editor.org/rfc/rfc9110.html#section-5.6.4).
+The context fields are optional when reading older saved captures. Commas
+inside URI targets or quoted parameters remain part of those values.
+Well-known-resource hints require an exact supported pathname; path case is
+preserved because a case variant may identify a different HTTP resource.
+COOP and COEP projections retain the exact declared token after validating the
+complete Structured Field Item, including parameters such as `report-to`.
+These are declarations, not an assessment of the document's effective policy;
+absent, malformed or non-token items project as `null`.
+
 REA retains local observation data and removes only URL userinfo credentials, explicitly declared scenario secrets, and values from structured credential/storage fields:
 
 - URL observations preserve the complete local URL, including query values, fragments, duplicate parameters, and their order. Only username/password userinfo is removed; malformed original URL strings remain available with `origin: null`.
 - DOM snapshots retain node types, node names, value lengths, and attribute names, but not text or attribute values.
-- Accessibility structure and roles are retained by default, while names and descriptions require `include_accessibility_text: true`.
+- Supported DOM URL attributes are projected separately as resolved destinations. Empty `action` on HTML or XHTML forms and empty `formaction` on HTML or XHTML submit controls resolve to the document URL; other relative URL attributes resolve against the captured document base.
+- Accessibility structure and reported roles are retained by default, while names and descriptions require `include_accessibility_text: true`. The browser protocol may omit an individual role; that role remains unknown for comparison. Other observed accessibility differences can still establish a change.
 - Network observations retain method, status, MIME type, size, type, initiator stack location, and complete URLs with userinfo credentials removed. Redirects reported by CDP are attached as ordered `redirects` on the one request ID, with the prior request URL, CDP response URL, method/type, response status/MIME/encoded length, request timestamp, and the next `requestWillBeSent` event timestamp. CDP delivers `redirectResponse` with that next request event, so the latter is an event boundary rather than an exact response timestamp. Redirect response headers and bodies are discarded; hops outside the selected origins are discarded with the request chain. Headers are discarded after an allowlisted projection of length/encoding, structured CSP/Link/policy fields, and untrusted agent hints. Cookies and authorization headers are never retained.
-- Request/response bodies are not requested or parsed by default. When selected, only allowed-origin JSON media types are read, converted immediately to complete value-free typed paths, types and observation counts, and discarded. Each path segment is either `{ "kind": "property", "name": "raw property name" }` or `{ "kind": "array-element" }`; array-element types include primitives and nested arrays. Literal `*` properties remain distinct from array elements. `Network.getResponseBody` is never sent unless JSON body shape capture is selected. Comparison uses body shapes only when both captures selected them, and compares request/response shapes only where both observations provide those facts. Unavailable or truncated selected coverage makes equivalence unknown; observed differences still report changed.
+- Request/response bodies are not requested or parsed by default. When selected, only allowed-origin JSON media types are read, converted immediately to complete value-free typed paths, types and observation counts, and discarded. Each path segment is either `{ "kind": "property", "name": "raw property name" }` or `{ "kind": "array-element" }`; array-element types include primitives and nested arrays. Literal `*` properties remain distinct from array elements. `Network.getResponseBody` is never sent unless JSON body shape capture is selected. A JSON-typed request body that CDP reports through `hasPostData` but omits from `postData`, or that cannot be parsed as JSON, makes `json_body_shapes` unavailable. That request's `body_shapes.status` is `partial`, not `included`, when only its response shape was captured. Comparison uses body shapes only when both captures selected them, and compares request/response shapes only where both observations provide those facts. Unavailable or truncated selected coverage makes equivalence unknown; observed differences still report changed.
   For normalized passive inputs with complete request inventories on both sides,
   request additions and removals remain observable even when selected body shapes,
   response metadata, or WebSocket coverage are unavailable. HTTP request and
   WebSocket inventory coverage are evaluated separately; complete HTTP coverage
-  cannot prove a missing WebSocket connection. Limited request inventory still
-  prevents request-absence claims. This comparison rule does not make live CDP's
+  cannot prove a missing WebSocket connection. A missing request in a limited inventory cannot prove absence; a retained
+  request absent from a complete peer can prove a change. This comparison rule does not make live CDP's
   attach-limited inventory complete or establish body-shape equality.
   An unavailable HTTP resource type also leaves request identity incomplete:
   observations with the same method and URL remain possible matches across types,
@@ -301,15 +392,19 @@ REA retains local observation data and removes only URL userinfo credentials, ex
   uses `websocket_connections` and `websocket_frames`. Drops in other event kinds
   and the aggregate total do not by themselves limit HTTP inventory. Proven
   modifications remain reportable despite incomplete inventory.
-  Repeated identical HTTP content samples establish multiplicity changes only
-  with complete request inventories; limited inventories retain known content
-  comparisons without interpreting duplicate sample counts as changes.
+  Repeated HTTP samples retain multiplicity. A partial group that is contained
+  in a complete peer cannot prove modification, even when the omitted response
+  had different content. A retained sample absent from the complete peer can
+  prove modification. Two partial groups cannot establish response
+  correspondence from their different retained samples alone.
   WebSocket stream inventories and connection counts likewise require complete
-  WebSocket coverage. Redirect resource types use only shared known metadata:
+  WebSocket coverage. Retained frame positions are compared only with complete
+  connection/frame inventories; matching direction/opcode sequences alone do
+  not prove alignment after drops. Redirect resource types use only shared known metadata:
   an omitted type leaves equality unknown rather than proving a modification.
-- Console observations with a stack source retain call type, argument types, timestamp, and source location. When console text capture is selected, already-delivered primitive values are retained verbatim; objects, getters, and remote properties are never expanded.
-- WebSocket observations retain direction, opcode, and payload byte length. When shape capture is selected, text frames are classified as text or complete value-free JSON shape; binary bytes, hashes, prefixes, and raw frames are never retained.
-- Storage observations always redact values. Key names, IndexedDB names, and cache names require `include_storage_keys`; stable content fingerprints require the additional `include_storage_fingerprints` selection. Cookie fingerprints cover cookies applicable to the current main-frame URL, including its path, rather than every cookie path on that origin. Other storage remains scoped to the origin. Cache bodies above 64 KiB and partial IndexedDB remote objects make fingerprint coverage incomplete, so identical observations remain `unknown`.
+- Console observations with a stack source retain call type, argument types, the CDP `Runtime.Timestamp` in milliseconds since Unix epoch, and source location. An absent or invalid timestamp is `null` and makes `console_events` incomplete; an observed zero remains zero. Argument types are listed by CDP argument position. An absent or non-array argument inventory is `null`, and a malformed argument keeps its position as `unknown`; either makes `console_events` incomplete, while an observed empty inventory remains `[]`. A malformed stack or top stack frame is not replaced by its caller: the console event is excluded, and a request initiator location or WebMCP `registration_source` is `null`; each is reported as `invalid_protocol_value`. This epoch-millisecond clock is separate from the monotonic-second timestamps in `observe-web-session`. When console text capture is selected, already-delivered primitive values are retained verbatim with their original `argument_index`; objects, getters, and remote properties are never expanded. Text capture is `unavailable` when the argument inventory is unknown.
+- WebSocket observations retain direction, opcode, and payload byte length. A missing or malformed CDP opcode is returned as `null`, with opcode-dependent payload byte length and shape also `null`, and marks `websocket_frames` incomplete. Missing or wrong-typed payload data is not treated as empty; its byte length is `null`, and a text payload shape is unknown. Numeric opcode `0` and an observed empty payload remain distinct from missing data. Opcode is nullable in the public inspection schema, so clients must handle unknown protocol values. When shape capture is selected, text messages are classified as text or complete value-free JSON shape; binary bytes, hashes, prefixes, and raw payloads are never retained.
+- Storage observations always redact values. Key names, IndexedDB names, and cache names require `include_storage_keys`; stable content fingerprints require the additional `include_storage_fingerprints` selection. Cookie fingerprints cover cookies applicable to the current main-frame URL, including its path, rather than every cookie path on that origin. Other storage remains scoped to the origin. Cache fingerprints hash all decoded response bytes, including bodies above 64 KiB, subject to the [64 MiB per-message CDP limit](#request-boundary). Missing or undecodable cache bodies and partial IndexedDB remote objects make fingerprint coverage incomplete, so identical observations remain `unknown`.
 - Unavailable or malformed storage inventories mark `storage_keys` unavailable, even without fingerprints. Comparisons suppress inferred key additions and removals while preserving modifications proven by complete fingerprints on both sides. Previously exported captures are not retroactively corrected; recapture them to obtain these coverage markers.
 - Script metadata is included only when CDP supplies a URL on an allowed origin. Stable keys exclude transient CDP script IDs, and exact transient raw URLs are used only during script/resource reconciliation. URL-less scripts are excluded because their origin cannot be established. When CDP supplies an execution-context association, the accepted script retains its authorized frame ID for later attribution. Script metadata admission bounds retained identity and source-map strings to 8 MiB and 50,000 records. When a source-map URL cannot fit, the script identity remains and source-map coverage reports `resource_budget_exhausted`; rejected script observations are counted without retaining their IDs. Source content is omitted unless explicitly requested and is returned inline as a self-verifying artifact with its SHA-256 digest.
 
@@ -331,6 +426,57 @@ validation and resource bounds. Omitted script metadata makes its affected
 section explicitly truncated.
 Disallowed-origin entries are filtered before retention.
 
+Script metadata and selected source reads use one fixed admitted-script snapshot.
+New in-scope script IDs still retained when the final CDP read completes are
+counted in `scripts.total` and `completeness.dropped_events.scripts`, with
+`scripts` marked truncated; they do not extend the source-read pass. Repeated
+late events for the same ID count once. The original snapshot is not replaced
+or reduced by later updates to an existing ID. These omissions do not imply
+that a source byte or time budget was exhausted.
+
+Selected script-source collection shares a 32 MiB aggregate UTF-8 retention
+budget and the `command_timeout_ms` deadline (default 20000 ms) across the
+capture. The byte budget has no runtime override. An oversized source is
+omitted while later smaller sources may still fit. When the deadline expires,
+remaining source reads are skipped. Script metadata, per-item omission reasons,
+and `script_sources` completeness remain available. A single CDP response can
+still be materialized up to the separate 64 MiB transport limit before source
+admission; the retention budget does not bound that transient allocation.
+
+Bundle parsing and semantic inspection run sequentially in an owned child,
+without executing captured code. These are separate phases. `max_heap_mb`
+defaults to 1024 MiB V8 old-space (range 128..16384). `analysis_timeout_ms`
+defaults to 30000 per script, including startup and transfer.
+`total_analysis_timeout_ms` defaults to 300000 across script analysis. A
+per-script deadline above the effective total is rejected. A failed or
+timed-out script becomes an explicit unknown while completed findings and later
+scripts remain available. Total-deadline expiry returns a typed timeout with
+partial Evidence and identifies unstarted scripts. `source_map_fetch_timeout_ms`
+defaults to 30000 and is an item limitation, not that analysis timeout. These
+controls are also available as CLI flags with hyphenated names. Worker protocol
+and result byte budgets have no runtime override.
+
+One web-bundle capture/analysis is admitted per REA process at a time. Completed
+script results share an 8 MiB serialized transfer budget, with host heap
+reservation checked before decoding and validating another result.
+When the host reservation is what reaches zero, the failure names that heap
+headroom and the remaining fixed budget. When the 8 MiB allowance itself is
+exhausted, the failure names that fixed budget and says it has no runtime
+override. Either stop ends further script launches and returns a typed
+resource failure with completed facts and explicit unstarted-script unknowns.
+Worker control frames are limited to 64 KiB. Cleanup must confirm child termination
+before releasing its private files; failed cleanup retains its exact owner and
+blocks replacement until retry succeeds. Cancellation retains completed bundle
+facts, including cancellation during final cleanup.
+
+The worker heap setting is not a process-family RSS limit. Node enforces V8
+old-space flags; Bun ignores them. Windows Bun instead uses an owned process
+commit limit with a 512 MiB runtime reserve; other Bun hosts have no enforced
+memory limit. An allocation failure before that Windows limit is reported as
+reached still names the configured limit and says it was not reached. These
+bounds do not establish an aggregate host-memory limit across other REA
+operations or external processes.
+
 Resource-tree inventories distinguish a valid empty `resources` array from a
 missing or malformed collection. Invalid resource entries are omitted with
 incomplete coverage, while valid resources from that frame and other frames
@@ -342,13 +488,31 @@ An explicitly empty MIME type is valid. The optional content size may be absent;
 an invalid supplied size remains unknown with incomplete resource coverage while
 the resource's other facts are retained. Zero and nonnegative fractional sizes
 are preserved.
+`resource_key` remains the exact observation identity used by script
+reconciliation. Capture comparison groups resources by URL, origin, type and
+MIME type, then compares known sizes within a group. If a matched group contains
+an unknown size, comparison reports unknown rather than treating the observation
+as a resource addition or removal; complete inventory coverage still establishes
+when an entire group is absent.
 
-CDP discovery has a 10-second deadline. Each WebSocket handshake and command
-response has its own 20-second deadline, separate from the caller-selected
-observation window. A stalled handshake or command reports `provider_timeout`;
-caller cancellation still reports cancellation. WebSocket messages remain subject
-to the [64 MiB limit](#request-boundary). REA does not impose a message-count or
-pending-command quota. On cancellation, it closes only its own WebSocket,
+CDP discovery has a 10-second response deadline. HTTP discovery documents
+use an 8 MiB budget. Browser-socket discovery (`Browser.getVersion`,
+`Target.getTargets`) uses the 64 MiB protocol message budget. Neither budget
+nor the discovery deadline has a runtime override. `list_browser_targets` accepts no
+`command_timeout_ms`: HTTP discovery and browser-socket discovery
+(`Browser.getVersion`, `Target.getTargets`) stay on that fixed deadline, and
+the socket closes without a later command. Other CDP tools use
+`command_timeout_ms` (default 20000) for each WebSocket handshake and command
+response, separate from the caller-selected observation window, discovery, and
+the 5000 ms cleanup grace. A browser-level socket's handshake and the following
+`Target.attachToTarget` each receive that full deadline; they do not share one
+budget. A stalled handshake or command reports
+`provider_timeout` and names `command_timeout_ms` only when that operation
+accepts it; caller cancellation still reports cancellation. WebSocket messages
+remain subject to the [64 MiB limit](#request-boundary), which has no runtime
+override. REA does
+not impose a message-count or pending-command quota. On cancellation, it closes
+only its own WebSocket,
 waiting up to one second before terminating it. Malformed unsolicited events
 and malformed replies to pending commands poison the connection so later
 commands fail closed. A correlated reply's session ID must match the command's
@@ -357,7 +521,7 @@ rejects it as unknown; other command rejections retain their reported failure.
 
 Network and console coverage starts only after REA attaches and enables the relevant CDP domains. `prior_activity_available` is always `false`; absence from these arrays is not evidence that an event never occurred. Scripts are different: scripts the page parsed before REA attached are still reported under `scripts`. A page that loaded before attach can therefore list its scripts while `network.requests` lacks the requests that fetched them. If CDP reuses a request ID without supplying redirect metadata, REA retains the predecessor observation and marks network coverage incomplete.
 
-Source maps are not fetched by inspection. Bundle analysis fetches them only when `fetch_source_maps` is selected, with exact-origin redirect checks and no credentials, cookies, or referrer. Each fetch operation has a 30-second deadline and a 64 MiB aggregate streamed-response budget across redirects. A deadline or byte-limit failure remains explicitly unavailable and makes bundle completeness partial. Source-map decoding also bounds raw resolved identities to 64 MiB and aggregate inline evidence to 32 MiB. Failed maps preserve their failure context when it fits; if another context cannot fit, collection stops with the full requested count, the returned processed count, and an explicit unknown-results limitation. Retained mappings and original sources are returned inline. See [source-location limits](web-source-location.md) for the structural decoder bounds. Original source artifacts with `.ts`, `.tsx`, `.mts`, or `.cts` suffixes carry `text/typescript`; other source names retain `text/javascript`.
+Source maps are not fetched by inspection. Bundle analysis fetches them only when `fetch_source_maps` is selected, with exact-origin redirect checks and no credentials, cookies, or referrer. Each fetch operation uses `source_map_fetch_timeout_ms` (default 30000) and a 64 MiB aggregate streamed-response budget across redirects. The byte budget has no runtime override. A fetch overrun is an item limitation, not a thrown analysis timeout. A deadline or byte-limit failure remains explicitly unavailable and makes bundle completeness partial. Source-map decoding also bounds raw resolved identities to 64 MiB and aggregate inline evidence to 32 MiB. Failed maps preserve their failure context when it fits; if another context cannot fit, collection stops with the full requested count, the returned processed count, and an explicit unknown-results limitation. Retained mappings and original sources are returned inline. See [source-location limits](web-source-location.md) for the structural decoder bounds. Original source artifacts with `.ts`, `.tsx`, `.mts`, or `.cts` suffixes carry `text/typescript`; other source names retain `text/javascript`.
 
 source-map items retain the script's `declared_url` separately from the observed
 `effective_url`. successful maps use the latter as their source-resolution base.
@@ -369,6 +533,10 @@ Current fetches emit a string or null; historical observations may omit
 An observed URL must fit before its response body is read. If it cannot fit,
 remaining map results are explicitly unknown; failed cancellation retains its
 cleanup context and resource constraint.
+Indexed maps whose decoded mappings cross the next section's starting position
+are invalid under [ECMA-426](https://tc39.es/ecma426/2024/), rather than complete
+maps with clipped mapping rows. Fetch collection and source tracing share this
+validation.
 
 ## Non-goals and threat model
 

docs/browser-scenario-contract.md

@@ -23,9 +23,21 @@ that rejection does not roll back browser effects. REA removes its owned
 future-document scripts, restores live frame methods, then disconnects.
 Add secret declarations, storage seeds, and additional artifact or event
 capture only when needed. Action, secret, and storage counts are not capped.
-Duration, action, and navigation timeouts remain fixed provider-owned liveness
-limits. `capture.limits` controls retained event and network-content budgets;
-it does not change those liveness deadlines.
+Duration, action, and navigation timeouts remain separate from CDP command
+deadlines. `command_timeout_ms` (default 20000; omit it to keep that default
+and leave it out of Evidence) bounds launch startup, the connect-over-CDP
+handshake, target lookup, environment commands, and storage CDP commands,
+including replies from attached child targets. Set it in the scenario JSON.
+`capture-browser-scenario` has no `--command-timeout-ms` option.
+It does not change action `timeout_ms`, `wait_for_timeout.duration_ms`, or
+navigation. Service-worker restoration keeps a fixed 5000 ms grace and does
+not inherit a raised `command_timeout_ms`. CDP discovery stays on its own
+fixed deadline. When service workers are blocked, the registration check
+before discovery has its own fixed 10000 ms deadline; `command_timeout_ms`
+does not apply because discovery has not started. The discovery deadline
+covers endpoint discovery only. The following WebSocket handshake and
+`Target.attachToTarget` each use `command_timeout_ms`. `capture.limits` controls retained event and network-content
+budgets; it does not change those liveness deadlines.
 
 ## Minimal launch example
 
@@ -107,6 +119,11 @@ Through MCP, pass the same object as the `capture_browser_scenario` arguments.
 The final step's DOM and accessibility artifacts retain the page state at that
 capture time inline. REA still launches and cleans up the browser profile.
 
+Action `timeout_ms` and `wait_for_timeout.duration_ms` accept integers from 1
+through 2,147,483,647 milliseconds, matching the underlying timer representation.
+Larger values receive an input error before browser launch. Omit `timeout_ms`
+to keep an action pending until completion or request cancellation.
+
 The click's `timeout_ms` covers the click action. A non-navigating `fetch` or
 delayed render can finish later, so use `wait_for` to observe a page-specific
 ready condition before final capture. A page that updates synchronously can
@@ -144,6 +161,36 @@ distinction: it closes and deletes provider-owned profiles, but only disconnects
 from an external CDP browser. Real-browser verification checks both outcomes and
 confirms that an attached external browser remains alive.
 
+That cleanup requires REA to remain running. Abrupt termination, including
+Windows process-tree termination or POSIX `SIGKILL`, can leave the launch
+profile and Playwright artifacts behind. A profile can retain cookies, history,
+cache and origin storage. Later REA runs currently do not reclaim these paths.
+Playwright 1.58.2 can also leave an artifacts directory when launch preparation
+fails before a browser process exists, such as for a missing executable.
+POSIX wrappers can request awaited cancellation with `SIGINT` or `SIGTERM`;
+Windows process termination is not an equivalent graceful-cancellation channel.
+Do not identify owned artifacts by directory prefix or creation time alone.
+See [#4659](https://github.com/morluto/rea/issues/4659) for the remaining recovery
+and wrapper-cancellation work.
+
+Keeping that browser alive does not guarantee its emulation state survives.
+Chromium can clear caller-owned `navigator.platform` emulation when a page CDP
+session disconnects, including sessions that sent no emulation commands. REA
+cannot snapshot the other client's complete user-agent/client-hint settings and
+does not replay a partial reconstruction. Use launch mode to keep this effect
+inside a provider-owned browser, or have the external owning client reapply its
+complete configuration after capture. Connected captures report this limitation;
+see [issue #4089](https://github.com/morluto/rea/issues/4089).
+
+`environment.locale` sets Intl formatting in both modes. Launch also sets
+`navigator.language`, `navigator.languages`, and the request `Accept-Language`.
+Connect preserves the caller's page and request languages; every attached
+capture identifies this locale limitation, including retained partial
+observations. Use launch mode when the application must select the requested
+language through navigator values or request headers. Artifact completeness
+covers the selected observations, not equivalence of launch and connect
+language environments.
+
 HTTP(S) URLs may include ordinary query values and fragments directly. Use
 ordered structured query entries when a value needs to reference a declared
 environment-backed secret. Form, storage, and cookie values are either literal
@@ -177,7 +224,9 @@ statements remain under the attached debugger client's control.
 The result is Evidence with an initial state followed by one record per
 declared action. Each step reports action status, elapsed time, sanitized URLs,
 event bounds, and independently typed capture states for screenshot, DOM,
-accessibility, URL, history, and storage. Console, page-error, network,
+accessibility, URL, history, and storage. Actions after a failed or cancelled
+action are not run and are recorded as `cancelled` with a not-executed error.
+Console, page-error, network,
 WebSocket, frame, worker, popup, and cancelled-download events are attributed to
 steps. Retained events and requested artifacts are returned inline; omitted
 events are counted by kind and identify affected network transactions.
@@ -338,6 +387,24 @@ observed) and excludes the sequence reference from normalization.
 Playwright scenario capture does not supply initiator stacks; passive CDP
 observation retains its own initiator evidence.
 
+## Storage value fingerprints
+
+New storage snapshots include `fingerprint_version: 2`. Non-secret, well-formed
+JavaScript strings retain their SHA-256 over UTF-8. Strings with unpaired UTF-16
+surrogates use SHA-256 over a raw `0xff` byte followed by their exact UTF-16LE
+code units; this prefix cannot occur in UTF-8. Declared secrets remain
+`redacted-secret` with a null digest, without a second or legacy digest.
+
+A missing version identifies legacy UTF-8 fingerprints, which replaced unpaired
+surrogates and cannot establish value identity. Storage containing any legacy
+hashed value is conservatively compared as `unknown`, including old-to-old
+comparisons; recapture both sides to compare reliably. This also conservatively
+withholds some otherwise provable legacy differences. Empty or entirely
+secret-redacted storage keeps its existing comparison semantics. The version is
+checked before normalization and excluded from content comparison: changing
+encoding metadata alone does not establish a website change. Existing captures
+and their completeness records are not rewritten.
+
 ## Scenario comparison
 
 `compare_web_captures` and `compare-web-captures` also accept two complete
@@ -370,6 +437,15 @@ sorted canonically before application, making the same policy reproducible
 regardless of input order. Rules operate only on already-redacted durable
 capture fields and are preserved in the result.
 
+Cookies, local storage, and session storage are compared as unordered inventories
+after caller literal rules apply. Enumeration order alone does not establish a
+storage change. Canonical ordering uses the complete entry, preserving cookie
+scope/flags, value fingerprints, and duplicate entries even when redaction or
+literal rules make names collide. Captured storage arrays retain their original
+order. The `compare_storage_entries_as_unordered_collection` built-in rule
+participates in the normalization commitment; new storage comparison digests
+therefore differ from earlier order-sensitive comparisons.
+
 Each aligned step reports changed, unchanged, or unknown. Changed and unknown
 action, screenshot, DOM, accessibility, URL, history, storage, and event
 artifacts carry their normalized before/after digests and capture states.

docs/ci-analysis.md

@@ -0,0 +1,90 @@
+# Static analysis in GitHub Actions
+
+The [minimal workflow](https://github.com/morluto/rea/blob/main/examples/github-actions/rea-analysis.yml) builds a small
+JavaScript application, analyzes its files without executing them, and uploads
+machine-readable results. It uses the existing CLI, not a separate GitHub Action
+or an MCP client. It does not promise support for every REA provider.
+
+## Run the example
+
+1. Copy the example to `.github/workflows/rea-analysis.yml` in your repository and
+   commit it to the default branch. No existing application or package manifest is
+   needed for the self-contained demonstration.
+2. Enable GitHub Actions if necessary. Select **REA static analysis example** in
+   the Actions tab and choose **Run workflow**.
+3. Download the `rea-analysis` artifact after the run. Its files are:
+
+   | File              | Meaning                                                                                            |
+   | ----------------- | -------------------------------------------------------------------------------------------------- |
+   | `analysis.json`   | Original CLI JSON, including complete Evidence on success or the error report when one is emitted. |
+   | `analysis.stderr` | Original diagnostics/progress; an empty file is normal when there are none.                        |
+   | `exit-code.txt`   | Original analysis process exit code.                                                               |
+
+The example pins Node.js 24.22.1, REA 6.7.0, and action commit revisions. It installs
+REA into the runner's temporary directory rather than changing the application's
+dependencies. This pins the tool version, not every transitive npm dependency;
+use a reviewed, committed package lock and `npm ci` if your integration requires
+that stronger dependency-reproducibility guarantee.
+
+## Prerequisites and trust
+
+The runner is GitHub-hosted Ubuntu 24.04 with Bash, npm, and `ps` from `procps`.
+Other Linux images must supply `procps`; REA's supervised process controls use it.
+This static JavaScript operation needs no Hopper, Ghidra, IDA, Java, browser,
+Docker, GPU, external API key, or user-supplied GitHub secret. The workflow only
+requests `contents: read`; artifact upload uses the runner's normal Actions
+credentials.
+
+Replace **Build a known artifact** with your reviewed build, and set
+`ARTIFACT_PATH` to its application directory (relative to the checkout) or an ASAR.
+The demonstration creates its own tiny application; it deliberately throws if
+executed. REA reads it statically. Your replacement build itself can execute code:
+review it separately and do not add untrusted pull-request builds to a privileged
+`pull_request_target` workflow. The example is manually triggered, not an
+untrusted-PR execution policy.
+
+Reports can contain paths, URLs, source-derived facts, and caller-selected data.
+Only analyze artifacts and publish reports you are authorized to share. Do not
+rely on REA inferring secrets from their names. Review artifact access and
+retention for your repository; the example retains reports for seven days.
+
+## Failure, partial results, and unsupported capabilities
+
+The analysis step captures stdout and stderr separately, records `$?`, and exits
+with that same status. There is no `continue-on-error`, `|| true`, output filter,
+or pipeline that could turn a failed analysis green. Before checkout or other
+fallible steps, the workflow creates a fresh report directory under `RUNNER_TEMP`.
+Analysis and upload both use that directory, never a checked-in `reports/rea` or
+other workspace files. The upload step uses `always()` when that directory was
+prepared, so available diagnostics survive an analysis failure. If installation
+or the build failed before analysis began, it is empty and upload warns rather
+than publishing stale results or inventing a successful result. Runner loss or
+cancellation can still prevent upload.
+
+`--json --full-output` selects a machine-readable envelope; it does **not** mean
+that analysis was complete. A successful operation can retain partial analysis,
+unknowns, and limitations. Exit zero is not a security verdict, proof of complete
+coverage, or a policy gate for those findings. Inspect the saved Evidence and
+coverage before making such a decision. See [output and exit status](cli.md#output-and-exit-status).
+
+A nonzero exit can represent invalid input, unavailable tooling, an unsupported
+host/capability, or another operation failure. Keep the original JSON error (when
+emitted), stderr, and exit code, and use their category/message to diagnose it.
+An early startup failure may leave `analysis.json` empty; do not parse that as
+successful analysis. Select a supported operation or explicitly provision its
+[documented prerequisites](installation.md); do not silently switch providers,
+execute the artifact, or ignore the failure to obtain a green job.
+
+To exercise the failure path, temporarily set `ARTIFACT_PATH` to
+`build/missing-artifact` without changing the demo build. Analysis must fail,
+`exit-code.txt` must be nonzero, and the available reports must still upload.
+To exercise unsupported input, set `ARTIFACT_PATH` to `build/rea-ci-demo/index.js`:
+this operation accepts an application directory or ASAR, not a single source file.
+The retained JSON error has category `unsupported_target`; the step exits nonzero.
+Restore the correct path before using the example normally.
+
+To exercise partial analysis, append `export const broken = ;` to the demo's
+JavaScript. REA 6.7.0 retains `data.normalized_result.statistics.parse_failures`
+and an incomplete-coverage limitation while the operation can still exit zero.
+Inspect those facts instead of treating process success as full coverage. The
+report JSON and diagnostics remain unfiltered in all three cases.

docs/cli.md

@@ -3,6 +3,8 @@
 Use the CLI for a direct inspection or a script. It uses the same application
 workflows and evidence contracts as REA's MCP server. Each CLI invocation is a
 separate process; save results when you want to use them in a later command.
+For a pinned, noninteractive workflow with retained reports, see
+[static analysis in GitHub Actions](ci-analysis.md).
 
 ## Run a command
 
@@ -30,6 +32,13 @@ jq -c '{source: {kind: "inline", evidence: .}, view: {kind: "summary"}}' app-evi
 rea inspect-analysis-view app-view.json
 ```
 
+The request can also name the saved file instead of embedding it:
+`{"source": {"kind": "evidence-file", "path": "app-evidence.json"}, "view": {"kind": "summary"}}`.
+The CLI resolves a relative path against the working directory; MCP clients
+pass an absolute path. REA reads and authenticates the file, then projects it
+like retained Evidence, so a large file is bounded by the file reader rather
+than charged as inline input.
+
 For example, a 334 MB Obsidian application Evidence file yields a summary view
 of about 10 KB. Replace the view with
 `{"kind": "page", "collection": "modules", "offset": 0, "limit": 32}` for a
@@ -37,6 +46,119 @@ module page. See
 [JavaScript application workflows](javascript-application-workflows.md) for
 views, feature traces and comparisons that take the same saved Evidence.
 
+## JSON Lines record output
+
+Use `--format jsonl` to consume row-oriented results one record per line:
+
+```bash
+cat > observations-view.json <<'JSON'
+{"source":{"kind":"evidence-file","path":"app-evidence.json"},"view":{"kind":"javascript-facts","property_keys":[],"offset":0,"limit":32}}
+JSON
+rea inspect-analysis-view observations-view.json --format jsonl > observations.jsonl
+```
+
+This request selects up to 32 observation identities and their provenance,
+without observation property values.
+
+This changes JSONL object-collection framing. Scripts that need one complete
+report or a scalar field should use `--format json`. For a full application
+graph, first project a supported analysis view rather than selecting an
+arbitrary nested graph array.
+
+A top-level array emits its elements in source order. An object must have one
+clear `rows`, `items`, `records`, or `occurrences` array, directly or under
+`page`. Evidence uses the same boundaries inside `normalized_result`.
+An arbitrary report or a scalar selection is refused with an actionable error;
+use `--format json` for a complete document. If several collections exist,
+select one with `--filter-output`, for example `normalized_result.rows`.
+
+Object collections finish with a control record:
+
+```json
+{
+  "$rea": "jsonl/v1",
+  "type": "metadata",
+  "records": {
+    "path": ["normalized_result", "rows"],
+    "emitted": 2,
+    "source": 10,
+    "filter": "normalized_result.rows[0,2]"
+  },
+  "context": {
+    "evidence_id": "ev_…",
+    "normalized_result": {
+      "coverage": {
+        "status": "page",
+        "examined": 10,
+        "total": 100,
+        "next_offset": 10,
+        "exhausted": false
+      }
+    }
+  }
+}
+```
+
+This abbreviated example illustrates framing. The actual `context` preserves
+all non-row fields, including Evidence subject, provider, authority, confidence,
+locations, links, limitations, parent/view digests, pagination and coverage.
+With `--full-output`, it also retains the command envelope and its metadata.
+A data filter selects row fields/slices while this context remains available;
+excluded peer record collections and unselected Evidence `raw_result` are not
+restored. An explicitly selected `raw_result` remains selected. Retain the
+original Evidence for complete observations.
+
+`records.emitted` and `records.source` count the selected output and its source
+array; they do not assert analysis completeness. Source coverage and unknowns
+retain their original meaning. Existing row identities and indices remain
+unchanged unless your field selection excludes them. Empty object collections
+still emit metadata; an empty top-level array emits no lines. Bare top-level
+arrays have no synthetic coverage record.
+
+Command and output-validation failures emit `{"$rea":"jsonl/v1","type":"error","ok":false,"result":…}`
+and retain the original diagnostic payload. SDK validation/sentinel errors use
+the SDK's exit status rather than guessing from data fields. `$rea` is reserved
+at the top level of a data record, so control records are unambiguous; collisions
+are refused before rows are written. Nested `$rea` values are ordinary data.
+A missing final metadata record or nonzero exit can indicate an interrupted
+export; do not treat already emitted rows as a complete result.
+
+`--token-count` estimates this exact representation, including metadata, one
+line at a time. Errors remain error records even with token counting enabled.
+Token windows cannot preserve JSON Lines and are rejected. JSON encoding and
+stdout writes are incremental, including large escaped strings; this is output
+streaming over a completed result, not incremental analysis. The analysis still
+needs its normal resources. A single record must fit the runtime string limit
+for token estimation; remove `--token-count` to stream a larger record.
+
+Standalone version and MCP doctor diagnostics remain single JSON documents
+in JSON Lines mode. Incur's streaming generator events and successful sentinel
+responses retain their SDK protocol. MCP application results and Evidence
+contracts are unchanged.
+
+## Shell completion
+
+The CLI includes optional shell completion. Run `rea completions --help` to
+see the available shells and their setup commands. For Bash, load the hook
+into the current shell:
+
+```bash
+eval "$(rea completions bash)"
+```
+
+The current Bash hook requires support for `complete -o nosort`; macOS's
+system Bash 3.2 cannot load it. Use a newer Bash for this hook.
+
+For future interactive non-login Bash sessions, add the command to
+`~/.bashrc`. Interactive login Bash sessions read the first available file in
+`~/.bash_profile`, `~/.bash_login`, and `~/.profile`; add the command to the
+file your shell reads, or have that file source `~/.bashrc`. Remove the command
+from the startup file to stop loading completion, and run `complete -r rea` to
+remove the registration from the current shell. These are manual, opt-in
+steps.
+Completion is a discovery aid; use `rea COMMAND --help` to confirm the
+available options and subcommands.
+
 ## Managed assemblies (.NET)
 
 For a .NET PE/CLI assembly, inspect its identity and deployment form before
@@ -56,10 +178,22 @@ or repeat it to select several types.
 Selection limits decoding to the selected types; member references remain an
 assembly-wide inventory, so the output can still be large.
 
+Select `--detail summary --evidence-bundle /absolute/path/members.json` to
+receive identity, selection, coverage, limitations and collection counts while
+saving the complete original Evidence in a new bundle. Summary requires that
+explicit destination; existing files are preserved. Later
+`inspect-analysis-view` calls can page the exact parent from the saved bundle
+without reopening the assembly. See
+[CLI summary and saved member rows](managed-code-analysis.md#cli-summary-and-saved-member-rows)
+for the complete command sequence. Omitted or `--detail complete` retains the
+full inline response.
+
 Choose the workflow from the artifact's contents rather than its `.exe` or
 `.dll` extension. The managed parser requires PE/CLI metadata; it does not
-unpack .NET single-file hosts or decode IL2CPP metadata. NativeAOT method bodies
-require native analysis. See the [managed-code guide](managed-code-analysis.md)
+unpack .NET single-file hosts or decode IL2CPP. For the image, type and method
+names in a Unity `global-metadata.dat`, use
+[IL2CPP metadata](il2cpp-metadata.md). NativeAOT method bodies require native
+analysis. See the [managed-code guide](managed-code-analysis.md)
 for supported deployment forms, selection semantics, and coverage limits.
 
 ## SQLite snapshots
@@ -83,6 +217,47 @@ The shared CLI/MCP workflow reports compiler version, modules, replacements,
 build settings, original bytes and SHA-256. See [Go binaries](go-binaries.md)
 for supported containers and absent or malformed metadata.
 
+Inspect a Godot PCK without running the game:
+
+```sh
+rea inspect-artifact game.pck --detail summary --evidence_output evidence.json --json
+rea extract-artifact game.pck --json
+```
+
+Formats 2, 3 and 4 retain header, directory and integrity facts. Compact delivery
+keeps complete Evidence in the requested bundle. See [Godot PCK](godot-pck.md)
+for embedded packs, batch queries and unknown encrypted coverage.
+
+Inspect local model declarations without loading weights:
+
+```bash
+rea inspect-model-artifact /absolute/path/model.gguf --json
+rea inspect-model-artifact /absolute/path/model.safetensors --hash full --json
+```
+
+Default hashing reads only headers and leaves full-file identity unknown.
+`--hash full` streams the complete file for SHA-256. Both return typed metadata,
+original tensor declarations and derived totals. MCP summaries retain complete
+Evidence for filtered metadata, tensor and type-total pages; CLI compact delivery
+uses `--detail summary --evidence_output FILE`. See [model files](model-artifacts.md).
+
+Inspect Unity IL2CPP metadata names without Unity or a disassembler:
+
+```bash
+rea inspect-il2cpp-metadata /absolute/path/global-metadata.dat --json
+```
+
+The command decodes a version 24, 27, 29, 31 or 39 `global-metadata.dat` and reports assembly
+images, type, method and member names, string literals, index ranges, header
+sections and SHA-256. With `--binary PATH` it also reads the matching
+`GameAssembly.dylib`, `GameAssembly.dll` or `libil2cpp.so` (never runs it) and
+links each method to its code address; `--ghidra-seeds PATH` then writes Ghidra
+function seeds. It does not decrypt metadata. MCP defaults to compact counts
+with retained Evidence; query indexed rows with filters, batch indices and
+selected fields. The CLI defaults to complete Evidence;
+`--detail summary --evidence_output FILE` exports the full result to a new file
+and returns a compact overview. See [IL2CPP metadata](il2cpp-metadata.md).
+
 Inspect an embedded cargo-auditable Rust dependency list without installing
 Rust or configuring a disassembler:
 
@@ -95,6 +270,49 @@ dependency edges, and the section's file range. A missing `.dep-v0` section is
 null metadata, not proof that the binary has no dependencies. See
 [Rust binaries](rust-binaries.md).
 
+Inspect a PE Authenticode certificate table without executing the image or
+verifying signatures:
+
+```bash
+rea inspect-pe-certificates /absolute/path/to/application.exe --json
+```
+
+The workflow reports `WIN_CERTIFICATE` entries and, for PKCS#7 signedData,
+embedded X.509 serial, issuer, subject, and validity fields. A missing security
+directory is observed absence of that table. See
+[PE Authenticode certificate tables](pe-certificates.md).
+
+Inspect ELF PT_NOTE and SHT_NOTE tables without executing the image:
+
+```bash
+rea inspect-elf-notes /absolute/path/to/program --json
+```
+
+The workflow reports note owners, types, descriptor hashes, and GNU build-id
+or ABI-tag fields when those descriptors match. A missing note table is
+observed absence of those declarations. See [ELF note tables](elf-notes.md).
+
+Inspect a CPython 3.7–3.13 `.pyc` header and nested marshal code objects
+without executing bytecode or installing Python:
+
+```bash
+rea inspect-python-pyc /absolute/path/to/module.pyc --json
+```
+
+The workflow reports the pinned CPython version for the magic, PEP 552 hash or
+mtime/size fields, nested function names, and bytecode SHA-256. Bytecode is
+not decompiled. See [CPython pyc inspection](python-pyc.md).
+
+Inspect an Erlang BEAM file without executing bytecode:
+
+```bash
+rea inspect-erlang-beam /absolute/path/to/app.beam --json
+```
+
+The workflow reports IFF chunk names, atom tables, exports, imports, lambdas,
+and the Code chunk header. Missing optional tables stay null. See
+[Erlang BEAM files](erlang-beam.md).
+
 Configure [Hopper or Ghidra](installation.md#hopper), the
 [IDA adapter](ida-provider.md), or [Kuna](kuna-provider.md) before analyzing a
 native target. Substitute
@@ -154,7 +372,7 @@ The selected provider remains bound to the session until an explicit switch
 or close. Provider failures are returned with their original reason. For an
 `ambiguous` selection error, choose from `details.candidate_ids`; for
 `provider_unavailable`, follow `message`, `remediation.action`, and the original
-reasons in `details.rejections`. For Hopper, Ghidra, IDA, and JADX, run
+reasons in `details.rejections`. For Hopper, Ghidra, IDA, JADX, and Kuna, run
 `rea doctor --provider ID --json` to diagnose the selected engine. Other adapters,
 such as `pwntools-elf`, report their configuration requirements directly and are
 not accepted by `doctor --provider`. See
@@ -168,6 +386,12 @@ be verified. See [MCP contracts](mcp-contracts.md) for session lifecycle and
 [Ghidra first-query deadlines](mcp-contracts.md#ghidra-first-query-deadlines-and-recovery)
 for import, client timeouts and recovery.
 
+If one-shot analysis fails and closing its session also fails, the original
+error details remain inline and `details.cleanup_error` describes the later
+close failure. Each diagnostic retains its own reported resources; the later
+failure does not establish whether an earlier resource was released. Captured
+output and partial observations remain in the original details.
+
 ## Save and reuse analysis snapshots
 
 A snapshot retains successful analysis results for later queries. REA reuses
@@ -175,7 +399,11 @@ an exact result when the target bytes, operation, parameters, provider and
 settings match. Mutations and cursor-dependent calls are excluded from the
 cache. Snapshot files are local and use owner-only permissions.
 
-Dispatching a metadata write invalidates the active session's snapshot cache,
+Importing an Evidence bundle refreshes retained subject display names and paths.
+It preserves cached analysis results, including observations for other targets.
+Snapshot bindings that use a replaced record retain its updated display path.
+
+Dispatching a provider metadata write invalidates the active session's snapshot cache,
 even if its reply fails: the provider may already have applied the write.
 Close and reopen the target before importing or saving another snapshot.
 
@@ -223,7 +451,12 @@ Exports preserve an existing destination unless `--overwrite` is explicit.
 For application traces and version comparisons, see
 [JavaScript application workflows](javascript-application-workflows.md).
 MCP [retained Evidence references](mcp-contracts.md#retained-application-evidence-inputs)
-belong to one connection. Separate CLI calls consume full saved records.
+belong to one connection. Separate CLI calls can consume full saved records.
+`inspect-analysis-view REQUEST.json --evidence-bundle BUNDLE.json` resolves
+the request's `retained-evidence` source ID from that authenticated bundle,
+including bundles exported from an MCP session. The same view JSON works for
+every supported parent kind; no target is reopened and no record is selected
+by file order. Omit `--evidence-bundle` when the request supplies inline Evidence.
 
 ## Import historical source
 
@@ -280,9 +513,37 @@ reported separately as configured-secret exclusions.
 Re-import saved graphs whose pattern-based exclusions lack `pattern` before
 using them in source-to-bundle comparisons.
 
+Entries that change during an import retain an unavailable-entry diagnostic
+alongside readable siblings. If enumeration of an already admitted root fails,
+the graph keeps its observed children and records the root failure as a
+top-level limitation. That partial inventory does not establish absence among
+unexamined children. POSIX directory-entry bytes that are not valid UTF-8 cannot
+be represented as graph paths; the containing directory (or root) is marked
+unsupported, and REA does not access the entry through a replacement-character
+name. Symlink targets are read as bytes too: a target that is not valid UTF-8
+is retained as an unreadable symlink with its bytes reported in the limitation,
+and is never resolved through a replacement-character path. Readable siblings
+remain inventoried.
+
 JavaScript and TypeScript import parsing requires valid UTF-8. Malformed source
 bytes retain their original hashes and sizes with a decoding diagnostic; REA
-does not infer module targets from replacement characters.
+does not infer module targets from replacement characters. Exact dynamic-import
+strings, expression-free templates and runtime-transparent TypeScript wrappers
+retain their targets; computed imports remain unknown. TypeScript import-equals
+dependencies remain `requires` relationships whether or not they are exported.
+
+REA derives historical CommonJS paths only when the selected loader and its
+base are proven: this includes the exact
+`createRequire(new URL("./file.cjs", import.meta.url))` form and
+`require.resolve` with `{ paths: [__dirname] }`. Other custom bases and search
+options remain partial with unknown resolution and the literal requested
+specifier. Overwritten resolver methods retain that uncertainty; they are not
+attributed to Node. `require.main` is a module object, so calling it does not
+establish a loader reference. These relationships do not establish execution;
+Node's [`require.resolve`](https://nodejs.org/docs/v24.22.1/api/modules.html#requireresolverequest-options)
+performs lookup without loading the target. An excluded `package.json`
+also leaves package-directory selection unknown rather than treating the
+manifest as absent and applying an `index` fallback.
 
 Historical-source import requires safe no-follow file opens on Linux or
 macOS. Native Windows returns `unsupported_host`; use Linux REA inside WSL
@@ -304,6 +565,158 @@ Choose the guide for the target and intended interaction:
 Runtime requests name the target and actions. Launched targets run with your
 user permissions; consult the chosen guide for host requirements and effects.
 
+## Serve MCP over streamable HTTP
+
+`rea mcp` speaks stdio to the client that launched it. To serve the same MCP
+catalog over streamable HTTP for one local client, such as an agent environment
+runtime, start the dispatcher entry with `--http`:
+
+```sh
+rea mcp --http --port 8765
+```
+
+`--host` and `--port` default to `MCP_HOST` and `MCP_PORT` from the environment,
+then to `127.0.0.1` and an ephemeral port. The bound URL is written to stderr as
+one JSON log line (`"msg":"MCP transport listening"`) whatever `REA_LOG_LEVEL`
+is set to; nothing is written to stdout. `rea mcp --http --help` lists the accepted options. The listener serves
+one client at a time: a session ends with the client's `DELETE`, and a client
+that connects while another session is open takes it over, so a client that
+exited without ending its session needs no restart. The listener runs until
+SIGINT or SIGTERM and does not stop when stdin closes. See
+[transports](mcp-contracts.md#transports) for the session, `Host` header and
+shutdown contract. Client registrations written by setup and the MCP registry
+entry keep stdio.
+
+## Serve an environment manifest
+
+`rea environment-serve <manifest.json> --evidence <capture.json>
+[--replay-only] [--host H] [--port N]` serves an environment manifest
+projected by `inspect-analysis-view` as an agent environment: an MCP server at
+`/mcp`, the environment card at `/.well-known/agent-env.json`, and a
+`/agentenv` data plane. The manifest's `parent_evidence_id` must equal the
+evidence file's `evidence_id`; otherwise the command refuses with
+`invalid_request` before binding. Before binding it derives the
+`environment-state` view of the capture with the manifest's `template_rules`,
+as `inspect-analysis-view` would, and serves from it: read tools whose views
+are state-backed answer from a live state, write tools are served only where
+the capture's later reads confirmed their effect, and
+`/agentenv` serves `data/reset`, `data/add` and `data/get`. When the state
+cannot be derived, or its catalog exceeds the discovery reservation, it logs a
+warning with the error code and message and serves replay; `--replay-only` serves replay without deriving. In replay every
+manifest tool replays the retained capture records it cites and `/agentenv`
+answers every method with JSON-RPC `-32601`. The stopped document's `state`
+member is `{"served": true, "evidence_id": …}`,
+`{"served": false, "reason": "replay-only"}` or
+`{"served": false, "reason": "unavailable", "code": …, "message": …}`; see
+[serve from the state](agent-environments.md#serve-from-the-state).
+
+```sh
+rea environment-serve manifest.json --evidence capture.json --port 8765
+```
+
+`--host` and `--port` resolve like `rea mcp --http`: flags, then `MCP_HOST` and
+`MCP_PORT`, then `127.0.0.1` and an ephemeral port. `ENVIRONMENT_NAME` renames
+the served environment; set but blank, it is refused as `configuration_invalid`
+like a blank `MCP_HOST`. The bound URL is written to stderr as one JSON log line
+(`"msg":"MCP transport listening"`). See
+[agent environments](agent-environments.md) for what the state and replay
+return and do not establish.
+
+## Export an environment bundle
+
+`rea environment-export <manifest.json> --evidence <capture.json> <dir>
+[--package-tarball <path>] [--name <override>]` writes the same manifest and
+capture pair as a bundle folder the agentenv framework runs with
+`agent-env run <dir>`. It reads each file once, authenticates and cross-checks
+the pair with the same checks as `environment-serve`, checks the bundle name,
+then prepares the state under that name as the bundle's server will at startup
+to predict what the bundle serves (replay when the state is unavailable),
+copies the bytes it authenticated, refuses to write into a directory that
+exists and is not empty, and never overwrites a file.
+
+```sh
+rea environment-export manifest.json --evidence capture.json ./grafana-example --json
+```
+
+The folder holds a `README.md` with the bundle's provenance, an env folder
+`envs/<name>/` with a `Dockerfile` that installs this exact `rea-agents`
+version and serves the copied `manifest.json` and `evidence.json` (written
+with mode `0600` on POSIX; Windows retains inherited ACLs) with
+`environment-serve` (from the state by default), a Python verifier under
+`artifacts/check_<name>/` that checks the served tool count and one served
+answer, and the smoke task `tasks/smoke.json`. `--package-tarball` copies a packed `rea-agents`
+tarball into the env folder and installs it instead of the registry package,
+which is how the hosted lane proves a checkout rather than a release. `--name`
+replaces the manifest's kebab-case name as the env folder and env id; a device
+name Windows reserves (`con`, `nul`, `com1` and the like) is refused. The
+command prints one JSON document listing the written files with their sizes,
+the pinned package source, the served `tool_count`, and `state` in the shape
+of the serve command's stopped document. See
+[agent environments](agent-environments.md#export-a-bundle) for the folder
+layout and what the smoke task checks.
+
+## Compare an environment with its application
+
+`rea environment-compare <manifest.json> --evidence <capture.json> --probe <probe.json> [--probe <probe.json> ...]`
+compares the environment the pair serves with probe captures of the application it came from. Each probe is capture
+Evidence of one call sequence recorded against the application after a reset; the environment's state is reset before
+each probe, every request to the application's origin is answered as the served environment answers it, and the
+answer is compared with the application's after ids are mapped and clocks and generated values are normalised. The command reads files only and
+makes no network call. It prints the report and exits `0`, or prints a structured `invalid_request` error and exits `1`
+for the same pair refusals as `environment-serve`, a probe that is not capture Evidence, or a browser probe captured
+without response bodies, or without the request bodies its writes need. See
+[agent environments](agent-environments.md#measure-fidelity-against-the-application) for the report.
+
+```sh
+rea environment-compare manifest.json --evidence capture.json --probe probe.json --json
+```
+
+## Report version and runtime
+
+`rea --version` prints the package version. Add an output format to get the
+version with the runtime and platform context that a support request needs:
+
+Version reports do not initialize analysis providers. Invalid global settings
+or output options still report their usual errors.
+
+```bash
+rea --version --json
+```
+
+```json
+{
+  "name": "rea-agents",
+  "version": "6.7.0",
+  "runtime": { "id": "node", "version": "24.22.1" },
+  "platform": "linux",
+  "architecture": "x64",
+  "supported_runtime": {
+    "range": "^22.23.4 || ^24.22.1 || >=26.12.1",
+    "satisfied": true
+  }
+}
+```
+
+| Field                                | Meaning                                                                                                        |
+| ------------------------------------ | -------------------------------------------------------------------------------------------------------------- |
+| `name`, `version`                    | Published package name and version; `version` matches plain `--version`.                                       |
+| `runtime.id`                         | `node` or `bun`.                                                                                               |
+| `runtime.version`                    | Version of that runtime.                                                                                       |
+| `runtime.node_compatibility_version` | Bun only: the Node.js version it reports compatibility with.                                                   |
+| `platform`, `architecture`           | Node.js `process.platform` and `process.arch`, such as `linux` and `x64`.                                      |
+| `supported_runtime.range`            | Supported Node.js range.                                                                                       |
+| `supported_runtime.satisfied`        | Whether this runtime is supported, judged as `rea doctor` judges it; Bun by its Node.js compatibility version. |
+
+The field names are stable. The report contains no paths, environment values,
+user or host names, and it starts no provider and reads no target. `--format`
+selects any CLI output format, such as `--format yaml`. JSON follows the usual
+rule: indented on a terminal, compact when piped or redirected. `--help` still
+takes precedence. With other output controls, `--version` keeps its existing
+behavior: valid controls such as `--filter-output` or `--full-output` print the
+plain version, and invalid combinations report the usual output-flag error,
+such as `UNSUPPORTED_OUTPUT_COMBINATION` for `--json` with `--token-limit`. Use
+`rea doctor` for provider and host readiness.
+
 ## Output and exit status
 
 `rea mcp doctor --json` checks production MCP startup and reports child logs in
@@ -322,6 +735,25 @@ redirected, which keeps large results smaller for agents and scripts; pipe
 through `jq .` for indented files. Output selection and formatting do not change
 operation status.
 
+`--format md` escapes backslashes and pipes in table headers and cells, and
+renders cell line breaks as `<br>`. These are presentation changes; retained
+Evidence and JSON output keep the original values. Table widths are measured
+iteratively, so an admitted row page is not limited by the JavaScript function
+argument count.
+
+`--filter-output` selects fields from successful command data. Classified failures
+retain their complete typed diagnostic projection even when the selector matches
+a field such as `code` or `message`, or names a missing field. `--full-output`
+controls whether the outer `{ok, error, meta}` envelope is included.
+Selected own JSON members, including `__proto__`, remain ordinary output data.
+
+Sibling array slices that select different fields are combined by original row
+index and emitted in source order, regardless of selector order. Selecting a
+whole parent row keeps that row intact and takes precedence over narrower fields
+selected from the same row. For example, `rows[0,2].name,rows[1,3].version`
+combines both fields for overlapping rows, while `rows[0,2],rows[1,3].version`
+retains rows 0 and 1 whole.
+
 For `cutter-command`, `--json-output` requests command-specific JSON from
 Cutter when available. This is separate from `--format json`, which selects the
 CLI result representation; the global `--json` flag selects that same CLI
@@ -338,6 +770,30 @@ output formatting.
 | `1`       | The operation could not complete. Structured output identifies invalid input, permissions, cancellation, timeouts or another failure when available.                                                                                                                                                                                                                                   |
 | `128 + N` | Signal-derived status. An unhandled signal may have ended the process, with the shell or runtime preserving the conventional status; a command may instead have handled the signal, waited for its operation and cleanup to settle, and assigned the same status (`130` for SIGINT, `143` for SIGTERM). In the handled case, it may also have printed a structured `cancelled` result. |
 
+`rea mcp --http` is long-running and prints no structured result. It exits
+`130` on SIGINT or `143` on SIGTERM after closing the listener, the session and
+its providers, and `1` when that cleanup fails.
+
+`rea environment-serve` is also long-running: it serves until SIGINT or
+SIGTERM, closes the listener, then prints one structured
+`{"status":"stopped", …}` document naming the URL, environment and tool count
+and exits `130` or `143` through the shared cancellation handler below.
+Refusals before binding (unreadable files, non-Evidence JSON, capture Evidence
+in a format other than HAR, a manifest or state whose parent is not the
+supplied evidence, an unbindable port) print a structured
+error and exit `1`, as does a listener that fails after binding.
+`rea environment-export` is one-shot: it prints the written files and exits
+`0`, or prints a structured `invalid_request` error and exits `1` for the same
+pair refusals, a non-empty or file-occupied directory, a missing package
+tarball or a name that is not a kebab-case label or is a Windows device name,
+writing nothing in each case.
+A write that fails after the first file (a permission, disk-space or
+pathname-length failure on a later path) propagates as the host's failure,
+naming its cause and, when it failed on one bundle file or the package
+tarball, that file. A failure while syncing the finished folder names only its
+cause. Either way the partial folder stays in place for inspection; remove it
+before retrying, since the next attempt refuses the non-empty directory.
+
 A `cancelled` result does not by itself determine the exit status; it depends on
 the command's signal handling. Commands routed through the shared
 `withCommandCancellation` handler in the CLI set status `130` for SIGINT or
@@ -350,11 +806,20 @@ browser, web and Electron commands (including `analyze-web-bundle`,
 `capture-electron-scenario`), and the platform and provider command families
 for Android/ADB, Apktool/JEB/Flutter, firmware, PCAP and signal analysis,
 WASM/EVM, Go/Rust and binary diagnostics, Frida instrumentation, the Cutter
-command, objdump, Rizin and GDB. In these paths, a returned structured
+command, objdump, Rizin and GDB, and the agent environment server
+(`environment-serve`). In these paths, a returned structured
 `cancelled` result can be printed before the wrapper assigns the
 signal-derived status. This describes the current source routing; it is not a
 claim that every command or provider runtime has been independently exercised.
 
+Commands with `--evidence_output` select their delivered view, then publish the
+complete Evidence bundle atomically. Cancellation before publication leaves no
+new file. A publication that has committed can win a later cancellation; the
+`cancelled` result then keeps the written path and byte count under
+`details.diagnostics.evidence_bundle`. If staging cleanup stays incomplete
+after commit, the `cleanup_incomplete` error reports the same file as
+`details.published_path` and `details.bytes`.
+
 When the reader of piped output closes early, as `rea ... | head` does, the
 CLI stops writing, prints no error and exits `141` (`128 + 13`, the status of a
 SIGPIPE-terminated writer). The operation itself may have completed; only its

docs/client-configuration.md

@@ -14,6 +14,12 @@ preserving your REA client entry. See the
 
 ## GitHub Copilot CLI
 
+Setup preserves an existing `tools` selection in the REA registration, including
+an empty list. New registrations default to `["*"]`. A restricted selection is
+aligned configuration; doctor does not widen it to expose every REA tool. See
+[Copilot's MCP configuration](https://docs.github.com/en/copilot/how-tos/copilot-cli/customize-copilot/add-mcp-servers)
+for tool-selection settings.
+
 For GitHub Copilot CLI 1.0.95, `gpt-4.1` model metadata can block ordinary
 chat with REA's full input-schema profile before sending a model request
 (`compaction_static_context_blocked`; [#1554](https://github.com/morluto/rea/issues/1554)).
@@ -52,10 +58,16 @@ in the file the installed app reads:
   `%APPDATA%` file it still reads; with neither file, setup creates the
   private copy.
 - If `%LOCALAPPDATA%\Packages` holds a `Claude_<publisher ID>` package from
-  another publisher, or cannot be read, REA cannot establish which file Claude
+  another publisher, an `AnthropicPBC.Claude_<publisher ID>` package, or cannot
+  be read, REA cannot establish which file Claude
   Desktop reads. Setup, doctor and uninstall report that instead of falling
   back to `%APPDATA%`. Doctor then reports `claude_desktop` as invalid, and
   uninstall stops before removing anything for any client.
+  The reported `AnthropicPBC.Claude_fnn82j28hfe8t` build can use the `Claude-3p`
+  deployment directory. REA recognizes that package as an unresolved location;
+  it does not assume that it reads the legacy AppData file or automatically
+  register in `Claude-3p`. See [#4432](https://github.com/morluto/rea/issues/4432)
+  for the reported build and manual registration evidence.
 
 Quit Claude Desktop fully before approving setup; closing its window is not
 enough. A running Claude Desktop can rewrite its configuration from memory and
@@ -67,12 +79,17 @@ to rerun it.
 A process that the MSIX app started, such as a terminal in Claude Desktop's
 Code tab, sees the package's private files at their `%APPDATA%` paths even
 though it reports no package identity. Setup compares one of those files at
-both paths. When it finds that redirected view, the app is running, so setup
-and uninstall refuse to change Claude Desktop's configuration; rerun them from
-a terminal that Claude Desktop did not start. If the package has no private
-file to compare, setup and uninstall cannot rule that out and also refuse; open
-Claude Desktop once if it has never run, quit it fully, then rerun. Doctor
-reads the same file from either kind of process.
+both paths: first one from the package's private `Claude` folder, and when
+that folder has none (some installations keep writing the real
+`%APPDATA%\Claude`), one from the package's other private roaming folders,
+which the package redirects in the same way. When it finds that redirected
+view, the app is running, so setup and uninstall refuse to change Claude
+Desktop's configuration; rerun them from a terminal that Claude Desktop did not
+start. If the package keeps no private roaming file at all, setup and
+uninstall cannot rule that out and also refuse. With Claude Desktop fully
+quit, add REA's entry by hand to the file setup names, which is the one the
+app reads, or remove it from both the private copy and the `%APPDATA%` file.
+Doctor reads the same file from either kind of process.
 
 The same process also writes everything under `%APPDATA%` and
 `%LOCALAPPDATA%` into the package's private storage, where other apps never
@@ -147,11 +164,14 @@ headroom. The default is 450000 ms, covering the default 330-second Ghidra
 startup; an explicit Ghidra startup timeout of 2700000 ms produces a 2820000 ms
 tool timeout. The startup allowance follows an explicitly selected Ghidra or
 Hopper provider, and otherwise uses the larger known Ghidra/Hopper deadline.
-An explicit `REA_JEB_MCP_TIMEOUT_MS` request budget also contributes to that
-maximum. For `ida` or `auto` selection, a valid IDA registration's parsed
+An explicit `REA_JEB_MCP_TIMEOUT_MS` deadline, shared by catalog discovery and
+each tools/call, also contributes to that maximum. For `ida` or `auto` selection, a valid IDA registration's parsed
 `timeoutMs` also contributes to that maximum; its default is 300000 ms.
 Setup forwards these explicitly selected REA deadlines and the input schema
-profile into the registration; unrelated process values are omitted.
+profile into the registration; unrelated process values are omitted. An
+explicit `REA_JADX_OPERATION_TIMEOUT_MS` is forwarded with the JADX
+registration but does not contribute to the derived tool timeout; raise
+`toolTimeoutMs` to cover a longer selected JADX deadline.
 This headroom is a practical default, not a guarantee for every analysis
 duration. Setup preserves positive timeout values already in the `rea` entry.
 You can edit `startupTimeoutMs` and `toolTimeoutMs` there; doctor keeps positive
@@ -234,17 +254,58 @@ standalone CLI check.
 
 ## Codex and Devin
 
+REA manages its launcher and selected provider settings in the client
+registration. Codex's server and per-tool approval policies are user choices.
+Setup patches REA's connection fields in the selected file and retains other
+server settings, including `default_tools_approval_mode`, nested
+`[mcp_servers.rea.tools.<name>]` tables, tool allow/deny lists, and unrecognized
+client settings. Doctor accepts Codex's per-tool policy tables. Known malformed
+policy values block setup before it writes the configuration or a backup, and
+the diagnostic identifies the setting to repair. See the
+[Codex configuration reference](https://developers.openai.com/codex/config-reference/)
+for the supported policy values.
+
+| Settings                                  | Setup ownership                                                       |
+| ----------------------------------------- | --------------------------------------------------------------------- |
+| Launcher, transport and activation        | Refreshes the REA stdio connection and enables it.                    |
+| Explicit server environment               | Preserves configured overrides and merges selected provider settings. |
+| Codex positive timeouts and client policy | Retains caller choices; adds missing recommended REA timeouts.        |
+| Unrelated or future server settings       | Retains their parsed values.                                          |
+
+Dry-run applies the same policy validation and writes no files. A successful
+refresh keeps an exact recovery snapshot of the original configuration;
+repeating setup with the same settings performs no write. Uninstall removes
+the owned REA entry, including its policy, while retaining other servers and
+the recovery snapshot. A green registration check establishes known
+configuration alignment; it does not prove that a particular client or fork
+successfully exposes every REA tool.
+
 Codex treats an unset or empty `CODEX_HOME` as `~/.codex`. Its MCP server
 registration supports `startup_timeout_sec` for connection initialization and
 `tool_timeout_sec` for each tool call. New REA registrations use a 30-second
 startup timeout and a tool timeout derived from the known provider
 allowance plus 120 seconds of operation headroom. This yields 450 seconds for
 the default Ghidra allowance; an explicit Ghidra startup timeout of 2700
-seconds yields 2820 seconds. Setup preserves positive caller-selected timeout
+seconds yields 2820 seconds. That 30-second Codex figure is client guidance.
+REA's own phase budgets are `REA_MCP_INITIALIZE_BUDGET_MS` (default 15000),
+`REA_MCP_TOOLS_LIST_BUDGET_MS` (default 5000),
+`REA_MCP_FIRST_CATALOG_BUDGET_MS` (default 20000), and
+`REA_MCP_DOCTOR_DEADLINE_MS` (default 30000). Each is a positive decimal
+integer of milliseconds from 1 to 2147483647. Unset keeps the default.
+Invalid values fail configuration and name the setting. Setup copies a
+caller-selected value into the registration, and a change is read from the
+environment that starts REA, so it needs a restart. Doctor and packaged
+startup probes name the effective milliseconds, whether they came from the
+default or the setting, the default, the floor, and the maximum. REA does
+not enforce Codex's 30 seconds as the doctor deadline. The 120000 ms
+tool-operation headroom and the 64 KiB doctor stderr cap stay fixed. Setup preserves positive caller-selected timeout
 values in the REA entry, and doctor reports a non-blocking advisory when a tool
 timeout is shorter than the known provider allowance. An explicitly selected
 JEB request budget also contributes to the allowance. A configured IDA
 registration's parsed `timeoutMs` contributes for `ida` or `auto` selection.
+An explicit `REA_JADX_OPERATION_TIMEOUT_MS` is forwarded with the JADX
+registration but does not contribute; raise `tool_timeout_sec` to cover a
+longer selected JADX deadline.
 These values are client-side deadlines, so increase them in `~/.codex/config.toml` if a
 workload needs more time. Devin honors `XDG_CONFIG_HOME` for its
 `devin/mcp_config.json` on non-Windows systems.

docs/client-verification.md

@@ -17,8 +17,8 @@ shared form would leave later fields unknown instead. The existing
 `REA_MCP_INPUT_SCHEMA_PROFILE=compact` profile lets callers stay below
 its character limit by advertising reduced nested validation structure, while
 REA continues to validate calls against the complete canonical input schema.
-Compact applies to the server's advertised inputs and is not an automatic
-full-profile fallback; see the [compact input schema profile](mcp-contracts.md#compact-input-schema-profile)
+Compact is the default for the server's advertised inputs; an explicit `full`
+choice is preserved. See the [compact input schema profile](mcp-contracts.md#compact-input-schema-profile)
 for details of its reduced forms.
 
 Custom renderer hosts can raise `inputMaxChars` through `renderToolSignature` or
@@ -457,11 +457,22 @@ disposable HOME and CODEX_HOME under `~/.cache/rea-agent-evaluations/`. Set
 `REA_AGENT_EVAL_ROOT` to choose another parent directory outside the OS temporary
 directory: current Codex refuses helper aliases under `/tmp`, preventing shell
 commands and skill reads even if MCP works. Each owned run directory is removed
-after verification unless fixture retention is requested. The lane
+after verification unless fixture retention is requested. Commands share the
+owned-process supervisor: native ownership preparation has its own bounded
+deadline before each short command budget, and timeouts join bounded process-tree
+and output cleanup before returning. If cleanup remains incomplete, the lane retries its retained
+owner and preserves the run directory (including disposable auth) rather than
+removing files still in use; the error retains ownership and recovery details.
+Command stderr is limited to 32 KiB and combined version output to 8 KiB;
+exceeding these diagnostic budgets fails explicitly. Scenario JSON events stream
+to the evaluator without retaining a second raw stdout copy. Cleanup attempts the
+remaining run tree even if auth removal fails, and reports unresolved cleanup
+together with the original verification failure. The lane
 installs the packaged skill and MCP registration through public `rea setup`,
 rather than copying a skill into a project that also inherits the caller's
 skills and configuration. Only authentication is copied, when present, and it
-is removed even with `REA_AGENT_EVAL_KEEP_FIXTURES=true`. No user configuration,
+is removed after confirmed process cleanup even with
+`REA_AGENT_EVAL_KEEP_FIXTURES=true`. No user configuration,
 rules, plugins or history is copied; shell snapshots and additional agents are
 disabled. Existing API-key environment authentication remains available.
 The native missing-provider scenario preapproves only `open_binary` and

docs/design/limits-and-recovery.md

@@ -0,0 +1,124 @@
+# Admission limits and recovery
+
+How to add a REA-owned admission ceiling, budget, or deadline, and what a caller
+does when one is hit. This is a design rule, not an inventory. Ceilings that
+should become settings are tracked in
+[#6176](https://github.com/morluto/rea/issues/6176). Naming honesty for limit
+errors is [#5910](https://github.com/morluto/rea/issues/5910).
+
+## Name a default, then a setting when operators need headroom
+
+Put the number in a named constant with a unit. Do not leave it as a magic
+number at the call site.
+
+When a workload may need more room, add a setting whose unset value is that
+constant. One module under
+[`src/config/`](https://github.com/morluto/rea/blob/main/src/config) owns the
+default, any floor, the maximum, the schema, the parser, and the recovery
+text. Register that schema on `environmentSchema` in
+[`environment.ts`](https://github.com/morluto/rea/blob/main/src/config/environment.ts).
+`parseConfig` calls `parseEnvironment`. Doctor uses that same parser through
+`nativeAnalysisConfigurationError`, so startup and `rea doctor` accept the
+same values. Add the name to `SELECTED_REGISTRATION_SETTINGS` so `rea setup`
+copies it into the client registration. Unset keeps today's behavior. An
+invalid value is a `ConfigurationError` and projects as `configuration_invalid`.
+Do not fall back to the default.
+
+[`REA_REGEX_MATCHING_DEADLINE_MS`](https://github.com/morluto/rea/blob/main/src/config/regexMatchingDeadline.ts)
+([#6267](https://github.com/morluto/rea/pull/6267)) is the reference. Default
+5000 ms, floor 1 ms, maximum 2147483647 ms, because Node schedules a longer
+`setTimeout` delay as 1 ms. Setup copies the setting. A hit says:
+
+> The effective REA_REGEX_MATCHING_DEADLINE_MS value is 5000 ms (source default; default 5000 ms, floor 1 ms, maximum 2147483647 ms). Set REA_REGEX_MATCHING_DEADLINE_MS in the environment that starts REA to a positive decimal integer from 1 to 2147483647, then restart REA. The worker old-generation heap does not raise this deadline.
+
+[`REA_MCP_MAX_RESPONSE_BYTES`](https://github.com/morluto/rea/blob/main/src/config/mcpResponseBudget.ts)
+is the byte-budget shape: a decimal safe integer with a 10 MiB floor
+(`MINIMUM_RESPONSE_BYTES`), the same shared schema, and no silent fallback. A
+change needs a REA restart and a matching client receive buffer. It does not
+raise REA's reserved string limit or Node's string ceiling. The setup
+allowlist does not copy this setting. A new headroom setting should, following
+the regex deadline.
+
+A per-call budget is not an environment setting. JavaScript `max_heap_mb`
+(CLI `--max-heap-mb`) is an integer from 128 to 16384 MiB, default 1024. It
+applies on the next analysis, which starts a new worker, and does not restart
+REA. It does not raise the parent result budget, transfer ceilings, or fixed
+parser limits.
+
+Host memory and `NODE_OPTIONS` do not raise a REA admission unless that
+admission's recovery says so. The JavaScript parent result budget is one that
+does: its recovery says to set `NODE_OPTIONS` (for example
+`--max-old-space-size=<MiB>`) on the process that starts REA, then restart
+that process. `max_heap_mb` does not raise it.
+
+## Keep a fixed ceiling honest
+
+If the ceiling must stay fixed, say so in the failure. Name the limit and the
+effective value, say this release has no runtime override, and name what does
+not raise it. Do not invent a setting.
+
+JSON values stop at `MAX_JSON_DEPTH` (1000) in
+[`jsonValue.ts`](https://github.com/morluto/rea/blob/main/src/domain/jsonValue.ts):
+"JSON value exceeds maximum nesting depth of 1000. This release has no runtime
+override."
+
+## Recovery text
+
+The string the caller sees includes:
+
+- the limit or setting name
+- the effective value and unit
+- the source (`default` or `setting`)
+- the default, the maximum, and the floor when the setting has one
+- the next action, including where to set it ("in the environment that starts REA")
+- whether a restart is required
+- what the change does not raise
+
+Invalid settings never take this path. They are `configuration_invalid`.
+Doctor reports `runtime-configuration:<setting>` and says to correct it in the
+shell or the MCP registration, then rerun.
+
+## What agents do
+
+Follow `remediation.action`. Change a request field or a `REA_*` setting only
+when that text recommends the change. Restart only when it says the value is
+read at startup. A sentence that a control does not raise the limit is not a
+request to change it.
+
+- Dropped MCP connection: retry the call once only when it is safe to repeat
+  or known not to have started. If it may already have changed the target, or
+  it fails again, tell the user and ask them to restart the MCP client. Do not
+  start with `rea doctor`
+  ([#6353](https://github.com/morluto/rea/pull/6353)).
+- REA could not start its connection: retry once. If it fails again, tell the
+  user and ask them to restart the MCP client, then run `rea doctor`. Do not
+  retry a startup that returns the same unchanged result: tell the user REA
+  could not start, ask them to restart the MCP client, then run `rea doctor`.
+- Unexpected process-command failure: do not repeat the command. Tell the user
+  to check the input files, then run `rea doctor`.
+- `cleanup_incomplete`: the operation is not retryable. Retry that exact
+  cleanup owner, named in `details.resources`, and wait until it confirms
+  release before admitting new work on those resources
+  ([#6357](https://github.com/morluto/rea/pull/6357)).
+- `resource_constraint` for capacity, when `retryable` is true: wait for the
+  active admission to finish, then retry within the reported limits. Do not
+  restart REA to free the slot.
+- `configuration_invalid`: correct the named setting in the environment that
+  starts REA, then start REA again. Run `rea doctor` for the
+  `runtime-configuration:` check. Re-run `rea setup` for that client when the
+  setting is on the registration allowlist, then reconnect so the new process
+  sees it.
+- `invalid_request`: correct the reported fields from the current schema or
+  CLI help. Do not repeat the unchanged call.
+
+A restart is required for an environment setting and for a client still
+running the previous registration. It is not the first step for a transient
+MCP drop, a capacity wait, or retained cleanup.
+
+## Tests
+
+Lock the unset default, one accepted override, rejected floor, maximum, and
+syntax, and the recovery sentence through `parseEnvironment` and `parseConfig`.
+[`regexMatchingDeadline.test.ts`](https://github.com/morluto/rea/blob/main/src/config/regexMatchingDeadline.test.ts)
+is the pattern. Project the failure and assert `configuration_invalid` for a
+bad value and the recovery wording for a hit.

docs/desktop-applications.md

@@ -0,0 +1,151 @@
+# Desktop installation directories
+
+Use `inspect_desktop_application` when an installation directory is the target
+and its runtime is unknown. It reads directory entries, selected headers and
+bounded metadata, then reports runtime candidates, supporting files, component
+path markers and concrete follow-up operations. No native provider or active
+binary session is needed. It does not run the application, extract archives,
+follow child symlinks, contact services or change target files.
+
+## MCP
+
+```json
+{ "path": "/artifacts/MyApplication" }
+```
+
+Default `detail: "summary"` retains complete source Evidence in the session and
+returns candidates with representative supporting paths, counts, coverage,
+limitations and grouped next probes. `normalized_result.parent_evidence_id`
+identifies the complete record. A server without Evidence retention refuses
+summary delivery before reading the target; select `detail: "complete"` to
+receive every retained fact inline.
+
+Read exact files in one request:
+
+```json
+{
+  "source": {
+    "kind": "retained-evidence",
+    "evidence_id": "ev_<parent-sha256>"
+  },
+  "view": {
+    "kind": "desktop-application",
+    "collection": "files",
+    "paths": ["Game_Data/Managed/Assembly-CSharp.dll", "GameAssembly.dll"],
+    "fields": ["path", "read_status", "ranges", "facts", "issues"],
+    "offset": 0,
+    "limit": 16
+  }
+}
+```
+
+Collections are `files`, `omissions`, `candidates`, `components` and `next_probes`.
+Files and omissions accept exact `paths` and batched `file_ids` / `omission_ids`.
+Files also accept `has_issues: true` to select probe diagnostics directly, or
+`false` to select rows without reported issues. This filter uses the retained
+issue facts; it does not re-read targets or infer that uninspected files are healthy.
+When both are supplied, all filters apply; an empty selection returns no rows.
+Omission records retain the observed path, kind and stable `omission_id`; their
+payloads are not read, and other special-entry metadata remains unavailable.
+`candidates` accepts exact `families`.
+Omitted selectors match all retained rows; an empty selector matches none.
+Every collection supports `fields`, `offset` and `limit`; its stable row ID is
+always included. Selection changes delivery, not source inspection. Pages keep
+source coverage and parent Evidence links. Page exhaustion does not establish
+absence in a partial source. Complete Evidence remains accessible through
+`get_evidence` and Evidence bundle export.
+
+The summary's `probe_issues` counts files with reported issues and their retained
+issue entries. It previews at most two such files, including stable `file_id`,
+read status and the first issue. Path/reason previews use at most 256 UTF-16 units
+each without splitting surrogate pairs; `text_truncated` marks shortened text,
+and `preview_truncated` marks additional files. These are display prefixes;
+full caller paths and all reasons remain in Evidence. Counts describe retained
+observations, while source coverage still identifies partial inspection.
+
+Use `inspect_analysis_view` with `collection: "files", has_issues: true` and
+`fields: ["path", "read_status", "issues"]` for exact reasons, with the usual
+paging and parent links. It intersects with exact paths and batched file IDs.
+
+An ASAR inventory probe first selects its exact path with `open_binary`; then
+call `inspect_artifact` on that active target. Directory inspection itself does
+not select or replace the session target. The JavaScript application workflow
+accepts its own selected path directly.
+
+## CLI
+
+The one-shot CLI returns complete Evidence by default. For compact output,
+explicitly retain it in a bundle:
+
+```bash
+rea inspect-desktop-application /artifacts/MyApplication \
+  --detail summary --evidence_output /reports/application.evidence.json --json
+```
+
+This returns `summary` Evidence and an `evidence_bundle` publication receipt.
+Existing output files are never overwritten. Summary delivery requires
+`--evidence_output`, checked before analysis. `inspect-evidence-bundle` selects
+saved records; an exact saved record can also feed `inspect-analysis-view`.
+`--detail complete` returns source Evidence inline and can optionally write the
+same bundle. Export writes only the caller-selected output file.
+
+## Facts and uncertainty
+
+| Family       | Evidence used                                                                                                                                                                                                                                  |
+| ------------ | ---------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- |
+| .NET         | A PE optional header declares a nonempty CLI directory.                                                                                                                                                                                        |
+| Unity Mono   | A managed assembly below `*_Data/Managed` or `Contents/Resources/Data/Managed` has an exact UnityEngine assembly reference.                                                                                                                    |
+| Unity IL2CPP | A recognized IL2CPP metadata magic (with the raw version when complete) coexists with a conventionally named native `GameAssembly.dll`, `GameAssembly.so`, `GameAssembly.dylib` or `libil2cpp.so` image. Build correspondence remains unknown. |
+| Godot        | A `GDPC` header at file start or a validated executable trailer offset.                                                                                                                                                                        |
+| GameMaker    | A bounded `FORM` container starts with a `GEN8` chunk. The scan probes conventional `data.win`, `game.unx`, and `game.ios` data files for that header.                                                                                         |
+| Unreal       | `Content/Paks` path conventions or an IoStore TOC signature; arbitrary `.pak` filenames are insufficient.                                                                                                                                      |
+| Electron     | A `resources/app.asar` or `resources/default_app.asar` path convention; archive validity is unverified.                                                                                                                                        |
+
+When the four-byte metadata magic is present but version bytes are incomplete,
+the fact retains `raw_version: null`; it does not offer an IL2CPP metadata
+inspection probe. A complete magic/version prefix with supported raw version
+24, 27, 29, 31 or 39 offers an `inspect_il2cpp_metadata` follow-up at its
+admitted file path, including when no native library was found. This probe does
+not establish full metadata validity or correspondence to a native binary;
+version 24 inspection still applies only to supported header layouts. See
+[IL2CPP metadata](il2cpp-metadata.md).
+
+Candidates are derived or inferred from these facts and do not prove loading or
+execution. BepInEx, MelonLoader, Easy Anti-Cheat and BattlEye labels identify
+path conventions only; runtime presence stays unknown. No bypass is provided.
+
+`file_id` includes the admitted root namespace, recorded path and available filesystem identity. The caller-selected root path and admitted canonical path/root identity are both retained; `local_path` identifies each observed file for follow-up operations. The selected root may be a symlink; child symlinks are not followed. Exact
+metadata integers remain decimal strings. Read ranges have offsets, byte counts
+and SHA-256 digests; typed facts expose CLI RVAs, metadata/pack versions and
+assembly tokens. Directory and symlink identities unavailable from traversal
+remain null.
+
+The directory's `snapshot_sha256` commits to retained paths, omission kinds,
+available metadata and range digests, composed through the admitted root,
+entry and omission identities. Caller ceilings, coverage counters, diagnostic
+wording and interpretation facts remain in Evidence, outside source identity.
+The same observations under different generous limits share a snapshot digest;
+Evidence IDs still preserve request parameters and reported coverage. Different
+retained entries or read ranges change the source snapshot.
+Omission records share the entry and representation budgets and can be fetched
+independently through their view.
+**It is not a digest of all file contents.** These are sequential stable entry
+observations, not an atomic directory snapshot. Unread bytes remain unknown.
+
+## Resource bounds and reuse
+
+Defaults are 20,000 examined entries and 8 MiB of total probe reads. Callers may
+select up to 100,000 entries and 64 MiB. A separate 32 MiB conservative
+representation budget bounds retained paths and facts before growth. Limits
+and failed probes produce partial coverage. Traversal and reads are serial.
+Each reader admits one inspection at a time; overlapping requests receive a
+retryable capacity refusal without cancelling the admitted caller.
+
+Unity reference decoding stops after a witness in each `Data/Managed` directory;
+other assemblies' references remain unexamined, while their headers are still
+checked. AssemblyRef facts decode names and tokens only; they do not decode or
+hash unrelated key/culture/hash payloads or establish runtime binding. Pack contents, encrypted data, metadata/address correspondence and
+section-embedded PCKs without a trailer are separate capabilities.
+
+Filesystem results are not cached across calls. Retained views reuse validated
+immutable metadata without rescanning files or retaining another inventory.

docs/electron-observation.md

@@ -2,6 +2,12 @@
 
 REA can attach to an Electron/Chromium CDP endpoint and inspect existing `file://` renderer pages without evaluating JavaScript or invoking Electron APIs. The request supplies the endpoint directly; REA accepts only literal-port loopback HTTP endpoints. Selecting an endpoint exposes every Electron target it serves, including local file paths and page metadata.
 
+Disconnecting the inspection's page session can clear caller-owned
+`navigator.platform` emulation inside Chromium. REA cannot restore another
+client's complete user-agent/client-hint configuration. Use a disposable target
+or let the owning client reapply its configuration; see the shared
+[CDP state limitation](browser-observation.md#provider-authorities).
+
 This passive runtime surface is distinct from the target-free static
 [`analyze_javascript_application`](javascript-artifact-reconstruction.md)
 workflow. Static analysis reads a local ASAR or extracted directory without
@@ -27,6 +33,12 @@ encoded path separators, ASAR directories or links, and nonexistent paths are
 rejected. The endpoint exposes every eligible target and its local path
 metadata.
 
+Cancellation stops further file authorization after an already-started filesystem
+operation settles. ASAR descriptor-close failures remain typed incomplete cleanup
+with an exact retryable owner; they are not treated as a missing or unauthorized
+page. A later authorization and provider shutdown retry retained descriptors
+before reporting their release.
+
 ## Workflow
 
 `list_electron_targets` returns every eligible target served by the supplied endpoint in one inline array.
@@ -50,12 +62,27 @@ rea inspect-electron-page http://127.0.0.1:9223 TARGET_ID \
 
 The normalized result contains canonical local paths, frame and DOM
 structure, resource metadata, stable script/resource identities, explicit
-completeness, and content-addressed script artifacts. Script metadata
-retains its execution-context frame ID when CDP supplies one. The capture also
+completeness, and content-addressed script artifacts. Resource sizes are unknown
+when omitted or invalid; supplied invalid sizes also produce an explicit
+completeness diagnostic.
+
+DOM node indices retain their snapshot positions. Invalid rows are omitted with
+completeness diagnostics, so `total_nodes` can exceed the retained node count
+and a parent index can refer to an omitted source node.
+
+Script metadata retains its execution-context frame ID when CDP supplies one and
+the observed frame generation is authorized. Frames describe the current inventory;
+script observations can retain earlier authorized frames that have since detached.
+The capture also
 inventories worker, service-worker, and shared-worker targets with
 validated opener-target and parent-frame IDs. Worker discovery uses passive
 target metadata; REA does not attach to or execute code in those targets.
-Collection counts and aggregate script-source bytes are not capped. Like every
+Script-source collection retains at most 32 MiB of UTF-8 text across scripts
+and shares a 20-second collection deadline. Omitted sources retain their script
+metadata and an explicit reason; later smaller sources may still fit after an
+oversized source is omitted. The separate 64 MiB CDP response limit bounds one
+wire message before retention admission, rather than total transient memory.
+The source budget adds no script-count cap. Like every
 target, frame, script, and resource, a worker URL must resolve to a local
 file before it is retained. Relationship IDs improve
 attribution but do not prove which static module started a worker or that its
@@ -92,7 +119,9 @@ The CLI and MCP surfaces accept the same schema. For a JSON request file:
 rea capture-electron-scenario scenario.json --json
 ```
 
-The result records action status and targets, correlated app/window/WebContents,
+The result records one status and target per declared action; actions after a
+failed or cancelled action are not run and are recorded as `cancelled` with a
+not-executed error. It also records correlated app/window/WebContents,
 preload, session, navigation, shell, permission, popup, download, protocol,
 native-addon, process, and IPC timeline events. IPC channels and argument-shape
 metadata retain their observed lengths; payload values are never retained.
@@ -101,8 +130,18 @@ scenario controls; their attempted and observed outcomes remain in the
 timeline. The active hook blocks and records external shell/navigation,
 permission, download, popup, updater, and OS-integration effects. The timeline
 is explicitly partial when attachment starts after application activity. The
-startup has a 60-second deadline. Actions wait until completion or request
-cancellation; a click whose target never appears remains pending until cancelled.
+startup, including the initial state snapshot, has a 60-second deadline.
+Actions wait until completion or request cancellation; a click whose target
+never appears, or a `wait` without `window_index` in an app that opens no
+window, remains pending until cancelled. After cancellation, the final state
+read has 5 seconds before REA keeps the initial snapshot instead; the
+cancellation keeps the caller's reason and records that expiry in its
+diagnostics.
+Blocked synchronous shell APIs keep their return contracts:
+`showItemInFolder` returns `undefined` and `writeShortcutLink` returns `false`.
+Their `shell-attempt` events record the blocked phase and argument shapes;
+neither result means the requested host operation occurred. The existing
+`openExternal` and `openPath` blocks continue to reject their promises.
 CLI SIGINT/SIGTERM requests cancellation and awaits owned-process cleanup.
 Window IDs, URLs, and cached titles come from one main-process Electron snapshot,
 so a crashed renderer does not require a successful renderer-side title query.
@@ -116,6 +155,80 @@ If cleanup cannot be verified, the error retains collected observations in
 `details.partial_observation` and marks their application cleanup `unverified`.
 Cleanup uncertainty does not authorize signaling unrelated processes.
 
+Main-process IPC recording keeps original-function removal for `on` and `once`
+registrations and records a fired `once` callback once. Native Node.js owns
+`once` registration, firing and callable dispatch as well as removal order and
+traversal, including listeners registered by removal observers;
+the hook translates only its own recording wrappers at the removal boundary.
+Wrapper metadata uses weak keys and does not retain removed registrations.
+Wrapper metadata operations, private collection methods and traversal constructors are captured
+before target startup; later replacement of their global bindings does not
+change listener registration or removal.
+
+Native runtimes differ in whether `once` consults callback `.call` or `.apply`.
+The hook observes actual callback entry without imposing another runtime's
+property reads or argument-list representation. Application methods selected by
+native dispatch keep their original receiver and argument-list object; opaque
+methods produce a `custom-call` or `custom-apply` observation gap when callback
+execution cannot be established.
+
+If native wrapper creation or registration-method lookup could expose private
+instrumentation through an application getter, setter or Proxy, registration
+stays native and unwrapped. The `once-metadata` observation gap identifies that
+unobserved registration attempt; it does not prove registration or delivery. The original callback,
+metadata writes and native method receivers are preserved, including when
+registration fails. Later registrations resume normal observation. Non-callable
+`call`/`apply` properties produce failed observation gaps when native dispatch reads
+them; callback entry is not fabricated.
+
+An `on` listener's application-defined `.apply` receives the original argument-list
+object unchanged. Borrowing the exposed method preserves its dynamic receiver.
+Because that method owns argument interpretation and may not
+invoke the callback, the hook records a bounded `ipc-observation-gap` timeline
+entry with the registered channel rather than an IPC invocation. Its empty
+`argument_shapes` means unobserved arguments, not a zero-argument callback;
+callback execution and result remain unknown. Coverage and retention include
+these gaps even when their individual records are dropped. Normal native
+dispatch continues to record actual invocation frames. Evidence snapshots copy
+retained arrays and counters without invoking application iterators or Array
+species.
+
+Listener-chain recognition distinguishes owned, unowned and unknown metadata.
+Data links may be inherited; absence is established only after completing the
+bounded prototype lookup. Exhausting either walk, encountering a getter or seeing a Proxy leaves
+ownership unknown. The hook preserves native registration without adding another
+recording wrapper and retains a `listener-chain` observation gap after successful
+registration. This gap does not prove IPC delivery. An existing inner recording
+wrapper can still report its actual invocation; otherwise callback activity
+remains unobserved. Setter-only properties have a known undefined read.
+Data-descriptor parsing is shared with removal normalization.
+
+This does not establish complete JavaScript function introspection equivalence.
+Caller-defined `.listener` properties can change the identity exposed by
+`listeners()` and meta-events; accessors and Proxy metadata are not followed for
+wrapper recognition. Re-registering a raw recording wrapper on another channel
+retains the original registration's channel attribution.
+Mutating `.listener` on a raw recording wrapper can also prevent removal through
+application decorators that bypass the hook's removal methods. When a removal
+request reaches the hook's methods and its registration order is known, private
+callback ownership preserves removal by the original function and the newest
+matching registration when public metadata is changed or deleted. Native removal
+still determines getter execution, exceptions and removal meta-events.
+
+If registration or removal methods differ from the Node methods captured at hook
+initialization, the hook leaves listener callbacks unwrapped and records a
+`listener-registration` observation gap. The application keeps its original
+callback and receiver; callback execution remains unobserved. This comparison
+assumes Node's EventEmitter prototype has not itself been globally decorated
+before hook initialization. The provider loads the hook before the application
+and excludes `NODE_OPTIONS` from the child environment.
+
+An opaque later listener or unsupported emitter state can prevent proof of
+removal order. The hook then delegates removal
+to the native method and records a `listener-removal` gap rather than claiming
+which registration was removed. These gaps are distinct from dropped records
+and do not establish IPC delivery or successful removal.
+
 ### Target fuse prerequisites
 
 Active scenario capture requires that the target Electron executable honors Node CLI inspector flags (`--inspect` or `--remote-debugging-port`) so that Playwright can establish its DevTools protocol session.
@@ -135,4 +248,10 @@ application path and capture outcome to the static graph, while frames, scripts,
 workers, and execution claims remain unavailable until a separate passive runtime
 capture provides them.
 
-ASAR observation reads the header through one open file handle, validates its declared lengths against the archive size, and limits encoded JSON headers to 16 MiB before parsing. The header cache also retains at most 16 MiB of encoded headers across eight archives. Oversized headers produce a scope refusal.
+ASAR observation reads the header through one open file handle and validates its
+declared lengths against the archive size. Header/index admission is the same
+budget used by static ASAR reads: 16 MiB / 16777216 bytes by default, or
+`REA_ASAR_MAX_HEADER_BYTES` (at most 256 MiB) when set. An oversized header is
+not parsed; selecting a member path does not bypass admission, and the page
+stays unauthorized. The header cache separately retains at most that many bytes
+of already-admitted encoded headers across eight archives. Evicting that cache does not raise admission.

docs/elf-notes.md

@@ -0,0 +1,70 @@
+# ELF note tables
+
+`inspect_elf_notes` / `inspect-elf-notes` answers whether one selected local
+ELF32/ELF64 image declares `PT_NOTE` program headers or `SHT_NOTE` sections,
+which note owners and types those tables store, and which GNU build-id or
+ABI-tag fields those notes carry. It reads the file without executing it and
+does not require a disassembler, debug package, or package catalog.
+
+```bash
+rea inspect-elf-notes /absolute/path/to/program --json
+```
+
+The CLI resolves relative paths against the caller's working directory. MCP
+requires an absolute filesystem path and preserves that exact selection:
+
+```json
+{
+  "name": "inspect_elf_notes",
+  "arguments": { "path": "/artifacts/program" }
+}
+```
+
+## Observations
+
+The inline Evidence result contains the artifact path, SHA-256, and byte length.
+`tables` lists each declared note program header or section, including the
+section name when a string table is present. Zlib-compressed section-name
+tables are decompressed within the selected file-size budget. An unsupported
+compression type leaves section names unknown and adds a limitation; note bytes
+remain independently examined. Each `entries[]` item reports owner, type,
+descriptor SHA-256, and a hex preview for descriptors of 64 bytes or less.
+
+For GNU notes, `gnu.build_id_hex`, `gnu.abi_tag`, or `gnu.gold_version` are
+filled when the descriptor matches that type. Other owners keep `gnu` null.
+
+`PT_NOTE` and `SHT_NOTE` are examined independently. The same file range may
+appear twice when both tables declare it.
+
+A missing table list is observed absence of those declarations. It is not proof
+the image never carried notes in another container or overlay.
+
+## Limits and failures
+
+The selected file is limited to 64 MiB by default (512 MiB maximum). At most
+32 note entries are examined by default across every table. Retained note
+names, section names, and GNU descriptor text share a scan budget based on
+the source bytes plus any decoded section-name table, capped at
+`max_file_bytes`. The raw source and decoded table must also fit together
+within that ceiling. If admission fails, the error reports the required
+`max_file_bytes`; raising it can admit the decoded table up to the operation's
+512 MiB maximum. Once admitted, raising that setting does not increase the
+scan budget for the same source and decoded names. Repeated table ranges count again. Each descriptor range is hashed once,
+and each GNU version's trailing NUL run is scanned once; neither walk may
+exceed that byte ceiling. Malformed ELF headers, truncated header
+tables, declared note tables that extend past the file, and truncated
+note entries fail without partial success. ELF extended
+numbering (`PN_XNUM`, including a section count stored when `e_shnum` is 0) is
+unsupported and returns an unsupported-target outcome rather than invalid-input
+recovery.
+
+`SHF_COMPRESSED` note sections return `unsupported_target` before their payload
+is interpreted or decompressed. Their compression header is storage metadata,
+not a note header. The current result requires physical file offsets for each
+note and descriptor, which decoded compressed notes cannot provide. This refusal
+does not label compressed note storage as malformed or allocate its declared
+decoded size. Compressed section-name tables remain supported within the limits
+above.
+
+This workflow does not execute the image or verify GNU build-id uniqueness
+against package catalogs.

docs/erlang-beam.md

@@ -0,0 +1,67 @@
+# Erlang BEAM files
+
+`inspect_erlang_beam` / `inspect-erlang-beam` answers whether one selected
+local file is a `FOR1`/`BEAM` IFF container, which chunks that form stores, and
+which atom, export, import, lambda, and Code-header fields those chunks carry.
+It reads the file without executing bytecode and does not require an Erlang
+installation or BEAM emulator.
+
+```bash
+rea inspect-erlang-beam /absolute/path/to/app.beam --json
+```
+
+The CLI resolves relative paths against the caller's working directory. MCP
+requires an absolute filesystem path and preserves that exact selection:
+
+```json
+{
+  "name": "inspect_erlang_beam",
+  "arguments": { "path": "/artifacts/app.beam" }
+}
+```
+
+## Observations
+
+The inline Evidence result contains the artifact path, SHA-256, and byte length.
+`chunks` lists every IFF chunk name, file range, and payload SHA-256.
+Repeated names retain separate chunk occurrences. Decoded tables and the Code
+header use the last occurrence of each name, matching OTP's loader; the chunk
+inventory still exposes earlier payloads and their source ranges.
+The BEAM form size bounds the parsed chunks. Bytes after the declared form are
+ignored by the parser but remain part of the selected artifact's byte length
+and SHA-256.
+
+When a selected `AtU8` or `Atom` table is decoded, `atoms` is the 1-based name table and
+`module` is atom 1. Both the pre-OTP 28 table and the OTP 28+ compact table
+(negated count, compact-encoded name lengths) are read. `exports`, `imports`, and `lambdas` resolve atom indices to
+those names when the table is present; unresolved indices keep `function` or
+`module` null.
+
+`code` reports the Code chunk instruction-set, opcode maximum, label count,
+function count, and the instruction stream's file range. Missing optional
+tables are null, not empty-everywhere.
+
+A zero-byte `AtU8` falls back to an available legacy `Atom` table, following
+[OTP 20.3's atom selection](https://github.com/erlang/otp/blob/a113f6117fd696ea6f84ed754055b4ec97a7ccb2/erts/emulator/beam/beam_load.c).
+A nonempty `AtU8` retains UTF-8 precedence, including an explicit zero atom
+count; malformed nonempty data fails instead of falling back. An empty atom
+chunk without an alternate table still fails its required count header.
+Absent and zero-byte optional `FunT` chunks produce `lambdas: null`, while a
+four-byte zero row count produces `lambdas: []`. This follows
+[OTP 29's positive-size lambda parsing rule](https://github.com/erlang/otp/blob/550d7b7898706c7822362c42e7b93120c1d1f29a/erts/emulator/beam/beam_file.c).
+Empty payloads retain their ranges and SHA-256 in the chunk inventory. These
+rules apply to the last occurrence of each chunk name; an empty final occurrence
+does not revive an earlier payload of that name. Nonempty truncated tables remain
+format failures.
+
+## Limits and failures
+
+The selected file is limited to 64 MiB by default (512 MiB maximum). At most
+64 chunks and 4096 atoms are examined by default, and at most 65536 export,
+import, and lambda rows together. Malformed `FOR1`/`BEAM` headers, truncated
+chunks, and `AtU8` names that are not valid UTF-8 fail without partial success.
+
+This workflow does not execute bytecode, inflate `LitT` literals, or decode
+`Attr`/`CInf` external-term payloads. Complete coverage describes this static
+chunk and table report; it does not establish that an Erlang runtime can load
+or execute the module.

docs/error-contract.schema.json

@@ -28,6 +28,33 @@
           "required": ["action"],
           "additionalProperties": false
         },
+        "recovery": {
+          "oneOf": [
+            {
+              "type": "object",
+              "properties": {
+                "kind": { "type": "string", "const": "inspect_input_contract" },
+                "operation": { "type": "string", "minLength": 1 }
+              },
+              "required": ["kind", "operation"],
+              "additionalProperties": false
+            },
+            {
+              "type": "object",
+              "properties": {
+                "kind": { "type": "string", "const": "inspect_timeout_state" },
+                "operation": { "type": "string", "minLength": 1 },
+                "timeout_ms": {
+                  "type": "integer",
+                  "exclusiveMinimum": 0,
+                  "maximum": 9007199254740991
+                }
+              },
+              "required": ["kind", "operation", "timeout_ms"],
+              "additionalProperties": false
+            }
+          ]
+        },
         "details": {
           "type": "object",
           "propertyNames": { "type": "string" },
@@ -64,6 +91,33 @@
           "required": ["action"],
           "additionalProperties": false
         },
+        "recovery": {
+          "oneOf": [
+            {
+              "type": "object",
+              "properties": {
+                "kind": { "type": "string", "const": "inspect_input_contract" },
+                "operation": { "type": "string", "minLength": 1 }
+              },
+              "required": ["kind", "operation"],
+              "additionalProperties": false
+            },
+            {
+              "type": "object",
+              "properties": {
+                "kind": { "type": "string", "const": "inspect_timeout_state" },
+                "operation": { "type": "string", "minLength": 1 },
+                "timeout_ms": {
+                  "type": "integer",
+                  "exclusiveMinimum": 0,
+                  "maximum": 9007199254740991
+                }
+              },
+              "required": ["kind", "operation", "timeout_ms"],
+              "additionalProperties": false
+            }
+          ]
+        },
         "details": {
           "type": "object",
           "propertyNames": { "type": "string" },
@@ -100,6 +154,33 @@
           "required": ["action"],
           "additionalProperties": false
         },
+        "recovery": {
+          "oneOf": [
+            {
+              "type": "object",
+              "properties": {
+                "kind": { "type": "string", "const": "inspect_input_contract" },
+                "operation": { "type": "string", "minLength": 1 }
+              },
+              "required": ["kind", "operation"],
+              "additionalProperties": false
+            },
+            {
+              "type": "object",
+              "properties": {
+                "kind": { "type": "string", "const": "inspect_timeout_state" },
+                "operation": { "type": "string", "minLength": 1 },
+                "timeout_ms": {
+                  "type": "integer",
+                  "exclusiveMinimum": 0,
+                  "maximum": 9007199254740991
+                }
+              },
+              "required": ["kind", "operation", "timeout_ms"],
+              "additionalProperties": false
+            }
+          ]
+        },
         "details": {
           "type": "object",
           "propertyNames": { "type": "string" },
@@ -136,6 +217,33 @@
           "required": ["action"],
           "additionalProperties": false
         },
+        "recovery": {
+          "oneOf": [
+            {
+              "type": "object",
+              "properties": {
+                "kind": { "type": "string", "const": "inspect_input_contract" },
+                "operation": { "type": "string", "minLength": 1 }
+              },
+              "required": ["kind", "operation"],
+              "additionalProperties": false
+            },
+            {
+              "type": "object",
+              "properties": {
+                "kind": { "type": "string", "const": "inspect_timeout_state" },
+                "operation": { "type": "string", "minLength": 1 },
+                "timeout_ms": {
+                  "type": "integer",
+                  "exclusiveMinimum": 0,
+                  "maximum": 9007199254740991
+                }
+              },
+              "required": ["kind", "operation", "timeout_ms"],
+              "additionalProperties": false
+            }
+          ]
+        },
         "details": {
           "type": "object",
           "propertyNames": { "type": "string" },
@@ -172,6 +280,33 @@
           "required": ["action"],
           "additionalProperties": false
         },
+        "recovery": {
+          "oneOf": [
+            {
+              "type": "object",
+              "properties": {
+                "kind": { "type": "string", "const": "inspect_input_contract" },
+                "operation": { "type": "string", "minLength": 1 }
+              },
+              "required": ["kind", "operation"],
+              "additionalProperties": false
+            },
+            {
+              "type": "object",
+              "properties": {
+                "kind": { "type": "string", "const": "inspect_timeout_state" },
+                "operation": { "type": "string", "minLength": 1 },
+                "timeout_ms": {
+                  "type": "integer",
+                  "exclusiveMinimum": 0,
+                  "maximum": 9007199254740991
+                }
+              },
+              "required": ["kind", "operation", "timeout_ms"],
+              "additionalProperties": false
+            }
+          ]
+        },
         "details": {
           "type": "object",
           "propertyNames": { "type": "string" },
@@ -208,6 +343,33 @@
           "required": ["action"],
           "additionalProperties": false
         },
+        "recovery": {
+          "oneOf": [
+            {
+              "type": "object",
+              "properties": {
+                "kind": { "type": "string", "const": "inspect_input_contract" },
+                "operation": { "type": "string", "minLength": 1 }
+              },
+              "required": ["kind", "operation"],
+              "additionalProperties": false
+            },
+            {
+              "type": "object",
+              "properties": {
+                "kind": { "type": "string", "const": "inspect_timeout_state" },
+                "operation": { "type": "string", "minLength": 1 },
+                "timeout_ms": {
+                  "type": "integer",
+                  "exclusiveMinimum": 0,
+                  "maximum": 9007199254740991
+                }
+              },
+              "required": ["kind", "operation", "timeout_ms"],
+              "additionalProperties": false
+            }
+          ]
+        },
         "details": {
           "type": "object",
           "propertyNames": { "type": "string" },
@@ -244,6 +406,33 @@
           "required": ["action"],
           "additionalProperties": false
         },
+        "recovery": {
+          "oneOf": [
+            {
+              "type": "object",
+              "properties": {
+                "kind": { "type": "string", "const": "inspect_input_contract" },
+                "operation": { "type": "string", "minLength": 1 }
+              },
+              "required": ["kind", "operation"],
+              "additionalProperties": false
+            },
+            {
+              "type": "object",
+              "properties": {
+                "kind": { "type": "string", "const": "inspect_timeout_state" },
+                "operation": { "type": "string", "minLength": 1 },
+                "timeout_ms": {
+                  "type": "integer",
+                  "exclusiveMinimum": 0,
+                  "maximum": 9007199254740991
+                }
+              },
+              "required": ["kind", "operation", "timeout_ms"],
+              "additionalProperties": false
+            }
+          ]
+        },
         "details": {
           "type": "object",
           "propertyNames": { "type": "string" },
@@ -280,6 +469,33 @@
           "required": ["action"],
           "additionalProperties": false
         },
+        "recovery": {
+          "oneOf": [
+            {
+              "type": "object",
+              "properties": {
+                "kind": { "type": "string", "const": "inspect_input_contract" },
+                "operation": { "type": "string", "minLength": 1 }
+              },
+              "required": ["kind", "operation"],
+              "additionalProperties": false
+            },
+            {
+              "type": "object",
+              "properties": {
+                "kind": { "type": "string", "const": "inspect_timeout_state" },
+                "operation": { "type": "string", "minLength": 1 },
+                "timeout_ms": {
+                  "type": "integer",
+                  "exclusiveMinimum": 0,
+                  "maximum": 9007199254740991
+                }
+              },
+              "required": ["kind", "operation", "timeout_ms"],
+              "additionalProperties": false
+            }
+          ]
+        },
         "details": {
           "type": "object",
           "propertyNames": { "type": "string" },
@@ -316,6 +532,33 @@
           "required": ["action"],
           "additionalProperties": false
         },
+        "recovery": {
+          "oneOf": [
+            {
+              "type": "object",
+              "properties": {
+                "kind": { "type": "string", "const": "inspect_input_contract" },
+                "operation": { "type": "string", "minLength": 1 }
+              },
+              "required": ["kind", "operation"],
+              "additionalProperties": false
+            },
+            {
+              "type": "object",
+              "properties": {
+                "kind": { "type": "string", "const": "inspect_timeout_state" },
+                "operation": { "type": "string", "minLength": 1 },
+                "timeout_ms": {
+                  "type": "integer",
+                  "exclusiveMinimum": 0,
+                  "maximum": 9007199254740991
+                }
+              },
+              "required": ["kind", "operation", "timeout_ms"],
+              "additionalProperties": false
+            }
+          ]
+        },
         "details": {
           "type": "object",
           "propertyNames": { "type": "string" },
@@ -352,6 +595,33 @@
           "required": ["action"],
           "additionalProperties": false
         },
+        "recovery": {
+          "oneOf": [
+            {
+              "type": "object",
+              "properties": {
+                "kind": { "type": "string", "const": "inspect_input_contract" },
+                "operation": { "type": "string", "minLength": 1 }
+              },
+              "required": ["kind", "operation"],
+              "additionalProperties": false
+            },
+            {
+              "type": "object",
+              "properties": {
+                "kind": { "type": "string", "const": "inspect_timeout_state" },
+                "operation": { "type": "string", "minLength": 1 },
+                "timeout_ms": {
+                  "type": "integer",
+                  "exclusiveMinimum": 0,
+                  "maximum": 9007199254740991
+                }
+              },
+              "required": ["kind", "operation", "timeout_ms"],
+              "additionalProperties": false
+            }
+          ]
+        },
         "details": {
           "type": "object",
           "propertyNames": { "type": "string" },
@@ -388,25 +658,82 @@
           "required": ["action"],
           "additionalProperties": false
         },
+        "recovery": {
+          "oneOf": [
+            {
+              "type": "object",
+              "properties": {
+                "kind": { "type": "string", "const": "inspect_input_contract" },
+                "operation": { "type": "string", "minLength": 1 }
+              },
+              "required": ["kind", "operation"],
+              "additionalProperties": false
+            },
+            {
+              "type": "object",
+              "properties": {
+                "kind": { "type": "string", "const": "inspect_timeout_state" },
+                "operation": { "type": "string", "minLength": 1 },
+                "timeout_ms": {
+                  "type": "integer",
+                  "exclusiveMinimum": 0,
+                  "maximum": 9007199254740991
+                }
+              },
+              "required": ["kind", "operation", "timeout_ms"],
+              "additionalProperties": false
+            }
+          ]
+        },
         "details": {
-          "type": "object",
-          "properties": {
-            "logical_path": { "type": "string" },
-            "declared_sha256": {
-              "anyOf": [{ "type": "string" }, { "type": "null" }]
+          "anyOf": [
+            {
+              "type": "object",
+              "properties": {
+                "logical_path": { "type": "string" },
+                "declared_sha256": {
+                  "anyOf": [{ "type": "string" }, { "type": "null" }]
+                },
+                "calculated_sha256": {
+                  "anyOf": [{ "type": "string" }, { "type": "null" }]
+                },
+                "unpacked": { "type": "boolean" }
+              },
+              "required": [
+                "logical_path",
+                "declared_sha256",
+                "calculated_sha256",
+                "unpacked"
+              ],
+              "additionalProperties": false
             },
-            "calculated_sha256": {
-              "anyOf": [{ "type": "string" }, { "type": "null" }]
-            },
-            "unpacked": { "type": "boolean" }
-          },
-          "required": [
-            "logical_path",
-            "declared_sha256",
-            "calculated_sha256",
-            "unpacked"
-          ],
-          "additionalProperties": false
+            {
+              "type": "object",
+              "properties": {
+                "logical_path": { "type": "string" },
+                "algorithm": { "type": "string", "const": "md5" },
+                "declared_md5": {
+                  "type": "string",
+                  "pattern": "^[a-f0-9]{32}$"
+                },
+                "calculated_md5": {
+                  "anyOf": [
+                    { "type": "string", "pattern": "^[a-f0-9]{32}$" },
+                    { "type": "null" }
+                  ]
+                },
+                "unpacked": { "type": "boolean" }
+              },
+              "required": [
+                "logical_path",
+                "algorithm",
+                "declared_md5",
+                "calculated_md5",
+                "unpacked"
+              ],
+              "additionalProperties": false
+            }
+          ]
         }
       },
       "required": [
@@ -446,6 +773,33 @@
           "required": ["action"],
           "additionalProperties": false
         },
+        "recovery": {
+          "oneOf": [
+            {
+              "type": "object",
+              "properties": {
+                "kind": { "type": "string", "const": "inspect_input_contract" },
+                "operation": { "type": "string", "minLength": 1 }
+              },
+              "required": ["kind", "operation"],
+              "additionalProperties": false
+            },
+            {
+              "type": "object",
+              "properties": {
+                "kind": { "type": "string", "const": "inspect_timeout_state" },
+                "operation": { "type": "string", "minLength": 1 },
+                "timeout_ms": {
+                  "type": "integer",
+                  "exclusiveMinimum": 0,
+                  "maximum": 9007199254740991
+                }
+              },
+              "required": ["kind", "operation", "timeout_ms"],
+              "additionalProperties": false
+            }
+          ]
+        },
         "details": {
           "type": "object",
           "propertyNames": { "type": "string" },
@@ -482,6 +836,33 @@
           "required": ["action"],
           "additionalProperties": false
         },
+        "recovery": {
+          "oneOf": [
+            {
+              "type": "object",
+              "properties": {
+                "kind": { "type": "string", "const": "inspect_input_contract" },
+                "operation": { "type": "string", "minLength": 1 }
+              },
+              "required": ["kind", "operation"],
+              "additionalProperties": false
+            },
+            {
+              "type": "object",
+              "properties": {
+                "kind": { "type": "string", "const": "inspect_timeout_state" },
+                "operation": { "type": "string", "minLength": 1 },
+                "timeout_ms": {
+                  "type": "integer",
+                  "exclusiveMinimum": 0,
+                  "maximum": 9007199254740991
+                }
+              },
+              "required": ["kind", "operation", "timeout_ms"],
+              "additionalProperties": false
+            }
+          ]
+        },
         "details": {
           "type": "object",
           "propertyNames": { "type": "string" },
@@ -518,6 +899,33 @@
           "required": ["action"],
           "additionalProperties": false
         },
+        "recovery": {
+          "oneOf": [
+            {
+              "type": "object",
+              "properties": {
+                "kind": { "type": "string", "const": "inspect_input_contract" },
+                "operation": { "type": "string", "minLength": 1 }
+              },
+              "required": ["kind", "operation"],
+              "additionalProperties": false
+            },
+            {
+              "type": "object",
+              "properties": {
+                "kind": { "type": "string", "const": "inspect_timeout_state" },
+                "operation": { "type": "string", "minLength": 1 },
+                "timeout_ms": {
+                  "type": "integer",
+                  "exclusiveMinimum": 0,
+                  "maximum": 9007199254740991
+                }
+              },
+              "required": ["kind", "operation", "timeout_ms"],
+              "additionalProperties": false
+            }
+          ]
+        },
         "details": {
           "type": "object",
           "propertyNames": { "type": "string" },
@@ -554,6 +962,33 @@
           "required": ["action"],
           "additionalProperties": false
         },
+        "recovery": {
+          "oneOf": [
+            {
+              "type": "object",
+              "properties": {
+                "kind": { "type": "string", "const": "inspect_input_contract" },
+                "operation": { "type": "string", "minLength": 1 }
+              },
+              "required": ["kind", "operation"],
+              "additionalProperties": false
+            },
+            {
+              "type": "object",
+              "properties": {
+                "kind": { "type": "string", "const": "inspect_timeout_state" },
+                "operation": { "type": "string", "minLength": 1 },
+                "timeout_ms": {
+                  "type": "integer",
+                  "exclusiveMinimum": 0,
+                  "maximum": 9007199254740991
+                }
+              },
+              "required": ["kind", "operation", "timeout_ms"],
+              "additionalProperties": false
+            }
+          ]
+        },
         "details": {
           "type": "object",
           "propertyNames": { "type": "string" },
@@ -590,6 +1025,33 @@
           "required": ["action"],
           "additionalProperties": false
         },
+        "recovery": {
+          "oneOf": [
+            {
+              "type": "object",
+              "properties": {
+                "kind": { "type": "string", "const": "inspect_input_contract" },
+                "operation": { "type": "string", "minLength": 1 }
+              },
+              "required": ["kind", "operation"],
+              "additionalProperties": false
+            },
+            {
+              "type": "object",
+              "properties": {
+                "kind": { "type": "string", "const": "inspect_timeout_state" },
+                "operation": { "type": "string", "minLength": 1 },
+                "timeout_ms": {
+                  "type": "integer",
+                  "exclusiveMinimum": 0,
+                  "maximum": 9007199254740991
+                }
+              },
+              "required": ["kind", "operation", "timeout_ms"],
+              "additionalProperties": false
+            }
+          ]
+        },
         "details": {
           "type": "object",
           "propertyNames": { "type": "string" },
@@ -626,6 +1088,33 @@
           "required": ["action"],
           "additionalProperties": false
         },
+        "recovery": {
+          "oneOf": [
+            {
+              "type": "object",
+              "properties": {
+                "kind": { "type": "string", "const": "inspect_input_contract" },
+                "operation": { "type": "string", "minLength": 1 }
+              },
+              "required": ["kind", "operation"],
+              "additionalProperties": false
+            },
+            {
+              "type": "object",
+              "properties": {
+                "kind": { "type": "string", "const": "inspect_timeout_state" },
+                "operation": { "type": "string", "minLength": 1 },
+                "timeout_ms": {
+                  "type": "integer",
+                  "exclusiveMinimum": 0,
+                  "maximum": 9007199254740991
+                }
+              },
+              "required": ["kind", "operation", "timeout_ms"],
+              "additionalProperties": false
+            }
+          ]
+        },
         "details": {
           "type": "object",
           "propertyNames": { "type": "string" },
@@ -662,6 +1151,33 @@
           "required": ["action"],
           "additionalProperties": false
         },
+        "recovery": {
+          "oneOf": [
+            {
+              "type": "object",
+              "properties": {
+                "kind": { "type": "string", "const": "inspect_input_contract" },
+                "operation": { "type": "string", "minLength": 1 }
+              },
+              "required": ["kind", "operation"],
+              "additionalProperties": false
+            },
+            {
+              "type": "object",
+              "properties": {
+                "kind": { "type": "string", "const": "inspect_timeout_state" },
+                "operation": { "type": "string", "minLength": 1 },
+                "timeout_ms": {
+                  "type": "integer",
+                  "exclusiveMinimum": 0,
+                  "maximum": 9007199254740991
+                }
+              },
+              "required": ["kind", "operation", "timeout_ms"],
+              "additionalProperties": false
+            }
+          ]
+        },
         "details": {
           "type": "object",
           "propertyNames": { "type": "string" },
@@ -698,6 +1214,33 @@
           "required": ["action"],
           "additionalProperties": false
         },
+        "recovery": {
+          "oneOf": [
+            {
+              "type": "object",
+              "properties": {
+                "kind": { "type": "string", "const": "inspect_input_contract" },
+                "operation": { "type": "string", "minLength": 1 }
+              },
+              "required": ["kind", "operation"],
+              "additionalProperties": false
+            },
+            {
+              "type": "object",
+              "properties": {
+                "kind": { "type": "string", "const": "inspect_timeout_state" },
+                "operation": { "type": "string", "minLength": 1 },
+                "timeout_ms": {
+                  "type": "integer",
+                  "exclusiveMinimum": 0,
+                  "maximum": 9007199254740991
+                }
+              },
+              "required": ["kind", "operation", "timeout_ms"],
+              "additionalProperties": false
+            }
+          ]
+        },
         "details": {
           "type": "object",
           "propertyNames": { "type": "string" },
@@ -734,6 +1277,33 @@
           "required": ["action"],
           "additionalProperties": false
         },
+        "recovery": {
+          "oneOf": [
+            {
+              "type": "object",
+              "properties": {
+                "kind": { "type": "string", "const": "inspect_input_contract" },
+                "operation": { "type": "string", "minLength": 1 }
+              },
+              "required": ["kind", "operation"],
+              "additionalProperties": false
+            },
+            {
+              "type": "object",
+              "properties": {
+                "kind": { "type": "string", "const": "inspect_timeout_state" },
+                "operation": { "type": "string", "minLength": 1 },
+                "timeout_ms": {
+                  "type": "integer",
+                  "exclusiveMinimum": 0,
+                  "maximum": 9007199254740991
+                }
+              },
+              "required": ["kind", "operation", "timeout_ms"],
+              "additionalProperties": false
+            }
+          ]
+        },
         "details": {
           "type": "object",
           "propertyNames": { "type": "string" },

docs/evm-bytecode.md

@@ -19,8 +19,10 @@ rea inspect-evm-interface ./runtime.bin raw --json
 
 Initial real verification covers flat legacy bytecode on Linux x64 and unchanged
 bundled EVMole 0.9.3. EOF-style EF00 containers return `unsupported_target`,
-with the selected path and format reason. Provider/configuration failures retain
-their distinct unavailable-provider or capability classification.
+with the selected path and format reason. Selected-path resolution errors
+`ENOENT`, `ENOTDIR`, `ELOOP`, and `ENAMETOOLONG` are `invalid_request` path
+issues. Permission, resource, cancellation, provider, and cleanup failures retain
+their distinct classifications.
 The host must already provide util-linux `prlimit`, normally `/usr/bin/prlimit`;
 `REA_EVM_PRLIMIT_COMMAND` can select another absolute executable. REA installs
 no system tools and changes no user configuration. [Upstream provenance and
@@ -43,7 +45,8 @@ license](https://github.com/morluto/rea/blob/main/third_party/evmole/README.md)
   map from the selected bytes. Integers use decimal strings; byte strings use hex,
   and scalar keys retain their types. The recorded offset/length addresses the
   original bytecode, so summary delivery does not copy an opaque CBOR payload.
-  Nested containers have partial scalar coverage; unsupported or malformed
+  Nested containers have partial scalar coverage; map keys are still checked for
+  duplicates throughout nested maps and arrays. Unsupported or malformed
   candidates report `not-decoded` with a diagnostic. Decoding is bounded to 64
   container levels and the terminal payload's encoded size.
 - An unavailable trailer does not prove metadata absence or Solidity origin.
@@ -106,7 +109,13 @@ CLI keeps its existing complete result.
 
 Inputs are limited to 4 MiB, complete replies to 16 MiB and combined diagnostic
 output to 1 MiB. An owned worker has a 256 MiB JavaScript heap, 3 GiB virtual
-address-space limit, 30 CPU seconds and a 30-second command deadline. Virtual
+address-space limit and 30 CPU seconds. The worker command deadline is
+`timeout_ms` (default 30000 ms, integer 1 through 2147483647) on the next call,
+or `--timeout-ms` on `rea inspect-evm-interface`. Omission keeps 30000 ms.
+Those byte, address-space, and CPU ceilings have no runtime override, and
+`timeout_ms` does not raise them. An oversized selected carrier returns a
+non-retryable `resource_constraint` with resource `file-size` and its fixed
+input allowance in `reported_limits.maximum_input_bytes`. Virtual
 address space differs from resident memory. The worker retains inherited tighter
 soft/hard limits, reports its configured soft limits inline and leaves hard limits
 unchanged. WASM trap reservations are disabled

docs/firmware-analysis.md

@@ -50,6 +50,27 @@ git submodule update --init third_party/binwalk third_party/unblob
 Normal REA builds and npm packages do not include or build these sources or
 engine binaries. See [upstream provenance](https://github.com/morluto/rea/blob/main/third_party/README.md).
 
+### Windows submodule checkout
+
+On Windows, the Unblob checkout stops with `error: invalid path`. Several
+upstream test fixture names under `tests/integration/` contain characters,
+trailing dots, or trailing spaces that NTFS does not allow; REA does not use
+those fixtures. Keep the clone, exclude that directory, and retry from the
+repository root in PowerShell:
+
+```powershell
+git -C third_party/unblob config core.protectNTFS false
+git -C third_party/unblob sparse-checkout set --no-cone '/*' '!/tests/integration/'
+git submodule update third_party/unblob
+```
+
+Both settings are required. While `core.protectNTFS` is enabled, Git for
+Windows also rejects index paths that the sparse pattern excludes; the pattern
+keeps the invalid names out of the worktree. The setting applies only to the
+Unblob submodule. In Git Bash, prefix the `sparse-checkout` command with
+`MSYS_NO_PATHCONV=1`; otherwise `!/tests/integration/` is rewritten as a
+Windows path.
+
 ## CLI and MCP
 
 ```sh
@@ -153,13 +174,39 @@ tree semantics, vulnerability scanning, emulation or hardware runtime capture.
 ## Resource and lifecycle policy
 
 - Nonempty regular firmware inputs: at most **128 MiB**, copied and hashed into
-  a private workspace before analysis. No in-place source modification.
+  a private workspace before analysis. Growth beyond that allowance during
+  copying receives the same typed input refusal as an already oversized file.
+  Supply a stable copy within the allowance. No in-place source modification.
 - One queued heavy operation per provider instance; Binwalk one thread and
   Unblob one worker. No automatic parallel extraction engines.
-- **120-second** deadline; inherited **1 GiB per-process address-space** limit,
+- Analysis defaults to **120 seconds**, from tool resolution and input snapshot
+  through inspection and its final private-workspace cleanup, or extraction
+  through report consumption and inventory, excluding queue waiting and retained
+  cleanup admission. Inspection has no publication phase. Select `analysis_timeout_ms` in either MCP operation or
+  `--analysis-timeout-ms` in either CLI command for slower analysis. Current
+  results retain the actual budget; an absent historical value means unknown.
+  Extraction publication and its final
+  cleanup start a separate **120-second** budget; select
+  `publication_timeout_ms` in MCP or `--publication-timeout-ms` in CLI for a slow
+  destination. Valid budgets are integers from 1 through 2,147,483,647 ms (the
+  timer representation limit). Publication timeouts identify their phase and
+  verified file progress; verified writes before atomic commit are not a claim
+  that output remains published after rollback. Caller cancellation applies to
+  both phases and final cleanup. Phase handoff, publication commit and final
+  completion also check monotonic elapsed time, so delayed timer callbacks
+  cannot revive an expired budget.
+- Analysis timeout recovery names the selected budget and the analysis control,
+  not the later publication control. Keep the client call timeout above analysis,
+  publication and cleanup headroom; keep the input/range/question unchanged.
+  Resolve any reported incomplete cleanup before retrying. Raising wall time
+  does not raise fixed CPU, memory, report, diagnostic or input limits.
+- Inherited **1 GiB per-process address-space** limit,
   120 CPU seconds and zero core-dump allowance through `prlimit`.
 - JSON reports: at most **8 MiB** and valid UTF-8; malformed encoding and
   oversized reports fail without repairing or truncating producer data.
+  Exit status 0 is not an empty result: a missing, empty, malformed or
+  oversized report fails with `unreadable_output`, its `report_state`, the
+  command, exit status and captured output. REA does not retry automatically.
   Diagnostic output is bounded with exact captured byte counts.
 - Default staging budget: **256 MiB** and **10,000 entries**, configurable using
   `max_output_bytes` / `max_output_files` (CLI `--max-output-bytes` /

docs/flutter-analysis.md

@@ -23,7 +23,9 @@ rea inspect-dart-aot /targets/Example.apk --abi armeabi-v7a
 The equivalent MCP methods are `identify_flutter_build` (with `path`) and
 `inspect_dart_aot` (with `path` and an optional `abi`; the default prefers
 arm64-v8a, then x86_64, then x86, then armeabi-v7a, then the remaining
-ABIs alphabetically).
+ABIs alphabetically). An explicit `abi` is the directory label exactly as
+`identify_flutter_build` reports it, so a payload under a nonstandard label
+stays selectable and keeps that label in the result.
 
 ## Optional retained AOT delivery
 
@@ -106,7 +108,12 @@ server; older servers keep the complete request after schema discovery.
   `build_id_candidates` preserves distinct IDs. `build_id` is populated only
   when all declared note ranges were examined and one ID was found;
   conflicting IDs or malformed/bounded note metadata make coverage partial.
+  Nonempty GNU build-ID descriptors up to 64 bytes are retained as lowercase
+  hexadecimal, including IDs shorter than four bytes.
   Matching bytes outside declared note ranges cannot supply a build ID.
+  Compressed `SHT_NOTE` sections are skipped with an unsupported-compression
+  limitation and partial coverage; IDs from other note ranges remain candidates.
+  Encoded note storage is never interpreted as note headers or inflated.
 - Labeled strings must fit their complete grammar. `dart_version_capped`
   and `toolchain_capped` independently report labels omitted by grammar or
   projection bounds; either makes identification coverage partial. Values
@@ -150,9 +157,16 @@ server; older servers keep the complete request after schema discovery.
 - **Selected ABI.** AOT inspection applies the library budget to the selected
   ABI after selection; an oversized unselected library does not block it.
   Undefined dynamic-symbol imports are not treated as snapshot exports.
+  A retained section directory without `SHT_DYNSYM` uses the same bounded
+  `PT_DYNAMIC` symbol description as a section-stripped image. Malformed
+  present symbol-table sections remain failures.
   Section-stripped images require a bounded symbol count from `DT_HASH` or
   `DT_GNU_HASH`; layouts without either are refused. Symbol names and their
-  terminators must fit their linked string table.
+  terminators must fit their linked string table. Only dynamic entries used for
+  symbol discovery are interpreted as numeric bounds. ABI-ignored payloads,
+  including [`DT_BIND_NOW`](https://gabi.xinuos.com/elf/08-dynamic.html#dynamic-section),
+  do not change discovery; their stored bytes still contribute to the selected
+  artifact's digest.
 - REA performs **no hash-to-SDK lookup**: mapping the snapshot hash to a
   Dart SDK release requires an external dataset and is deliberately out of
   scope for this family.

docs/frida-instrumentation.md

@@ -6,8 +6,26 @@ For remote devices, pass a host with an optional port (`host[:port]`; bracket IP
 
 MCP exposes `list_frida_devices`, `list_frida_processes`, `start_frida_session`, `load_frida_script`, `resume_frida_session`, `unload_frida_script`, `frida_session_status`, and `close_frida_session`. A spawned process starts paused so scripts can load first. Closing unloads session scripts and detaches. If a spawn remains paused, REA resumes it only after script cleanup or session detach confirms the instrumentation is released; if both cleanup paths fail, the target stays paused and the cleanup error is retained. REA never kills the target. MCP sessions end with the server process. The one-shot CLI command `instrument-with-frida` accepts one inline script or script file, observes messages for a bounded duration, and returns Evidence with the source hash and captured messages.
 
+`instrument-with-frida --json` returns the Evidence record directly, matching MCP. Consumers of the earlier `{evidence, cleanup_error}` CLI wrapper should read the record at the JSON root and the session cleanup observation at `normalized_result.cleanup_error`. If final provider cleanup fails or cancellation arrives during it, the typed error retains completed Evidence in `details.partial_observation`. Discovery commands retain their device or process report in the same field.
+
+For state and capture-coverage checks, pass `{"session_id":"SESSION_ID","detail":"summary"}` to `frida_session_status`. Each call reads the observed session state and retains the complete observation. The reply is an `inspect_analysis_view` Evidence summary with the session identity, state, scripts, `messages_retained`, and `messages_truncated`; message payloads remain in its `parent_evidence_id`. The count covers retained queued messages, not every target event. Omitted `detail` or `detail=complete` preserves the existing complete reply.
+
+A summary's selected fields look like this (excerpt):
+
+```json
+{
+  "summary": {
+    "state": "running",
+    "messages_retained": 50,
+    "messages_truncated": false
+  }
+}
+```
+
+Use `get_evidence_bundle` with `detail="summary"` and `filters={"evidence_id":"PARENT_ID"}` to discover parent metadata without message payloads. Filters apply only to summary delivery. Use `export_evidence_bundle` for the complete canonical bundle, then select the record whose `evidence_id` matches `normalized_result.parent_evidence_id` from the status summary. To use the existing CLI `inspect-analysis-view`, extract that record into a standalone Evidence JSON file and select it as an Evidence-file source with `view={"kind":"summary"}`. Reading retained Evidence does not refresh a live session or verify a process after exit or PID reuse.
+
 Failed startup cleanup retains acquired Frida handles, including a paused spawn that has not attached, for later cleanup retries. Its remote connection stays leased until session detach and target resume complete. CLI commands await provider cleanup before returning; MCP retains these resources for cleanup when its server closes. Server shutdown stops admitting new Frida operations and waits for admitted startup, discovery, and queued work before releasing their resources. A failed cleanup remains reported and can be retried by closing the server again, rather than treating an unregistered session as fully released.
 
-Frida scripts execute inside the selected target with that process's memory, filesystem, and network authority. Message capture is bounded to 16 MiB and reports truncation. Script source is not stored in Evidence; its selected path and SHA-256 digest are. Evidence retains non-authentication target selection such as device or endpoint, mode, program or PID, and spawn arguments. Authentication values remain excluded. Results report the observed session and messages, while unsupported platform metadata remains unknown. A request cancelled before a Frida operation starts has no provider effects. If cancellation arrives during a native Frida call, REA waits for that call to return, then closes any session it created and returns the observed session as a partial observation. A cancelled script load on an existing session releases only its new script before queued operations can run; the session and previously loaded scripts remain available. Once the provider commits a successful load, later request cancellation does not unload that script; use `unload_frida_script` to release it. If script cleanup fails, the script remains owned and a paused target cannot resume until the pending instrumentation is released. Queued script loads, resumes, and unloads check cancellation before dispatching native effects. Cancellation Evidence separately reports whether a target resume was observed, whether instrumentation cleanup completed, and that final target liveness remains unknown. REA does not kill the target. Remote discovery retains its observations and reports a separate cleanup error if the temporary connection cannot be removed.
+Frida scripts execute inside the selected target with that process's memory, filesystem, and network authority. Local device lookup waits 1000 ms. This release has no runtime override for that lookup; another REA timeout does not raise it. Remote endpoints do not use that lookup. Message capture is bounded to 16 MiB and reports truncation. Metadata, type, and text field caps, and the JSON node cap, are also fixed. Evidence names a cap when that observation actually dropped data. This release has no runtime override for those capture caps. Script source is not stored in Evidence; its selected path and SHA-256 digest are. For a file, `source_sha256` identifies the original selected bytes. Invalid UTF-8 continues to use replacement decoding; Evidence then records that limitation and `raw_result.source_representation` with the encoding, decoding mode, and `loaded_source_sha256` of the UTF-8 string passed to Frida. Valid UTF-8 files and inline scripts do not need this additional representation fact. Evidence retains non-authentication target selection such as device or endpoint, mode, program or PID, and spawn arguments. Authentication values remain excluded. Results report the observed session and messages, while unsupported platform metadata remains unknown. A request cancelled before a Frida operation starts has no provider effects. If cancellation arrives during a native Frida call, REA waits for that call to return, then closes any session it created and returns the observed session as a partial observation. A cancelled script load on an existing session releases only its new script before queued operations can run; the session and previously loaded scripts remain available. If cancellation is observed when the native call returns, partial Evidence retains the source identity and captured messages even if that call rejects, including messages emitted during awaited cleanup. The caller's cancellation reason remains primary and a native rejection is retained as a sanitized diagnostic. A failure already observed before cleanup stays primary if cancellation arrives during cleanup. The cancellation cutoff is REA's observation of native completion, rather than a claim about unobservable native event order. `script_id` is REA's load-attempt identity. When creation rejects without returning a handle, `script_cleanup: "no_handle"` means REA acquired no handle to unload; it does not prove that the engine performed no work. Once the provider commits a successful load, later request cancellation does not unload that script; use `unload_frida_script` to release it. If script cleanup fails, the script remains owned and a paused target cannot resume until the pending instrumentation is released. Queued script loads, resumes, and unloads check cancellation before dispatching native effects. Cancellation Evidence separately reports whether a target resume was observed, whether instrumentation cleanup completed, and that final target liveness remains unknown. REA does not kill the target. Remote discovery retains its observations and reports a separate cleanup error if the temporary connection cannot be removed.
 
 The [Frida Core API](https://frida.re/docs/frida-core-api/) documents device management, sessions, and script loading; the [JavaScript API](https://frida.re/docs/javascript-api/) documents target-side instrumentation. [Luma](https://luma.frida.re/) is Frida's standalone GUI; REA does not automate or control it. `npm run verify:frida` requires a Linux host with the optional Frida binding and uses a short-lived local Node.js fixture created by the verifier. Add `-- --address HOST:PORT --pid PID` to run the separate remote attach lane against a caller-selected target. Mock tests do not establish real device/server compatibility.

docs/ghidra-dos.md

@@ -23,13 +23,19 @@ not name its entry function `entry`. MCP uses `open_binary`, `list_procedures`,
 `procedure_info`, `read_function_instructions`, and `analyze_function` with the
 same admitted target and provider semantics.
 
-The classifier validates MZ page counts, initialized module length, header and
+The DOS MZ parser validates page counts, initialized module length, header and
 relocation-table bounds, relocation destinations, and the initialized entry
-point. It distinguishes DOS load-module and relocation bytes from the Windows
-new-header field. PE remains PE; NE, LE, and LX declarations are explicitly
-unsupported. A damaged declared Windows header does not silently become DOS.
-Inventory classification returns `unknown` when its bounded prefix cannot
-establish the format; this is distinct from the complete target admission check.
+point using the available structural bytes and actual file length. It reports
+those DOS facts separately from Windows-header evidence. A recognized PE
+signature remains PE; recognized NE, LE, and LX headers are explicitly
+unsupported. A nonzero Windows-header pointer without a recognized signature
+remains an unresolved candidate, even when the DOS structure validates:
+automatic target selection fails closed and reports both observations.
+REA does not infer DOS intent from padding or infer a damaged Windows header
+from the pointer alone. Inventory classification is a bounded-prefix format
+hint; it returns `unknown` for that candidate or when its prefix cannot
+establish the format. That result is distinct from complete DOS validation and
+does not decide target admission.
 
 Ghidra imports a private target snapshot with `MzLoader`, language
 `x86:LE:16:Real Mode`, and compiler specification `default`. These settings,
@@ -124,6 +130,10 @@ FileBytes digests, every source mapping and its memory digest, relocation record
 module partitions, relocation fixups and linear coordinates rather than relying
 on the imported executable SHA alone.
 
+Expected DOS relocation records retain each fixup's pre-application bytes and
+applied value. Their memory-byte windows come from the final relocated image,
+including changes made by later overlapping fixups.
+
 The result is `verified`, `mismatch`, or `unsupported`. Every check retains its
 expected and observed values. A memory-digest mismatch identifies the source
 range's file offset and analysis address; it does not locate the first differing
@@ -134,6 +144,10 @@ their contents and DOS runtime semantics are not verified. Independent verificat
 also covers explicit [raw-binary imports](raw-binary.md); other formats return
 `unsupported` with measured observations.
 
+MZ verification retains the initialized image described by the header's page
+count and streams appended overlay bytes into the complete original and modified
+source digests. Overlay size does not increase the retained image allocation.
+
 `read_bytes` returns initialized provider memory, including loader fixups; an
 unmapped or uninitialized byte ends the read with an explicit returned length and
 `complete: false`. `address_to_file_offset` uses the provider's source mapping;

docs/ghidra-nativeaot.md

@@ -69,6 +69,12 @@ Pass a returned MethodTable address to `inspect_native_data_type`. A direct PE
 query also derives the snapshot metadata from the authenticated Ghidra session
 handshake, so it does not require a preceding load-image call. In a persistent
 MCP session, load-image followed by type inspection reuses that same derivation.
+The bridge resolves the caller's address spelling once; metadata lookup uses
+its canonical default-space identity. Bare hexadecimal, prefixed hexadecimal,
+and encoded default-space selectors therefore select the same MethodTable.
+Other address spaces and unresolved identities are not matched by offset alone.
+Metadata already supplied by the engine is retained without reading or deriving
+an additional source snapshot.
 The database result remains `unavailable` when Ghidra has no DataType at that
 address; its `metadata_recovery.source` is `read-only-derived-overlay` and
 carries related type, interface, slot, and target observations. Generated

docs/ghidra-provider.md

@@ -65,6 +65,14 @@ PDB, or change analyzer settings.
 
 ## Loaded addresses and file offsets
 
+Ghidra computes overview procedure and string counts from the same inventories
+used by `list_procedures` and `list_strings`, without transporting those rows
+solely to count them. Counts include external functions and only Ghidra-defined
+strings; they do not establish discovery of unidentified functions or undefined
+string bytes. Complete inventories remain available through the list tools.
+Overview reuse follows the target/profile snapshot lifetime and is invalidated
+by session metadata edits.
+
 `rea analyze <path> --provider ghidra --json` and MCP `binary_overview` report
 the observed loaded `image_base` in `normalized_result`. They retain each
 segment's reported `image_base` and `address_space`. A provider that does not
@@ -139,6 +147,19 @@ overview merely because its source artifact is unchanged.
 
 ## Admitted function-analysis semantics
 
+Concurrent callers share one startup while retaining independent cancellation.
+Only the final departing caller abandons startup and waits for its cleanup.
+New attempts wait for that cleanup; failed cleanup retains its original error
+and owned directories until a successful close. Closing the provider reports a
+provider-state failure to live callers rather than cancelling their signals.
+
+Configured analysis extensions and seeds use owned file reads during profile
+resolution and launch snapshots. A changed read extent or descriptor state
+refuses the commitment. Failed source or snapshot descriptor closes remain
+owned and block new admission until cleanup succeeds; the physical failure
+and exact cancellation reason remain available. Provider shutdown joins all
+profile stages before retrying their retained descriptors.
+
 | Concern                 | Ghidra contract                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                            |
 | ----------------------- | -------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- |
 | Decompiler lifetime     | One persistent `DecompInterface` is opened for the imported Program and disposed during bridge shutdown. Native decompilation has no fixed per-function deadline. External functions or functions without bodies return `null`; cancellation and native failure remain distinct.                                                                                                                                                                                                                                                                                                                                           |
@@ -193,7 +214,11 @@ these tools before compilation. Keeping this matrix separate lets Linux
 host/provider acceptance run with only the host compiler.
 
 `npm run verify:ghidra:windows` checks source-owned native x86-64 PE applications;
-`-- --x86` selects x86 and `-- --dll` selects DLLs, and the flags can be combined.
+`-- --x86`, `-- --arm`, or `-- --arm64` selects another target architecture;
+`-- --dll` selects DLLs. Choose one architecture and optionally combine it with
+`--dll`. These static fixtures contain source-owned encoded instructions; they
+are not compiler-produced programs. See the [Windows verification scope](windows-ghidra-p0.md#native-authority-and-target-identity)
+for real-provider coverage and exclusions.
 These lanes check all 26 read-only operations, target/snapshot/import SHA-256 linkage,
 and project, endpoint, process, and runtime cleanup. The independent native
 lane checks DACLs, handle admission, cancellation, and Job Object lifecycle.

docs/go-binaries.md

@@ -19,6 +19,56 @@ requires an absolute filesystem path and preserves that exact selection:
 }
 ```
 
+For build identity and counts, select `detail: "summary"`. MCP retains the same
+complete Evidence and returns `normalized_result.parent_evidence_id`. Read
+selected dependency versions without reopening the file:
+
+```json
+{
+  "name": "inspect_analysis_view",
+  "arguments": {
+    "source": {
+      "kind": "retained-evidence",
+      "evidence_id": "ev_<parent digest>"
+    },
+    "view": {
+      "kind": "go",
+      "collection": "dependencies",
+      "paths": ["golang.org/x/crypto"],
+      "fields": ["path", "version", "replacement"],
+      "offset": 0,
+      "limit": 10
+    }
+  }
+}
+```
+
+Go views support `dependencies` (exact `paths`) and `settings` (exact `keys`),
+optional original `indices`, field selection, and pagination. Filters intersect;
+empty filter arrays select no rows and `fields: []` returns only row identity.
+Each row keeps its original `index`, scoped to the parent Evidence and collection;
+duplicate module paths and setting keys keep distinct indices. Page offsets
+count matching records. Exhausting retained rows does not establish complete
+runtime dependencies. Original bytes, unknown lines and all complete metadata
+remain in the parent and can be exported as an Evidence bundle.
+
+One-shot CLI summaries save complete Evidence before returning the projection:
+
+```bash
+rea inspect-go-binary ./application --detail summary --evidence_output ./go-evidence.json --json
+```
+
+`--evidence_output` is required for CLI summary delivery and never overwrites an
+existing file. The default CLI/MCP reply remains complete.
+
+[Measured workflows](https://github.com/morluto/rea/blob/42e1a17d6e6c0487f17be0bae248553c4caba7f7/experiments/analysis-efficiency/go-build-views.md) include
+catalog repetition and extra calls. Identity/count questions benefit from a
+single summary. In the tested Codex full-catalog configuration, a compound
+question about identity and one dependency uses fewer tokens in captured request JSON with
+one complete inspection than with a summary followed by a lookup. The complete
+reply remains useful for that initial question; retained views support later
+selected reads. Token figures are serialization estimates, not model usage.
+
 ## Observations
 
 The inline Evidence result contains the artifact path, SHA-256, byte length,
@@ -39,8 +89,8 @@ container format, architecture, word size and byte order. A recognized
 - `module.dependencies`, with the same fields and their original order;
 - `module.settings`, retaining ordered key/value entries and duplicate keys;
 - `module.unparsed_lines`, `module.unparsed_line_bytes_base64` and
-  `module.complete`, so unfamiliar, malformed, conflicting or undecodable records
-  do not disappear behind a complete result.
+  `module.complete`, which says whether every present module-text line was
+  parsed. It does not establish that the dependency inventory is complete.
 
 Missing checksums are `null`; empty fields and development/local replacement
 versions remain unchanged. Build settings such as `GOOS`, `GOARCH` or `vcs.revision`
@@ -50,19 +100,43 @@ base64 while other decodable module records are still reported; the reader does
 not insert replacement characters or reject the entire artifact for those
 strings.
 
+When no main-module (`mod`) record was decoded, `module.main` is `null` and
+dependency coverage is unknown. Successfully decoded dependency records remain
+visible. GOPATH-mode builds may contain only a package `path` and build settings;
+their empty dependency array does not establish the absence of dependencies.
+The result's limitations state this explicitly. A module-mode build with a main
+module record and no dependency records has an observed empty dependency list.
+
 ## Supported boundary
 
-The built-in TypeScript reader supports ELF, PE and thin Mach-O images with Go's
-inline or historical pointer build-info encoding. Source ranges refer to the
-original stable file snapshot, including pointer-backed strings. It does not
-scan arbitrary byte strings as compiler metadata or interpret a fat Mach-O
-container as one selected architecture.
+The built-in TypeScript reader supports ELF, PE and thin Mach-O images, and one
+selected slice of a universal Mach-O, with Go's inline or historical pointer
+build-info encoding. Source ranges refer to the original stable file snapshot,
+including pointer-backed strings. Legacy pointer-backed fields are decoded at
+the width declared by Go's build-info record, independently of the image's ELF
+class. It does not scan arbitrary byte strings as compiler metadata.
+
+A universal (FAT) Mach-O needs `architecture` (`x86`, `x86_64`, `arm`
+or `arm64`; CLI `--architecture`). REA reads its own FAT table, without `lipo`,
+and validates the selected slice's range, alignment and Mach-O header CPU
+against the table. The result keeps the universal file as `artifact` (its
+SHA-256 and size), adds `selected_slice` with the slice offset, length, raw CPU
+type/subtype and the FAT table digest, and reports every source range as a
+file offset inside that slice. Without `architecture`, or when the file has no
+such slice or several, the request fails on `architecture` and lists the slices
+present. `architecture` on any other image is rejected rather than ignored.
 
 For ELF images without a named build-info section, discovery checks each
 nonempty writable, non-executable load segment in program-header order. This
 includes sectionless PIE images with metadata in a later data segment. Named
 sections retain priority, and alignment, file mapping and record validation
-remain required.
+remain required. When ELF section names are unreadable, limitations report that
+coverage gap; the fallback search policy is unchanged.
+
+A named ELF or Mach-O build-info section selects the record's discovery range.
+For inline encoding, the following version and module strings may continue past
+that section into its file-backed load mapping; their bytes are still checked
+against the mapping and the existing aggregate string limit.
 
 Historical moduleless Go binaries can store their empty module-string header in
 declared zero-initialized virtual memory rather than file-backed bytes. The reader
@@ -84,11 +158,17 @@ host permission denial, changed artifacts and cancellation retain distinct
 errors. The selected file is never modified; this inspection creates no
 temporary files, starts no subprocesses and accesses no network endpoint.
 
+An input above 256 MiB returns `resource_constraint` with resource `file-size`
+and `reported_limits.maximum_input_bytes`. Select a suitable input within that
+fixed allowance; changing an output-write or transport budget does not admit it.
+
 Resource guards bound the selected file and cumulative header searches to 256 MiB
-each, aggregate embedded strings
-to 1 MiB, structural table decoding to 16 MiB and the complete normalized metadata
-report to 8 MiB. Identical search spans with the same virtual alignment are
-searched once; partially overlapping spans consume the cumulative search budget
+each, aggregate embedded string contents to 1 MiB (varint length prefixes are
+framing and do not count toward this allowance), structural table decoding to
+16 MiB and the complete normalized metadata report to 8 MiB. Inline strings must
+also fit within their actual mapped source range. Identical search spans with
+the same virtual alignment are searched once; partially overlapping spans
+consume the cumulative search budget
 and fail with a CPU resource constraint when it is exhausted. Oversized results
 fail without partial success. MCP also applies
 its [response transport budget](mcp-contracts.md), which counts both text and

docs/godot-pck.md

@@ -0,0 +1,113 @@
+# Godot PCK archives
+
+`inspect-artifact` and `extract-artifact` read ordinary Godot PCK formats 2, 3
+and 4 without running the game or requiring Godot. Official Godot 4.4.1, 4.6.1
+and 4.7.1 produced these respective formats in verification; other engine
+releases have not been verified. Format 1 is unsupported.
+
+Inspection retains the original header and directory facts, including engine
+version, flags, declared file count, source paths, offsets and declared MD5.
+Observed plaintext bytes receive content-addressed SHA-256 identities. MD5 is a
+container integrity claim, separate from the observed SHA-256. A mismatch fails
+by default; `--integrity-policy record-and-continue` retains the observed bytes
+and an explicit untrusted contradiction. An all-zero MD5 is an unknown claim,
+not a verified digest.
+
+## CLI
+
+Start with a compact summary and keep the complete Evidence bundle:
+
+```sh
+rea inspect-artifact /artifacts/game.pck --detail summary \
+  --evidence_output /evidence/game.json --json
+rea extract-artifact /artifacts/game.pck --json
+```
+
+Summary delivery requires `--evidence_output` and never overwrites that file.
+The response contains `summary` Evidence and an `evidence_bundle` receipt.
+Without `--detail summary`, inspection keeps its complete inline delivery.
+Extraction creates a new temporary directory and returns its path. It rejects
+unsafe paths and verifies container identity before creating output.
+
+The same CLI commands locate a PCK at the beginning of a file or through its
+12-byte EOF trailer in an executable carrier. They do not scan arbitrary
+payloads or discover packs behind additional trailing signatures. The root
+content SHA-256 covers the whole carrier; its content classification remains
+native, while the inventory records the PCK role and exact pack bounds.
+An EOF marker is only a candidate: an invalid size or a pointed-to location
+without a `GDPC` header leaves native classification in place. Once the
+trailer points to a pack header, malformed or unsupported PCK content remains
+an error. This recognition boundary follows the Godot loader at pinned source
+revision [06d1e7a](https://github.com/godotengine/godot/blob/06d1e7a34ccde4029eed53d018cfde77dfe771d5/core/io/file_access_pack.cpp#L2099-L2146).
+Ordinary native `open_binary` detection remains unchanged.
+
+## MCP and selected facts
+
+For a standalone pack, call `open_binary` with its absolute path, then
+`inspect_artifact` with `{"detail":"summary"}`. The complete inspection and its
+nested inventory stay in the session. Use the summary's
+`normalized_result.parent_evidence_id` as the source of subsequent views:
+
+```json
+{
+  "name": "inspect_analysis_view",
+  "arguments": {
+    "source": {
+      "kind": "retained-evidence",
+      "evidence_id": "<parent_evidence_id>"
+    },
+    "view": {
+      "kind": "inventory",
+      "collection": "occurrences",
+      "logical_paths": ["scripts/main.gd", "project.binary"],
+      "fields": [
+        "logical_path",
+        "declared_integrity",
+        "hash_status",
+        "format_metadata"
+      ],
+      "offset": 0,
+      "limit": 32
+    }
+  }
+}
+```
+
+Occurrences retain `occurrence_id`, `artifact_id` and `parent_occurrence_id`
+even when `fields` is empty. Batch occurrence IDs with `occurrence_ids`, or
+join returned content IDs to `collection: "nodes"`, `artifact_ids: [...]` and
+`fields: ["sha256", "size"]`. Filters intersect; an omitted filter selects all
+available rows and an empty array selects none. Pagination follows the filtered
+canonical order. Multiple occurrences may share one content ID.
+
+Views reuse authenticated immutable inventory facts and lazy lookup indexes;
+they do not reopen the source. These are historical observations, including
+after the source changes or disappears. Close the target with `close_binary`
+when finished. To read an exported bundle through the CLI, select its complete
+inspection record as `source: {kind: "inline", evidence: ...}` in a JSON request
+file for `rea inspect-analysis-view request.json --json`.
+
+## Coverage and limits
+
+PCK summaries expose root format metadata, available graph counts,
+`available_hash_status_counts`, and `source_coverage`. Inventory pages preserve
+producer coverage alongside page coverage. `source_complete` means all declared
+retained graph rows are available; it does not establish that an encrypted or
+otherwise unknown directory was examined. Missing matches in unknown coverage
+are not evidence of absence.
+
+Encrypted directories retain their header and declared file count with unknown
+entry coverage. Individually encrypted members retain directory metadata with
+unknown content identity; REA does not decrypt them. Sparse bundles likewise
+retain header facts with unknown member coverage. Extraction refuses an
+unreadable directory or encrypted member. This capability does not recover
+keys, decode Godot resources, decompile GDScript bytecode, analyze signals,
+apply patch/removal entries or execute targets.
+
+Readable logical member bytes, including repeated physical ranges, share the ZIP
+archive expansion policy: at most the larger of 256 MiB or 500 times the PCK's
+encoded byte span. Embedded carrier bytes do not enlarge that budget. The full
+directory is checked before any member is yielded, hashed or extracted. Encrypted
+members are not read and do not consume this plaintext-work allowance. Small
+bounded aliases remain valid. This is a processing/output resource bound, separate
+from the directory metadata retention budget and format validity.

docs/gpu-code-objects.md

@@ -0,0 +1,224 @@
+# Static CUDA and HIP code objects
+
+`inspect_gpu_code_objects` answers which GPU targets and kernel declarations a
+local artifact carries. It reads ELF/PE sections, standalone CUDA fatbins,
+CUBIN/PTX and Clang HIP bundles, including CCOB v1/v2/v3. It neither executes the
+artifact nor invokes CUDA/ROCm tools. GPU execution, SASS disassembly and host
+launch correlation remain outside coverage.
+
+Start with a target inventory:
+
+```json
+{ "path": "/artifacts/library.so", "view": { "collection": "targets" } }
+```
+
+Target rows aggregate inventoried payloads and distinguish an unrequested kernel
+analysis from a complete or partial one. `objects` pages expose exact object
+IDs, compression declarations and encoded/header locations. Fields, exact
+filters, batch IDs and stable pagination select the facts delivered to the
+caller. IDs are positional within the recorded artifact SHA-256. Selecting `object_ids`
+in a new inspection requires `expected_sha256` from that inventory; a changed
+file is rejected before parsing selected payloads.
+
+`target` is the inventory declaration used for selection; `payload_target` is
+the independently observed inner declaration when available. Object pages keep
+both values, their kind, `target_consistency` and inspection coverage even with
+selected fields.
+Conflicting base architectures retain both values, mark object/report/target
+coverage partial and annotate the affected kernels. CUBIN flags establish a base
+architecture: `base-only` explicitly leaves feature-profile agreement unknown,
+including `a` and `f` variants. A standalone declaration is `single-source`, and
+unexamined payloads remain `unexamined`. An exact AMDGPU target is normalized at
+its note owner for comparison with a Clang bundle declaration; raw values remain
+intact. Unresolved feature comparisons stay `unknown`.
+
+A HIP target ID identifies a toolchain target, not its payload representation.
+Recognized raw or wrapped LLVM bitcode stays `unsupported` for kernel metadata
+inspection, with the bundle ID, range and selected payload digest retained.
+Bitcode validity and inner targets remain unexamined; malformed recognized ELF
+still reports `malformed`. Standalone bitcode is outside GPU container coverage.
+
+HIP bundle and metadata note recognition accept both the legacy `amdgcn` triple
+and LLVM's `amdgpu` subarchitecture triples, including declared generic processors
+such as `gfx11-generic` and `gfx9-4-generic`. Notes with the empty or explicit
+`unknown` environment use the same processor and feature comparison while
+retaining their raw triple spelling. CUDA's legacy fatbinary compression flag is
+retained as an unsupported codec with its encoded and decoded extents;
+compressed bytes are not treated as PTX text. CCOB is recognized by the exact
+four signature bytes `CCOB`; high-bit-corrupted bytes are not decoded as ASCII
+and accepted as a bundle. CCOB v1 has no stored total size: the completed zlib
+or Zstd stream determines its encoded extent. Embedded `CCOB` text is ordinary
+compressed data; zero-byte alignment after the stream belongs to neither
+payload. If decoding fails or its codec is unsupported, a later `CCOB` marker
+is only an uncertain recovery candidate. The affected row retains that
+limitation and discovery remains partial.
+
+The declared decoded length bounds decompression. An understated length can
+stop decoding before codec completion, so any subsequent recovery extent stays
+unknown; the decoder does not expand its allowance to prove the bad declaration.
+Interrupted observations omit unfinished v1 rows. A terminal capacity row
+qualifies its retained range as the containing tail and does not infer siblings.
+
+`cuda_target_header` retains the OS ABI, normalized raw `e_flags` and exact
+header location. CUDA OS ABI `0x33` and the observed `0x41` flags layout use
+different architecture fields; they are decoded at one ABI-aware owner.
+Unrecognized layouts keep the header and an unknown target instead of
+inventing an SM label. Header provenance is validated against its CUBIN object.
+
+Target filters select inventory declarations. An unselected payload's inner
+target remains unexamined; an empty filtered result does not establish that
+no enclosed code declares that target. Discovering every inner declaration
+requires decoding those payloads.
+
+HIP bundle IDs retain the declared processor and v2-v3 `+feature` suffixes for
+exact selection. If a recognized AMDGPU bundle ID does not expose a target, the same
+bounded note-discovery owner used for standalone AMDGPU files establishes it
+before filtering. Those directory/note reads share the inspection work budget;
+successive queries can select the target returned by inventory.
+
+Decode selected targets and return a kernel page in the same call:
+
+```json
+{
+  "path": "/artifacts/library.so",
+  "analysis": "kernels",
+  "targets": ["sm_100"],
+  "view": {
+    "collection": "kernels",
+    "names": ["kernel_name"],
+    "fields": [
+      "name",
+      "parameter_count",
+      "parameter_bytes",
+      "parameters",
+      "registers",
+      "shared_bytes"
+    ]
+  }
+}
+```
+
+Only selected payloads are decoded. Other inventory entries remain present with
+`not-selected` states, and overall payload coverage remains partial. An empty
+kernel page proves no matching retained fact; it does not establish that an
+unexamined payload contains no kernel. Missing fields remain null.
+
+Kernel representations have different meanings:
+
+- PTX reports parameter declarations and virtual registers. Shared bytes cover
+  supported static declarations inside that entry; they do not establish module
+  shared memory, dynamic allocation or physical device resources.
+  PTX 1.0–1.3 declares entry parameters in the body; PTX 1.4 and later uses the
+  entry parameter list. Entry-scoped `.pragma` statements before the body do
+  not end the declaration or suppress its body resources. A `.target
+  compute_xx` spelling is retained as written;
+  NVIDIA PTX defines it as a synonym for `.target sm_xx` when comparing its
+  target identity with the fatbin declaration. `.callprototype` formal
+  parameters are excluded from entry virtual-register counts. Parameter names
+  remain available when pointer qualifiers are understood or an unsupported
+  qualifier prevents layout decoding. A pointer's `.align` describes the
+  pointed-to memory and does not change parameter layout alignment. Spaced,
+  joined and partly joined pointer qualifiers have the same meaning. Parameter
+  alignment before or after the scalar type retains its layout and name.
+  Parameter names and following declarations are retained when an array
+  extent uses a parenthesized expression or when dimensions are
+  multidimensional, even though those forms do not currently yield a layout.
+  PTX target selection uses the highest-numbered module-scope `.target`
+  directive, including directives after entry declarations; unsupported or
+  conflicting target profiles leave the effective target unknown.
+  Scalar register declarations count each bare or percent-prefixed name and
+  each parameterized range, including compact keywords such as `.reg.b32`.
+  Unsupported declaration forms or overflowing totals
+  retain an unknown count and partial decoding coverage. Unsupported parameter
+  layouts and unknown shared-memory declarations likewise retain their known
+  facts, explicit limitations and partial coverage through retained pages.
+- CUBIN identifies CUDA entry symbols and reads numeric `nv.info` records,
+  including V2 kernel-parameter sizes, and native section declarations. Global
+  register declarations also cover compiler
+  versions that leave the older text-section register field empty.
+  Conflicting register declarations keep the physical count unknown, preserve the
+  kernel's other facts, and mark object/report/target decoding coverage partial.
+- AMDGPU reads v3+ MessagePack notes. V3 processor identity comes from the
+  `EF_AMDGPU_MACH` field in ELF `e_flags`; known processor values are selectable
+  even when the note omits `amdhsa.target`. This inventory reports the processor
+  only; it does not interpret version-specific feature flags. Unknown processor values remain unknown. A known
+  processor conflict between ELF flags and note metadata leaves the processor
+  unresolved. Argument counts include hidden ABI arguments
+  when present; SGPR/VGPR counts and fixed group/private segment sizes retain
+  their declared meanings. Legacy YAML metadata remains unsupported.
+
+Each selected PTX payload becomes a retained UTF-8 text artifact with a SHA-256,
+byte length and source coordinates. It is omitted from compact replies. Read it
+with `inspect_analysis_view` and the parent Evidence ID returned by the compact
+receipt:
+
+```json
+{
+  "source": { "kind": "retained-evidence", "evidence_id": "ev_..." },
+  "view": {
+    "kind": "gpu-code-objects",
+    "collection": "ptx",
+    "object_id": "...",
+    "offset": 0,
+    "limit": 4096
+  }
+}
+```
+
+Text offsets use UTF-16 code units and must fall on Unicode scalar boundaries.
+Complete Evidence retains the selected text; `export_evidence_bundle` provides
+its portable source record. Historical pages never reopen the input file.
+
+A location with `container_id: null` uses original file offsets. Other locations
+use that inventoried object's decoded byte coordinates. Follow its parent,
+encoded location and digest to trace decompressed metadata back to a native
+section and the original file. Decoded offsets are never presented as physical
+file offsets. Reads from an ELF or PE GPU section stay within that declared
+section extent, so neighboring native section bytes cannot complete a
+truncated embedded bundle header.
+
+The CLI uses the same service and contracts:
+
+```sh
+rea inspect-gpu-code-objects /artifacts/library.so --analysis kernels \
+  --targets '["sm_100"]' --detail summary \
+  --view '{"collection":"kernels","names":["kernel_name"]}' \
+  --evidence_output /analysis/library-gpu.evidence.json --json
+```
+
+A full-file SHA-256 always streams every byte. Inventory then reads bounded
+native directories and bundle headers; CCOB requires decompression to discover
+its inner directory. Kernel analysis reads selected payloads. `bytes_read`
+includes hashing and subsequent descriptor reads, so this is not a header-only
+hashing workflow.
+
+CUDA raw LZ4 blocks and Zstd, and CCOB zlib/Zstd are supported. Unknown versions
+or codecs retain declarations with explicit limitations. Lengths are checked
+before decoding, and CCOB's truncated MD5 is verified against decoded bytes.
+Resource limits are one active inspection, 128 MiB per input/output buffer,
+512 MiB of payload input, decoded output and AMD note-read work per inspection,
+and 16 MiB retained facts including selected PTX. Payload work is reserved before
+reading or allocating; a capacity refusal does not reread an aliased payload.
+Native tables and decoded section names each have a 16 MiB structural budget;
+AMDGPU ELF note discovery admits at most 8 MiB of source note bytes, with
+bounded depth and container/string allocations. Standalone AMD targets are read
+from those notes before target selection; selected kernel analysis reuses the
+same parsed declarations. Pages have a
+1 MiB response budget. Resource refusals and interrupted observations preserve
+known facts and explicit uncertainty.
+
+CUDA ELF parsing additionally shares a 262,144 record/symbol visit limit,
+128 MiB of cumulative metadata section reads, and a 16 MiB serialized-size
+accounting budget for temporary index facts. These limits are checked before
+index growth; they are not claims about JavaScript heap size. Equivalent scoped
+metadata byte ranges and repeated equal register declarations are coalesced;
+conflicting values remain unknown. Parameter sections are parsed once per ELF.
+
+Missing CUBIN symbol tables or unresolved section names produce `partial`
+object, report and target coverage even when some kernel facts survive. AMD note
+failures preserve earlier successfully parsed declarations and the original
+failure reason. Standalone PTX recognition tries a small header first, then
+continues within the existing 16 MiB text budget when leading comments or
+whitespace obscure its directives; admitted full-file probe bytes are reused.
+After those comments and whitespace, `.version` must be the first PTX module
+token; later PTX-looking text does not identify an arbitrary binary as PTX.

docs/harmony-analysis.md

@@ -36,8 +36,9 @@ The MCP tool `project_harmony_application_graph` takes the same
 
 - exact component paths, artifact IDs, and hashes (manifests, Ark bytecode,
   resources, native libraries, JavaScript, signing)
-- a packaging-model hint: Stage is claimed only from a `module.json` entry,
-  FA only from `config.json`, otherwise `unknown`
+- a packaging-model hint: Stage is claimed only from a `module.json` entry
+  outside a classified resource path, FA only from `config.json` outside a
+  classified resource path, otherwise `unknown`
 - FA resources under `assets/<module>/resources/` as well as Stage `resources/`
 - unavailable component paths with unknown IDs and hashes, including encrypted entries; these make projection coverage partial
 - App Pack child package paths without recursive inventory
@@ -45,6 +46,12 @@ The MCP tool `project_harmony_application_graph` takes the same
 - path-based N-API bridge hypotheses between bytecode and `libs/<abi>` native
   libraries
 
+File and slice occurrences supply component and child-package paths; directory
+trees and symlinks retain their identities in the source inventory without
+being classified as code files. Ark and JavaScript runtime hints summarize the
+classified bytecode and JavaScript lists; supported source extensions share the
+same classification rule as their component rows.
+
 Every claim is path-presence only. Manifest semantics (module identity,
 abilities, entry points, permissions, dependencies), Ark bytecode contents and
 versions, resource resolution, signing verification, and observed native

docs/ida-provider.md

@@ -8,6 +8,8 @@ Choose one lifecycle mode in an upstream MCP JSON registration. Set `REA_IDA_MCP
 
 Each mode accepts one upstream tool profile, checked against `tools/list` before any analysis runs. The attached profile requires the legacy GUI tool names; upstream 2.0.0 renamed those tools, and REA has not established an equivalent attached-mode lifecycle and ownership contract for that catalog. It therefore fails closed with the missing tool names in the error diagnostics.
 
+Startup failures identify the accepted REA analysis operation. Missing-profile and typed startup transport failures retain the original stage in `details.diagnostics.startup_operation`; existing diagnostics, timeout budgets and resource limits remain available. Startup remains shared within a session; each caller receives its own operation attribution. Cancellation retains live cleanup ownership, so a successful retry releases the originating cancellation outcome as well.
+
 The headless profile is defined by the database-supervisor API, not by an upstream version string. The upstream source snapshot [`c133c385`](https://github.com/mrexodia/ida-pro-mcp/tree/c133c3853faa111a9b00ee615c013b720d0c4acd) includes `idb_close` (introduced by [upstream PR #495](https://github.com/mrexodia/ida-pro-mcp/pull/495)) alongside `idb_open` and `idb_list`. Earlier builds may report the same `2.0.0` version while lacking that tool, so REA checks the live catalog rather than treating a version number or branch name as a compatibility guarantee.
 
 | Mode     | Upstream profile    | Known working                                          |
@@ -39,7 +41,16 @@ rea function /absolute/path/to/program main --json
 rea decompile /absolute/path/to/program main --json
 ```
 
-Open the **original input binary** in REA, not an `.idb` or `.i64` database. Its SHA-256 must match IDA's recorded input hash. REA checks the attached identity and image base around each analysis request. A mismatch or target switch fails with recovery guidance. The original input must remain readable by REA; WSL can use its mounted alias while the GUI runs on Windows.
+If the registration cannot be read or validated, provider diagnostics retain its
+selected `registration_path` and the configuration owner's setting/constraint
+facts in `configuration_details.settings`, plus `configuration_remediation`.
+`configured` means a path was selected, not that it is valid or an IDA connection
+was established; failed registrations remain unavailable. Correct
+`REA_IDA_MCP_CONFIG` in the next CLI invocation's environment or the MCP server's
+launch environment. Restart/reconnect that server after changing its environment;
+an export in another shell does not update a running server.
+
+Open the **original input binary** in REA, not an `.idb` or `.i64` database. Its SHA-256 must match IDA's recorded input hash. REA checks the attached identity and image base around each analysis request. A mismatch or target switch fails with recovery guidance. The original input must remain readable by REA; WSL can use its mounted alias while the GUI runs on Windows. Attached profiles, including native mode, refuse universal Mach-O inputs at admission: every slice shares the recorded whole-file hash, so REA cannot verify which slice the GUI loaded. Open a thin file extracted with `lipo -thin` in both IDA and REA, or use headless mode, which analyzes the selected slice.
 
 REA's close operation releases its MCP connection or proxy. It never saves or closes the existing GUI database. External edits remain visible because results are live and are not replayed from snapshots. These checks do not lock the GUI or prove that the database is unpatched.
 
@@ -131,7 +142,7 @@ Use a package version that actually includes this adapter; repository main can b
 {"name":"close_binary","arguments":{}}
 ```
 
-Opening an MCP target starts its selected provider session; headless opening waits for auto-analysis. Set the agent's MCP request deadline long enough for the target. The registration's optional `timeoutMs` defaults to 300000 for upstream requests. An upstream SDK request that reaches this limit returns `provider_timeout` with its operation and actual request budget. A freshly opened headless worker can briefly report `status: "busy"` from `server_health`; REA re-reads health every 100 ms until the worker answers, the caller cancels, or `timeoutMs` has elapsed since the first read. Each health request is capped by the time remaining in that overall wait, so a stalled read cannot extend the readiness deadline by another full request timeout. Expiry while the worker is still busy produces a retryable `provider_timeout`; no further read is started after the deadline. That error names the busy tool from the last observation, or `unknown` when the worker reported none or a value that is not tool-name-shaped. If a health request itself consumes the remaining wait budget, the same last-busy timeout is returned with the stalled request's captured output. Cancellation during the wait produces `cancelled`. Any other unexpected health payload, including a busy payload with an invalid shape, remains `unreadable_output`. Cancellation drains the current upstream request before cleanup; it does not force-stop GUI analysis. An open request that times out without a completed response retains the private workspace because worker inventory alone cannot prove the request has stopped. `binary_session` reports actual tool availability. Repeated doctor/setup calls are unnecessary when the registration already works.
+Opening an MCP target starts its selected provider session; headless opening waits for auto-analysis. Set the agent's MCP request deadline long enough for the target. The registration JSON named by `REA_IDA_MCP_CONFIG` has an optional `timeoutMs` integer from 1 to 2147483647. Omitted means 300000. REA applies an edited value on the next IDA connection after it re-reads that file; an already open session keeps the timeout it connected with. Changing the `REA_IDA_MCP_CONFIG` path itself is read at process start. An upstream SDK request that reaches this limit returns `provider_timeout` naming the effective milliseconds and `timeoutMs`. A shorter inner request is the time remaining inside that ceiling, and the error names both when they differ. A freshly opened headless worker can briefly report `status: "busy"` from `server_health`; REA re-reads health every 100 ms until the worker answers, the caller cancels, or `timeoutMs` has elapsed since the first read. Each health request is capped by the time remaining in that overall wait, so a stalled read cannot extend the readiness deadline by another full request timeout. Expiry while the worker is still busy produces a retryable `provider_timeout`; no further read is started after the deadline. That error names the busy tool from the last observation, or `unknown` when the worker reported none or a value that is not tool-name-shaped. If a health request itself consumes the remaining wait budget, the same last-busy timeout is returned with the stalled request's captured output. Cancellation during the wait produces `cancelled`. Any other unexpected health payload, including a busy payload with an invalid shape, remains `unreadable_output`. Cancellation drains the current upstream request before cleanup; it does not force-stop GUI analysis. An open request that times out without a completed response retains the private workspace because worker inventory alone cannot prove the request has stopped. `binary_session` reports actual tool availability. Repeated doctor/setup calls are unnecessary when the registration already works.
 
 For guided registration, set `REA_IDA_MCP_CONFIG` before `rea setup --client <client>`. Setup includes that file reference in its reviewed plan and preserves it after approval; it does not copy upstream credentials into the agent configuration. For `ida` or `auto` selection, Codex/Kimi setup and doctor use the effective client registration's parsed `timeoutMs` (default 300000 ms) when recommending the agent's per-tool deadline. Existing positive client deadlines remain caller choices. `rea doctor --provider ida --json` validates the registration without starting IDA and explicitly leaves runtime compatibility and target binding for analysis.
 

docs/il2cpp-metadata.md

@@ -0,0 +1,432 @@
+# IL2CPP metadata
+
+`inspect_il2cpp_metadata` / `inspect-il2cpp-metadata` decodes one selected local
+Unity IL2CPP `global-metadata.dat` file and returns its assembly images, type
+definitions, methods and named members with their owners, plus managed string
+literals, by stable index, namespace, range and count,
+with the header sections and serialized index widths. It reads the file without
+running it and requires no Unity installation, IL2CPP binary, active target,
+Hopper, Ghidra or IDA.
+
+Metadata versions 24, 27, 29, 31 and 39 are supported. For example, Unity 2019.4.41f2
+wrote version 24, 2020.3.49f1 version 27, 2022.3.30f1
+wrote version 29, 2022.3.62f3 wrote 31 and 6000.4.2f1 wrote 39; other Unity
+releases that write these numbers have not been checked. Other versions are
+rejected as unsupported. Version 24 decoding covers the 264-byte header layout
+used by the cited Unity 2019.4 build; earlier 24.x layouts are not inferred from
+the version number and are refused when their header or records differ. The path
+must name a regular file; directories and symbolic links are rejected.
+
+With `binary_path`, the tool also reads the matching native binary and links
+each method to its code address. See [Native method addresses](#native-method-addresses).
+
+## MCP
+
+MCP defaults to a compact summary and retains the complete decoded Evidence in
+the session. MCP requires an absolute filesystem path:
+
+```json
+{
+  "name": "inspect_il2cpp_metadata",
+  "arguments": {
+    "path": "/artifacts/Data/Managed/Metadata/global-metadata.dat"
+  }
+}
+```
+
+The summary retains the complete decoded Evidence in the session and returns a
+linked `inspect_analysis_view` overview: artifact identity, metadata version,
+index widths and `counts` (`sections`, `images`, `types`, `methods`,
+`nested_types`, `fields`, `properties`, `events`, `parameters` and
+`string_literals`) without rows. Use `normalized_result.parent_evidence_id` to page
+the original rows without reopening or decoding the file:
+
+```json
+{
+  "source": {
+    "kind": "retained-evidence",
+    "evidence_id": "<parent_evidence_id>"
+  },
+  "view": {
+    "kind": "il2cpp-metadata",
+    "collection": "types",
+    "offset": 0,
+    "limit": 64
+  }
+}
+```
+
+Collections are `sections`, `images`, `types`, `methods`, `fields`,
+`properties`, `events`, `parameters`, `string_literals`, `method_addresses`,
+`return_types`, `field_types`, `parameter_types` and `code_gen_modules`. Every returned row includes
+its source `index`, scoped by artifact SHA-256 and collection. All collections
+accept batch `indices` and `fields`; collections whose rows have a `name` also
+accept exact `names`. Types also accept `image_indices`, `namespaces` and
+`declaring_types`; methods, fields, properties and events accept
+`declaring_types`. Parameters accept exact `method_indices`, so a batch of
+method indices retrieves their parameters without reading the entire parameter
+table. `parameter_types` accepts the same `method_indices`, resolved through the
+retained `parameters` owners; Evidence without `parameters` refuses that filter. Filters intersect, duplicate indices are returned once, and
+rows retain metadata order. An empty filter selects no rows; `fields: []`
+returns indices only. Omitted fields return the complete selected rows.
+
+For example, after resolving a type index, retrieve selected method facts:
+
+```json
+{
+  "source": {
+    "kind": "retained-evidence",
+    "evidence_id": "<parent_evidence_id>"
+  },
+  "view": {
+    "kind": "il2cpp-metadata",
+    "collection": "methods",
+    "declaring_types": [1555],
+    "names": ["SendMessage"],
+    "fields": ["name", "declaring_type"]
+  }
+}
+```
+
+After finding method indices, parameters can be requested directly:
+
+```json
+{
+  "source": {
+    "kind": "retained-evidence",
+    "evidence_id": "<parent_evidence_id>"
+  },
+  "view": {
+    "kind": "il2cpp-metadata",
+    "collection": "parameters",
+    "method_indices": [9785],
+    "fields": ["name", "method"],
+    "limit": 64
+  }
+}
+```
+
+Historical Evidence that predates these member collections remains queryable.
+A missing collection has an unknown (`null`) summary count, and requesting it
+reports that it was not collected; it is never presented as an empty inventory. The same holds for Evidence that predates binary linking: its summary
+omits `binary`, and `method_addresses`, `method_address_states`,
+`return_types`, `field_types` and `parameter_types` counts are `null`.
+Existing empty collections retain a zero count. Full Evidence is preserved.
+
+`offset` counts matching rows, not source indices. `limit` defaults to 64;
+`coverage.next_offset` continues that filtered page. Coverage describes retained
+metadata facts only; runtime behavior and correspondence to native code remain
+unknown. Each view preserves the parent Evidence link, artifact identity and
+limitations. Queries work after the source file is removed. Select
+`detail: "complete"` to return every row inline within the separate MCP response
+budget.
+
+## CLI
+
+```bash
+rea inspect-il2cpp-metadata /absolute/path/global-metadata.dat --json
+```
+
+The CLI defaults to complete Evidence. For compact delivery, explicitly preserve
+the full Evidence in a new file:
+
+```bash
+rea inspect-il2cpp-metadata /absolute/path/global-metadata.dat \
+  --detail summary --evidence_output /tmp/metadata-evidence.json --json
+jq -c '{source: {kind: "inline", evidence: .records[0]}, view: {kind: "il2cpp-metadata", collection: "methods", indices: [9785], fields: ["name", "declaring_type"]}}' \
+  /tmp/metadata-evidence.json > /tmp/metadata-query.json
+rea inspect-analysis-view /tmp/metadata-query.json --json
+```
+
+Use the exported Evidence with the CLI's inline Evidence source format described
+in [CLI and Evidence](cli.md#run-a-command). Compact delivery requires
+`--evidence_output`; existing files are never overwritten. The full export
+preserves every decoded row for subsequent queries without the original file.
+
+`--binary PATH` links method addresses and `--architecture arm64|amd64` selects
+the slice of a universal Mach-O.
+
+## Observations
+
+The result contains:
+
+- `artifact`: path, SHA-256 and byte length;
+- `magic`: unsigned little-endian header signature (`0xFAB11BAF`, delivered as a JSON number);
+- `metadata_version`: `24`, `27`, `29`, `31` or `39`;
+- `index_widths`: the serialized widths (1, 2 or 4 bytes) of
+  `type_definition`, `generic_container`, `parameter` and `type` indices, and
+  `type_width_source`. Version 39 derives the type index width from the
+  `interface-offsets-record` or the `type-definition-record`; versions 24, 27, 29
+  and 31 always use 4-byte indices (`fixed-width-version`);
+- `sections`: all version-specific header sections in header order, each with `offset`, `size`
+  and `count`. Versions 24 to 31 record no counts, so `count` is `null` and
+  record counts come from the section sizes;
+- `images`: `name`, `type_start` and `type_count`;
+- `types`: `name`, `namespace`, `method_start`, `method_count`,
+  `nested_type_start`, `nested_type_count` and `declaring_type`;
+- `methods`: `name` and `declaring_type`;
+- `fields`, `properties` and `events`: `name` and `declaring_type`, the type
+  whose member range holds the row (`-1` when none does);
+- `parameters`: `name` and `method`, the method whose parameter range holds it;
+- `string_literals`: managed string literals as `value`, or `value_base64` with
+  the exact bytes when they are not valid UTF-8;
+- `limitations`.
+
+A row's position in its collection is its metadata index, so `type_start`,
+`method_start` and `declaring_type` refer to positions in `types` or `methods`.
+`nested_type_start` instead indexes the metadata's `nestedTypes` table, which is
+not returned; each nested type's `declaring_type` carries that relationship.
+A type's `declaring_type` is derived from the `nestedTypes` section and is `-1`
+for top-level types. The start of an empty range can retain the original value
+`-1`.
+
+The complete inventory of a large game is several megabytes and can exceed
+the MCP response budget. Summary delivery and filtered, projected pages keep
+unneeded rows out of agent context without discarding the source Evidence.
+
+## Native method addresses
+
+`binary_path` names the native binary built with the metadata: macOS
+`GameAssembly.dylib` (Mach-O), Windows `GameAssembly.dll` (64-bit PE) or
+Android `libil2cpp.so` (64-bit ELF, AArch64 or x86-64). A
+universal Mach-O needs `binary_architecture` (`arm64` or `amd64`); a thin binary
+rejects it. The file is read, never loaded or executed.
+
+A position-independent `libil2cpp.so` stores zero in most pointer slots and
+keeps the value in `R_AARCH64_RELATIVE` or `R_X86_64_RELATIVE` relocations
+(`DT_RELA`). The reader writes those values into a private copy before
+searching it, so addresses are the link-time addresses and RVAs are relative to
+the first loaded segment. Android packed
+relocations (`DT_ANDROID_RELA`) are refused. No target code is executed.
+
+The binary's `Il2CppCodeRegistration` points to one `Il2CppCodeGenModule` per
+assembly image. Each module record starts with its module name, its method
+pointer count and its method pointer array. The reader finds, for every metadata
+image, the single record whose name pointer reaches that image's name string and
+whose count equals the image's method count. It then requires exactly one
+`codeGenModulesCount`/`codeGenModules` pair that lists exactly those records. No
+symbols are needed, so stripped builds work. A method's token row id selects its
+slot in the module's method pointer array. Within one image, each method must
+claim a distinct MethodDef token row; duplicate rows are refused before any
+linked method table or seed output is published. Row ids may repeat across
+different images, and distinct rows may still resolve to the same address when
+the binary uses identical-code folding.
+
+For ELF, a repeated load mapping that resolves a complete module record to the
+same virtual address and file bytes contributes one record. Distinct virtual
+record addresses remain distinct candidates, and conflicting file mappings
+remain malformed.
+
+Module records establish names through their declared NUL-terminated string pointers. NUL/0xFF padding, spaces, Unicode and suffix-shared storage are supported. A pointer to a longer string does not match just its suffix; a pointer directly into a shared suffix may name it. Exact reversed UTF-8 matching avoids rescanning the file for every name and retains duplicate-name owners until module records resolve them. A contributor investigated a Linux
+`GameAssembly.so` from Unity 2023.2.13f1 whose static registration was not located. When no static pair is resolved on x86-64, the reader searches executable ranges for bytes consistent with RIP-relative LEA/MOV references to module-array candidates. Exactly one referenced array candidate must remain. `code_gen_modules_field` is then `null`, and Evidence is **inferred**: instruction boundaries, the absence of a static registration elsewhere, and runtime use were not verified. The module records and slots themselves remain file observations. Static-table associations are **derived**. Otherwise the result is `code-registration-not-found`.
+
+The binary/metadata pair is caller-selected. Name/count/token agreement is structural evidence, not a cryptographic same-build proof. File-offset locations refer to the primary metadata artifact; linked virtual addresses refer to `binary.artifact`. Both SHA-256 identities remain available in compact native pages as `artifact` and `binary_source`.
+
+Contributor-reported verification (not all rerun by the follow-up): macOS arm64 and x86_64 slices from Unity 2021.3.45f2, 2022.3.30f1, 2022.3.62f3 and
+6000.0.84f1 against dSYM symbols; Android arm64 and x86-64 from 2022.3.62f3
+against `libil2cpp.sym.so`; Windows x64 from 2023.2.13f1 (metadata version 29,
+the MIT-licensed game Between Two Worlds 1.0.5, `GameAssembly.dll` SHA-256
+`a3ae7f6ecd1f671297b7d98d12713ef024062d93cd7a4205298102656c93b67a`) against
+Il2CppDumper, where all 43,198 linked methods share an address with a
+same-named Il2CppDumper method and identical code folding groups 10,448 methods
+into 1,735 addresses. A macOS x86_64 build from Unity 2020.3.49f1 (metadata version 27, which Il2CppDumper reads as 27.1) links 10,555 methods, all at addresses where Il2CppDumper names the same method; its other addresses are generic instances. A build from Unity 2021.1.28f1 (also read as 27.1) gives the same result: 10,345 methods linked, every address named the same by Il2CppDumper, and its other 3,486 addresses generic instances.
+
+These checks compare addresses with symbols or with another static tool. They
+do not establish that a linked address is the code the game runs (a runtime may
+patch, hook or replace it), that builds from other Unity versions, compilers or
+obfuscators link the same way, or that a method without an address has no code:
+generic instances, adjustor thunks and invokers are not linked, and an
+`unknown` row was not looked up at all.
+
+The result adds:
+
+- `binary`: artifact identity, `format`, `architecture`, `image_base`,
+  `code_gen_modules_field` (the address of `codeGenModulesCount`, or `null` as
+  described above), `code_gen_modules_array` (the module pointer array) and
+  `code_gen_modules` (`image`, record `address`, `method_pointer_count`);
+- `method_addresses`: one row per method index with `state`, `address` and
+  `rva` (address minus `image_base`), both hexadecimal strings. `state` is
+  `code` when the module slot holds an unambiguously mapped executable address, `no-body` when an observed slot is zero,
+  as for abstract, extern and most generic methods, and `unknown` when no type
+  range in any image holds the method, a symbol relocation leaves its slot
+  unresolved, or dyld will bind a Mach-O slot. The raw word in a dyld-bound
+  slot is not treated as the resolved address.
+  `address` and `rva` are `null` unless `state` is `code`. The summary counts
+  each state in `counts.method_address_states`. Ghidra seeds are written only
+  for `code` rows.
+
+- `return_types`, `field_types` and `parameter_types`: one row per method,
+  field and parameter index with `type`, the index into the binary's
+  `Il2CppType` table, and `name`, rendered from that table: C# keywords for
+  primitives (`int`, `string`, `object`), `Namespace.Outer+Inner` for type
+  definitions, `T[]` for vectors, `T[*]` for rank-one multidimensional arrays,
+  `T[,]` for rank-two arrays and `T*`, generic instances as
+  `System.Collections.Generic.List<string>` and generic parameters by their
+  metadata name (`TKey`). Nested generic arguments stay with the component
+  that declares them, for example `Game.Outer<int>+Inner<string>` or
+  `Game.Outer<int>+Leaf` when the nested type adds no parameters. A trailing
+  `&` marks a by-reference type. `name` is `null` only for type index `-1`.
+
+`binary` also records `types_field` and `types_count`: the address of
+`Il2CppMetadataRegistration.typesCount` and its value. That record is found as
+the single place whose `fieldOffsetsCount` and `typeDefinitionsSizesCount` both
+equal the metadata type-definition count and whose type table is file mapped.
+
+Without `binary_path`, `binary` is `null` and the four binary collections are
+empty in the complete producer report to indicate unrequested observations. Their summary counts, including state counts, are `null`, and native page requests report unavailable observations. This does not establish an empty native inventory. Missing native fields in historical Evidence remain unknown too.
+
+For metadata version 27, type names are `null` and only type indices are
+reported: versions 27.0 to 27.2 share that header number but place the
+Il2CppType byref flag in different bits, so a rendered name could gain or lose
+`&` without warning.
+
+The builds checked are listed after the module search rules above.
+
+| Reason                            | Category    | Meaning                                                                                             |
+| --------------------------------- | ----------- | --------------------------------------------------------------------------------------------------- |
+| `unsupported-container`           | unsupported | Not a 64-bit little-endian Mach-O, PE or ELF image (ELF: AArch64 or x86-64 with dynamic section).   |
+| `slice-selection-required`        | unsupported | A universal Mach-O was given without `binary_architecture`.                                         |
+| `slice-not-found`                 | unsupported | The requested slice is absent, or a thin binary was given an architecture.                          |
+| `unsupported-pointer-encoding`    | unsupported | Mach-O chained fixups or Android packed (APS2) ELF relocations; pointers not decoded.               |
+| `malformed-container`             | input error | The container headers are damaged.                                                                  |
+| `code-gen-modules-not-found`      | input error | An image has no module record; often a binary from a different build.                               |
+| `ambiguous-code-gen-module`       | input error | More than one module record matches an image.                                                       |
+| `code-registration-not-found`     | input error | No unique module table or executable-byte array candidate was resolved.                             |
+| `invalid-method-token`            | input error | A token is outside its module's method-pointer rows, or a MethodDef row is reused within one image. |
+| `invalid-method-pointer`          | input error | A slot is unavailable or its nonzero address lacks unambiguous executable file coverage.            |
+| `metadata-registration-not-found` | input error | No single MetadataRegistration matches the type-definition count.                                   |
+| `invalid-type`                    | input error | A type index or `Il2CppType` record is out of range, unmapped or malformed.                         |
+| `unsupported-type-kind`           | unsupported | A valid type kind has no implemented rendering.                                                     |
+
+Format and selection failures name `binary_path`; capacity refusals remain typed resource errors. Failed linking preserves the metadata source identity, validated counts, selected binary path/architecture and its SHA-256 when observed, plus prior producer observations. It never publishes an incomplete native table as complete. The binary input is bounded to1GiB.
+
+Scans yield cooperatively in bounded batches. Name-address retention and reverse-trie construction are each limited to65,536 entries/nodes and fallback module-array candidates to64 (`native-index-capacity`); exceeding capacity does not prove absence. Rendered type text has a64MiB UTF-8 cache-workspace guard, separate from the64MiB encoded-report admission, and64 nested rendering frames (`native-type-text-capacity` / `type-nesting-capacity`). Genuine cycles are malformed; unsupported type kinds are explicit coverage failures. Native row/text admission runs before allocation or retention. Unused generic parameter names are read on demand from the same stable metadata snapshot. There is no cross-inspection file cache; validated immutable Evidence is reused for session views.
+
+ELF RELATIVE writes into zero-fill are kept as a private virtual-pointer overlay. Unresolved supported-width symbol relocations, including RELA PLT entries, remain unknown only for the bytes their ELF ABI relocation type covers; an adjacent 32-bit relocation does not make a separate 64-bit method slot unknown. Stored zeros never establish no-body. COPY, TLS descriptor, REL, packed APS2 and other unsupported relocation encodings fail closed. Input bytes are unchanged.
+
+Native address queries batch exact64-bit hexadecimal `addresses` and `states` (`code`, `no-body`, `unknown`), intersecting indices and pagination. Numerically equal address spellings match; original row indices and every shared-body alternative survive:
+
+```json
+{
+  "source": {
+    "kind": "retained-evidence",
+    "evidence_id": "<parent_evidence_id>"
+  },
+  "view": {
+    "kind": "il2cpp-metadata",
+    "collection": "method_addresses",
+    "addresses": ["0x180c56ed0"],
+    "states": ["code"],
+    "fields": ["address", "rva"],
+    "offset": 0,
+    "limit": 64
+  }
+}
+```
+
+## Ghidra function names
+
+The CLI can turn a linked report into a
+[`REA_GHIDRA_SEED_FILE`](installation.md) so Ghidra names every IL2CPP method
+before auto-analysis:
+
+```bash
+rea inspect-il2cpp-metadata global-metadata.dat --binary libil2cpp.so \
+  --ghidra-seeds il2cpp-seeds.tsv --json > /dev/null
+export REA_GHIDRA_SEED_FILE="$PWD/il2cpp-seeds.tsv"
+```
+
+Each linked method becomes one `function` seed named `Namespace.Type$$Method`
+(nested types joined with `.`, whitespace replaced by `_`), the Il2CppDumper
+convention. Seeds are at `seed-base + rva`. The default base is where Ghidra's
+loaders place the binary: `0x100000` for an ELF linked at 0, and the link base
+for Mach-O and PE; `--seed-base 0x...` overrides it for a different import. When
+identical code folding gives several methods one address, the first method names
+the function and a `# folded <address> <kept> also <names>` comment line after
+the seeds lists the others, so the kept name is visibly one of several. Ghidra still names a seeded function that is a thunk after its
+target. The file is atomically published with exclusive creation; an existing destination is never replaced. Failed staging retains its cleanup owner for retry. Source digest and link/seed-base comments make the mapping inspectable. Unknown slots are skipped separately from observed no-body slots. `--seed-base` requires `--ghidra-seeds`; overflowing rebased addresses and names/outputs beyond the existing Ghidra protocol limits (2,000 name characters,32MiB file,1,000,000 entries) are refused before publication. The CLI and reusable service use one formatter and owned output workflow; ordinary MCP inspection remains read-only.
+
+## Failures
+
+A file whose magic is not `0xFAB11BAF` fails as an unsupported target with
+`unrecognized-magic`. The file may be encrypted or may not be IL2CPP metadata.
+A version other than 24, 27, 29, 31 or 39 fails as an unsupported target with
+`unsupported-version`.
+
+Damaged metadata returns an input error (`invalid_format`) whose
+message names one reason:
+
+| Reason                    | Meaning                                                                                                          |
+| ------------------------- | ---------------------------------------------------------------------------------------------------------------- |
+| `truncated-header`        | The file is shorter than the header.                                                                             |
+| `unexpected-header-size`  | `stringLiterals.offset` is not the header size: 264 bytes for version 24, 256 for 27, 29 and 31, 380 for 39.     |
+| `section-out-of-bounds`   | A section extends beyond the file, or a non-empty section starts inside the header.                              |
+| `section-overlap`         | Two sections overlap. Zero-size sections are exempt and gaps are allowed.                                        |
+| `record-size-mismatch`    | A section size does not divide into its records, or the record size differs from the one the index widths imply. |
+| `type-index-width`        | The record used to derive the type index width implies no 1, 2 or 4 byte width.                                  |
+| `index-out-of-range`      | A serialized index or range points outside its target collection.                                                |
+| `nested-type-conflict`    | A type is listed as nested in two types, in itself, or in a cycle of declaring types.                            |
+| `declaring-type-mismatch` | A method inside a type's method range names a different type, or a method lies outside its type's range.         |
+| `image-range-overlap`     | Two images claim the same type definitions.                                                                      |
+| `invalid-string`          | A name index is outside the strings section, its string is unterminated, or it is not valid UTF-8.               |
+
+A version field that disagrees with the file's layout, for example a version
+39 file relabeled as 29, usually surfaces as `unexpected-header-size` or
+`record-size-mismatch`.
+
+An empty range's start must lie between -1 and the collection size, which is
+stricter than some runtime paths. String indices are byte offsets; a valid
+offset may select a suffix of another NUL-terminated name.
+
+An incomplete inventory is never reported as complete. Failures after a verified
+snapshot retain partial source Evidence with its SHA-256, byte length and any
+validated section row counts; these counts do not establish that names and
+relationships were decoded. A file that changes during reading and a cancelled
+request keep distinct errors. Decode checkpoints yield for cancellation, and
+close waits for active work. Each reader admits one inspection at a time.
+Resource guards bound the selected file to 256 MiB and the encoded complete
+report to 64 MiB, checking row counts before allocation and JSON/UTF-8/base64 expansion during
+decoding, before serialization. Names and identical literal spans are decoded
+once within one stable file snapshot; no filesystem results are cached across
+inspections. Ownership assignment also yields to cancellation. MCP also applies its
+[response transport budget](mcp-contracts.md); a complete report within the
+report limit can still require the summary and paging path.
+
+## Outside this tool
+
+- Decrypting or deobfuscating protected metadata.
+- Linking Mach-O binaries that use chained fixups, ELF images with Android
+  packed (APS2) relocations, and 32-bit images; generic method instances,
+  adjustor thunks and invokers.
+- Full method signatures beyond the rendered field, parameter and return type
+  names, generic method instances, and type names for metadata version 27.
+- Running Ghidra or another disassembler. `--ghidra-seeds` only writes a seed
+  file for [`REA_GHIDRA_SEED_FILE`](installation.md); use the
+  [native workflow](native-investigation.md) for native code.
+- Metadata versions other than 24, 27, 29, 31 and 39.
+
+Names come from the metadata file only. They do not establish which methods
+were compiled into the native binary or how they behave at runtime.
+
+## Real producer verification
+
+`npm run verify:il2cpp:metadata -- --metadata PATH` exercises a caller-supplied
+`global-metadata.dat` through the public CLI and MCP. It is opt-in, like
+`verify:pcap`, because no real game metadata ships with the repository. Normal
+inspection does not need it.
+
+The follow-up was checked against the public Cpp2IL
+[`Simple_2022_3_35` sample](https://github.com/SamboyCoding/Cpp2IL/tree/b5ad444b82267cb1e4b88b8b373c008105bdea52/TestFiles/Simple_2022_3_35):
+version 31, 1,892,268 bytes, SHA-256
+`2441635a7c234a1a34deccc6a9906d6f5abcbdab078a38b2bf488a52a4962c48`.
+An independent byte-layout oracle matched all 36 images, 2,468 types and 12,409
+methods. Corresponding managed DLLs independently matched names and ownership
+for 35 images, 2,463 types and 12,406 methods; the generated image has no matching
+DLL. This does not establish real-producer coverage for versions 29 or 39.
+
+Compact native summaries omit the `binary.code_gen_modules` table and keep its
+count in `counts.code_gen_modules`. Query `collection: "code_gen_modules"`
+with batch `image_indices` or source `indices`, selected `fields` and pagination.
+Module row `index` is its retained collection position; `image` joins the
+metadata image collection. Scalar registration facts and both source identities
+remain inline. Complete Evidence retains the original full table.

docs/index.md

@@ -47,20 +47,23 @@ For runtime instrumentation, see the [Frida guide](./frida-instrumentation.md).
 
 ## Choose a guide
 
-| Task                                                | Guides                                                                                                                                                                                                                                                                                                                      |
-| --------------------------------------------------- | --------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- |
-| Inspect native functions, references, or types      | [Native binaries](native-investigation.md), [Go metadata](go-binaries.md), [Rust dependencies](rust-binaries.md), [offline ELF layout](binary-diagnostics.md), [recorded crashes](recorded-crashes.md)                                                                                                                      |
-| Select a native provider or import format           | [Ghidra semantics](ghidra-provider.md), [Windows](windows-ghidra-p0.md), [DOS](ghidra-dos.md), [raw bytes](raw-binary.md), [MIPS](ghidra-mips.md), [PSP](ghidra-psp.md), [NativeAOT](ghidra-nativeaot.md), [IDA](ida-provider.md), [Kuna](kuna-provider.md), [JEB](jeb-analysis.md)                                         |
-| Analyze JavaScript or Electron artifacts            | [Static analysis](javascript-artifact-reconstruction.md), [tracing and comparisons](javascript-application-workflows.md), [application graph](javascript-application-graph.md), [source recovery](javascript-recovery.md)                                                                                                   |
-| Compare static JavaScript with runtime observations | [Runtime attribution](javascript-runtime-observation.md), [reconciliation](javascript-runtime-reconciliation.md), [Electron observation](electron-observation.md), [Electron fuses](electron-fuses.md)                                                                                                                      |
-| Inspect assemblies, packages, or firmware           | [.NET](managed-code-analysis.md), [Android](android-analysis.md), [Apktool resources](apktool-resource-analysis.md), [ADB acquisition](adb-device-analysis.md), [Apple](apple-application-analysis.md), [Harmony](harmony-analysis.md), [firmware](firmware-analysis.md), [WASM](wasm-artifacts.md), [EVM](evm-bytecode.md) |
-| Observe or compare a browser scenario               | [CDP observation](browser-observation.md), [scenarios](browser-scenario-contract.md), [runtime evidence](web-runtime.md), [historical captures](web-network-captures.md)                                                                                                                                                    |
-| Trace captured website source                       | [Source locations](web-source-location.md), [module traces](website-module-trace.md), [script export](website-script-export.md)                                                                                                                                                                                             |
-| Capture and compare a command                       | [Process capture](process-capture.md)                                                                                                                                                                                                                                                                                       |
-| Use CLI/MCP and retain evidence                     | [CLI](cli.md), [MCP contracts](mcp-contracts.md), [agent prompts](mcp-prompts.md), [obligation ledgers](reconstruction-obligation-ledgers.md)                                                                                                                                                                               |
+| Task                                                | Guides                                                                                                                                                                                                                                                                                                                                                                                                                                         |
+| --------------------------------------------------- | ---------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- |
+| Identify a desktop installation                     | [Runtime candidates, evidence and follow-up files](desktop-applications.md)                                                                                                                                                                                                                                                                                                                                                                    |
+| Inspect native functions, references, or types      | [Native binaries](native-investigation.md), [Go metadata](go-binaries.md), [Rust dependencies](rust-binaries.md), [CPython pyc files](python-pyc.md), [PE certificates](pe-certificates.md), [ELF notes](elf-notes.md), [Erlang BEAM](erlang-beam.md), [offline ELF layout](binary-diagnostics.md), [recorded crashes](recorded-crashes.md)                                                                                                    |
+| Select a native provider or import format           | [Ghidra semantics](ghidra-provider.md), [Windows](windows-ghidra-p0.md), [DOS](ghidra-dos.md), [raw bytes](raw-binary.md), [MIPS](ghidra-mips.md), [PSP](ghidra-psp.md), [NativeAOT](ghidra-nativeaot.md), [IDA](ida-provider.md), [Kuna](kuna-provider.md), [JEB](jeb-analysis.md)                                                                                                                                                            |
+| Analyze JavaScript or Electron artifacts            | [Static analysis](javascript-artifact-reconstruction.md), [tracing and comparisons](javascript-application-workflows.md), [application graph](javascript-application-graph.md), [source recovery](javascript-recovery.md)                                                                                                                                                                                                                      |
+| Compare static JavaScript with runtime observations | [Runtime attribution](javascript-runtime-observation.md), [reconciliation](javascript-runtime-reconciliation.md), [Electron observation](electron-observation.md), [Electron fuses](electron-fuses.md)                                                                                                                                                                                                                                         |
+| Inspect assemblies, packages, or firmware           | [Godot PCK](godot-pck.md), [.NET](managed-code-analysis.md), [Python packages](python-packaged-applications.md), [Android](android-analysis.md), [APK signatures](apk-signature.md), [Apktool resources](apktool-resource-analysis.md), [ADB acquisition](adb-device-analysis.md), [Apple](apple-application-analysis.md), [Harmony](harmony-analysis.md), [firmware](firmware-analysis.md), [WASM](wasm-artifacts.md), [EVM](evm-bytecode.md) |
+| Observe or compare a browser scenario               | [CDP observation](browser-observation.md), [scenarios](browser-scenario-contract.md), [runtime evidence](web-runtime.md), [historical captures](web-network-captures.md)                                                                                                                                                                                                                                                                       |
+| Serve recorded behavior as an agent environment     | [Agent environments](agent-environments.md)                                                                                                                                                                                                                                                                                                                                                                                                    |
+| Trace captured website source                       | [Source locations](web-source-location.md), [module traces](website-module-trace.md), [script export](website-script-export.md)                                                                                                                                                                                                                                                                                                                |
+| Capture and compare a command                       | [Process capture](process-capture.md)                                                                                                                                                                                                                                                                                                                                                                                                          |
+| Use CLI/MCP and retain evidence                     | [CLI](cli.md), [MCP contracts](mcp-contracts.md), [agent prompts](mcp-prompts.md), [obligation ledgers](reconstruction-obligation-ledgers.md)                                                                                                                                                                                                                                                                                                  |
 
 For development, use [testing](testing.md), [client verification](client-verification.md),
-[tool design](tool-design.md), and [releasing](releasing.md).
+[tool design](tool-design.md), [admission limits and recovery](design/limits-and-recovery.md),
+and [releasing](releasing.md).
 [Reconstruction readiness](reconstruction-readiness.md) describes a development
 conformance verifier. [The 6.2 migration](migration-6.2.md) records that release's
 compatibility changes; [the roadmap](roadmap.md) tracks future work.

docs/installation.md

@@ -24,6 +24,10 @@ Restart the agents named in the completion message. Quit Claude Desktop fully
 before approving, because a running app can overwrite the registration; on
 Windows, see [where Claude Desktop reads its configuration](client-configuration.md#claude-desktop).
 
+During package resolution, npm may print a deprecation warning for `prebuild-install`
+from transitive native dependencies (such as Frida). This warning originates from
+the upstream dependency tree and does not affect REA or indicate setup failure.
+
 For the latest published release, use `npx rea-agents@latest setup`.
 [Package-runner and global-installation options](#package-runners-and-global-installation)
 cover explicit versions, Bun, runtime requirements, and the curl wrapper.
@@ -56,17 +60,39 @@ Before applying changes, REA checks your current configuration. The plan lists:
 
 Malformed or unsafe existing configuration blocks the whole transaction before
 Hopper installation or any file write. Declining or pressing Ctrl-C makes no
-changes. Agent configuration writes preserve unrelated entries and comments,
-create backups, use atomic replacement, and verify their result. Setup and
-uninstall retain an existing `.rea.backup` rather than replacing the first snapshot.
-If that snapshot differs from the current configuration, setup previews and
-preserves a verified `.rea.backup.<sha256>.recovery` copy of the current bytes.
-Both snapshots remain available; a failed rollback reports the current recovery
-copy's path and the restore failure.
+changes. Client configuration files and bundled skill files have separate
+publication and recovery paths; the guarantees in this paragraph apply to
+client configuration. See [skill destinations](#skill-destinations) for skill
+rollback limits. Agent configuration writes preserve unrelated entries and comments,
+create backups, stage changes in a private directory, and verify their result.
+Before publication, they refuse detected changes to the selected configuration.
+Replacement uses an explicit destination under the selected canonical parent;
+it does not follow a newly inserted destination symlink. First-run setup uses
+an exclusive hard link and fails without overwriting a newly created destination;
+filesystems without hard-link support cannot publish a first-run configuration.
+Setup and uninstall retain an existing `.rea.backup` rather than replacing the
+first snapshot.
+If that snapshot differs from the current configuration, setup and uninstall
+preserve a verified `.rea.backup.<sha256>.recovery` copy of the current bytes
+before editing; setup also previews it. Both snapshots remain available.
+When an edit made a backup, a failed uninstall update or restore, and a failed
+setup readback or rollback, name the backup path that holds the bytes from just
+before that edit.
+If setup cannot verify a newly created configuration, it retains the file and
+reports manual recovery instead of deleting a pathname that may have changed.
+For an existing configuration, setup and uninstall recovery refuse a detected
+replacement or edit and name the recovery backup. Quit clients before editing their configuration;
+these checks do not lock out another writer modifying the same file in place,
+or exclude pathname changes between the final check and publication.
 Setup also retains valid environment overrides in REA's existing server entry,
 such as `REA_MCP_INPUT_SCHEMA_PROFILE`, `REA_MCP_OUTPUT_SCHEMA_PROFILE`, and
 `REA_MCP_MAX_RESPONSE_BYTES`.
 Detected provider settings replace only matching environment keys.
+Setup manages REA's launcher, selected provider settings, and supported timeout
+defaults. Client approval policies remain user choices, but the current Codex
+refresh drops those policies inside the REA entry; see the
+[Codex preservation limitation](client-configuration.md#codex-and-devin) before
+refreshing a customized registration.
 Setup and doctor report a malformed REA environment map or invalid runtime
 setting before setup writes files. Repair the named setting in the client
 configuration; setup does not silently discard the rest of the map.
@@ -86,9 +112,11 @@ your investigation. Failed steps and diagnostics remain in terminal history.
 
 Select exact clients in scripts with repeatable `--client` flags. Each explicit
 client skips interactive selection. Use `--all-detected` only when you intend
-to configure every detected supported client. `--skill=false` omits the
-workflow, and `--dry-run` returns a read-only plan with status `planned` and
-exit code 0:
+to configure every detected supported client. Combining `--client` with
+`--all-detected` configures both the explicitly selected and detected clients,
+even when an explicit client is not detected. Each client is configured once.
+`--skill=false` omits the workflow, and `--dry-run` returns a read-only plan
+with status `planned` and exit code 0:
 
 ```bash
 rea setup --client codex --client cursor --skill=false --dry-run
@@ -115,6 +143,15 @@ exits 1. Installing missing Hopper non-interactively additionally requires
 rea setup --yes --all-detected --install-hopper --json
 ```
 
+Client configuration writes detect changes to the file read by setup, including
+when a matching recovery backup already exists. A detected conflict reports
+`reason: "conflict"` and preserves the newer file; review it and rerun setup.
+Rollback likewise requires the file identity and content installed by that
+transaction. Requests for the same resolved path are serialized within one REA
+process. These checks do not provide an operating-system compare-and-swap or
+coordinate arbitrary external writers, so avoid editing client configuration
+while setup is applying it.
+
 ### Skill destinations
 
 The bundled skill is installed in each selected client's personal skill directory.
@@ -133,6 +170,26 @@ paths as `unmanaged`, names the affected paths, and provides those alternatives;
 an unreadable skill identity remains `unknown`. Uninstall retains and reports
 files it cannot safely remove, including their ownership manifest.
 
+The skill installer preserves byte-exact backups of changed existing files,
+including invalid UTF-8, BOMs, and permission modes, and verifies each published
+file through its owned descriptor. Before publication it checks the selected
+file identity; first installation uses an exclusive hard link and refuses a
+concurrently created destination. Filesystems without hard-link support cannot
+publish a first-install skill file.
+
+Rollback restores an existing file only while its published identity, state,
+and bytes still match that installation attempt. A recovery backup preserves
+the original bytes; it does not authorize overwriting a later user edit or
+replacement. Conflicts report the affected paths and recovery backups. If a
+newly created skill file cannot be verified, setup retains it and reports manual
+recovery rather than deleting by pathname. Staging cleanup failures remain
+owned and retryable, and setup does not report installation success while
+cleanup is unresolved.
+
+These checks do not lock out another writer modifying the same inode or exclude
+pathname changes between the last check and publication. Avoid editing managed
+skill files or running another installer while setup applies them.
+
 `rea setup --skill-client claude_code --yes --json` refreshes and verifies the
 selected skill destination independently of MCP registration selection. The
 final `doctor.scope.skill_clients` records which skill destinations were
@@ -179,6 +236,8 @@ as well, so it is broader than this Claude-only migration.
 history under `CLAUDE_CONFIG_DIR`. Setup follows the same
 variable: it writes the registration to `$CLAUDE_CONFIG_DIR/.claude.json` and
 the skill to `$CLAUDE_CONFIG_DIR/skills`.
+For REA setup, an empty `CLAUDE_CONFIG_DIR` is treated as unset: registration
+goes to `~/.claude.json` and the skill goes to `~/.claude/skills`.
 
 ```bash
 CLAUDE_CONFIG_DIR=~/.claude-rea npx rea-agents setup --client claude_code
@@ -260,14 +319,14 @@ Register with the variable unset:
 <!-- x-release-please-start-version -->
 
 ```bash
-claude mcp add --scope local rea -- npx -y rea-agents@6.5.0 mcp
+claude mcp add --scope local rea -- npx -y rea-agents@6.7.0 mcp
 ```
 
 On native Windows, use the same project directory and wrap the package-runner
 shim so the stdio client can launch it:
 
 ```powershell
-claude mcp add --scope local rea -- cmd.exe /d /c npx -y rea-agents@6.5.0 mcp
+claude mcp add --scope local rea -- cmd.exe /d /c npx -y rea-agents@6.7.0 mcp
 ```
 
 <!-- x-release-please-end -->
@@ -324,6 +383,13 @@ A scoped report has `scope.mode: "explicit"`. Its `scope_checks` determine
 `environment_healthy` summarizes the full audit. `--target` adds a target check,
 but the report remains audit-wide unless a scope option is also supplied.
 
+An invalid `REA_LOG_LEVEL` fails every scope, including one client or the
+installed skill, because the executable rejects it before any command. A
+selected client also fails for every other setting that MCP startup rejects,
+including `REA_GHIDRA_STARTUP_TIMEOUT_MS` and `REA_HOPPER_STARTUP_TIMEOUT_MS`.
+Those settings still fail a selected provider and the audit-wide result. A
+skill-only report leaves them informational.
+
 Follow the remediation for the failed check. If a native target has several
 available providers, choose one with `--provider` on the CLI or `provider_id`
 on `open_binary`. A `capability_unavailable` failure carries
@@ -360,11 +426,68 @@ Uninstall preserves Hopper, Node.js, Evidence files, captures, unrelated skills
 and other MCP servers. Purging removes only REA's cache and state under
 `~/.rea`. A client configuration that is malformed, unreadable, or at an unsafe
 path stops the operation before anything is removed, and a client that fails
-while being updated stops the remaining removals. A purge path that is a
+while being updated stops the remaining removals. Uninstall reports each
+candidate `configuration_path` with its outcome. A client with multiple files
+keeps their individual `actions` alongside its aggregate status and detail,
+including earlier removals when a later file fails. The paths name
+the discovered configuration candidates, not resolved physical file identities.
+For JSON clients, uninstall removes every repeated `rea` member in the effective
+server table only when each registration is REA-owned. An unowned earlier
+member, or an unowned OpenCode V1 entry that removing V2 would expose, stops
+preflight before any
+removal. Repeats inside an inactive earlier copy of the server table stay
+unchanged. A purge path that is a
 symbolic link is retained and reported rather than followed. Uninstall checks
 REA's managed shared and Claude Code personal skill locations. See the
 [CLI guide](cli.md#output-and-exit-status) for exit statuses.
 
+### Missing dependencies in an npx installation
+
+If REA exits with `ERR_MODULE_NOT_FOUND` before `doctor` can produce a report,
+inspect the missing path in the error. A path under npm's `_npx/<key>/node_modules`
+identifies a package-runner installation. In [#5259](https://github.com/morluto/rea/issues/5259),
+the missing file was `@modelcontextprotocol/client/dist/index.mjs`. That file is
+present in a clean Windows installation of REA 6.7.0; its absence alone does not
+establish an incompatible dependency release or explain how the file went missing.
+
+`npm cache verify` checks npm's downloaded content cache; it does not restore
+missing files in an already extracted npx installation. Repeating the same npx
+command can reuse that incomplete installation. To reinstall only the affected
+entry:
+
+1. Stop the REA processes and agent clients using that entry. In the same shell
+   environment as the failing command, identify the cache and its entries:
+
+   ```powershell
+   npm.cmd config get cache
+   npm.cmd cache npx ls
+   npm.cmd cache npx info REA_CACHE_KEY
+   ```
+
+   Replace `REA_CACHE_KEY` with the key from the failing `_npx` path. Check that
+   `info` reports that location and the expected `rea-agents` version. Its
+   `valid` label describes the cache entry metadata, not the presence of every
+   installed file. These subcommands require npm with `cache npx` support.
+
+2. Remove that specific entry and rerun the affected version with registry
+   access. For the version reported in #5259:
+
+   ```powershell
+   npm.cmd cache npx rm REA_CACHE_KEY
+   npx.cmd --yes rea-agents@6.7.0 doctor --json
+   ```
+
+   Keep the key argument: this recovery does not require clearing all npm caches
+   or deleting REA configuration, skills, or Evidence. On Linux/macOS, use `npm`
+   and `npx` without `.cmd`. A JSON doctor report confirms startup recovered;
+   it can still exit nonzero for missing analysis engines or registrations.
+   Restart the affected agent after repairing its cache entry.
+
+If a fresh installation still loses the file, retain the error, REA/Node/npm
+versions and installation details in the issue. Check installation logs and,
+on Windows, security software's protection history before attributing the cause.
+Do not rename `index.cjs` to `index.mjs` or disable protection to make startup pass.
+
 ### Update the owning installation
 
 `rea update` installs the exact resolved
@@ -503,8 +626,10 @@ Hopper is separate commercial software with its own license. Its free demo has
 vendor-defined limits, and a paid license is optional. REA reuses any detected
 installation and preserves Hopper during uninstall.
 
-The automated Hopper installer supports macOS 12+, Ubuntu 24.04+, Fedora 41+,
-Nobara 44+, 64-bit Arch Linux, CachyOS, or Omarchy. REA itself runs on any
+The automated Hopper installer supports macOS 12+ and x86_64 hosts running
+Ubuntu 24.04+, Fedora 41+, Nobara 44+, Arch Linux, CachyOS, or Omarchy. The
+official Linux packages are x86_64-only, so setup does not offer Hopper on other
+Linux architectures; configure [Ghidra](#ghidra) there instead. REA itself runs on any
 Linux distribution with a recognized apt, dnf, or pacman package family, where
 an existing Hopper launcher is selected through `HOPPER_LAUNCHER_PATH`. Ghidra
 and IDA have their own provider-specific host requirements; Windows Ghidra uses
@@ -583,12 +708,28 @@ host mount and the rest of `/tmp` remain unchanged; this fallback never invokes
 `sudo`. `rea doctor --provider hopper --json` reports the selected
 strategy and both host and effective mount facts.
 
+Hopper also embeds Python; the Python executable and standard library alone
+are insufficient on Ubuntu. Approved setup checks the system interpreter
+at `/usr/bin/python3` and, when necessary, installs its matching
+`libpython3.X` package, then verifies that its shared library can load.
+APT resolves the concrete package (for example, `libpython3.12t64` on Ubuntu
+24.04). Doctor includes these fresh observations in the `embedded_python`
+details of its `hopper-demo-runtime` check, alongside the private-display
+diagnostic. Failed inspections remain unknown rather than proving a library
+absent. Run `rea setup --yes --install-hopper` to repair a supported install.
+This prerequisite check does not launch the bridge or validate a custom Python
+interpreter configured inside Hopper; that interpreter needs its own matching
+shared runtime. Custom installation prefixes or shared-library ABIs require
+manual dependency repair.
+
 When the package-manager step fails, setup reports the failing command and its
 captured output, so a package conflict is distinguishable from an authorization
-failure. On CachyOS, an unversioned compatibility package such as
-`zlib-ng-compat` may conflict with Hopper's versioned `zlib` dependency. Resolve
-the package conflict yourself or install Hopper's official package manually,
-then rerun `rea setup`.
+failure. Hopper is optional, so setup still applies the approved client
+registrations and skill, then reports `needs_human` with the Hopper failure. On
+Arch Linux and CachyOS, the unversioned compatibility package `zlib-ng-compat`
+may conflict with Hopper's versioned `zlib` dependency. Resolve the package
+conflict yourself, install Hopper's official package manually, or use another
+provider such as Ghidra; rerun `rea setup` after installing Hopper.
 
 The supported Linux demo forwards additional launches into its existing
 application, even across private displays. REA therefore reserves one Linux
@@ -623,7 +764,9 @@ and skips directories.
 On Linux, it prefers executable `/opt/hopper/bin/Hopper`, then executable
 `~/.local/share/rea/hopper/bin/Hopper`, followed by `/opt/hopper-v4/bin/hopper`
 and `/usr/bin/hopper`; if none exists, the former remains the diagnostic
-fallback. `HOPPER_LAUNCHER_PATH` overrides those choices:
+fallback. `rea doctor` checks only the current platform's candidates, in the
+same order; other platforms have no automatic candidates.
+`HOPPER_LAUNCHER_PATH` overrides those choices:
 
 ```bash
 export HOPPER_LAUNCHER_PATH=/absolute/path/to/Hopper
@@ -635,6 +778,11 @@ bridge readiness. For large targets, set `REA_HOPPER_STARTUP_TIMEOUT_MS` to a
 decimal integer from 1 to 2,147,483,647 milliseconds in the environment that
 starts REA. CLI and MCP use the same setting; invalid values fail configuration
 validation. An MCP client's own request timeout must also allow enough time.
+`open_binary` validates and selects a target without waiting for native provider
+readiness. During Hopper startup, `binary_session.provider_operation_health`
+reports `busy` at stage `launch`; `analysis_activity` describes bridge requests
+and can remain idle until the socket is ready. Cleanup work reports `busy` at
+stage `connection`.
 
 If macOS denies Apple Events, allow the application that starts REA under
 **System Settings → Privacy & Security → Automation**, then retry. REA releases
@@ -644,9 +792,12 @@ Arguments containing quotes, backslashes, or line breaks remain unchanged but
 cannot establish that a failed handoff opened no document. Other failed handoffs
 retain ownership until document closure is confirmed. If a timed-out bridge
 becomes ready later, `close_binary` connects to that run, authenticates its
-identity, and requests document shutdown. After a connection drops, the bridge
-allows 30 seconds for an authenticated cleanup reconnect, then releases Hopper's
-Python execution thread. Expiry does not establish document closure, so REA
+identity, and requests document shutdown. Once its listener is ready, the bridge
+allows 30 seconds for the first connection and capability validation. After a
+connection drops, it allows the same grace period for an authenticated cleanup
+reconnect, then releases Hopper's Python execution thread. Initial analysis is
+outside this listener grace period, and authenticated native work is not bounded
+by it. Expiry does not establish document closure, so REA
 retains unconfirmed ownership. Retry close while the bridge remains available.
 
 If a Linux launcher exists but cannot start, inspect its shared libraries:
@@ -736,9 +887,17 @@ Once document and process closure are confirmed, the lease can be released even
 if temporary-file cleanup fails. Another client can then launch, but the owning
 client must finish its retained cleanup before starting again.
 An unconfirmed external document can be retried while its authenticated bridge
-remains connected. If that bridge has disconnected, another close cannot confirm
-the document: inspect and close the reported document in Hopper before ending
-the owning REA connection.
+remains available. The same authentication and process-identity capture apply
+when the first successful handshake occurs during cleanup after a startup
+timeout. REA can confirm closure if that exact authenticated process exits,
+including while a shutdown request is in flight. A helper exit alone is
+insufficient, and an application that never authenticated has no verified
+process identity. If no matching bridge becomes available, inspect and close
+the reported document in Hopper before ending the owning REA connection.
+Export the Evidence bundle before `close_binary` clears session records.
+`binary_session.cleanup_pending` remains true while provider cleanup is retained,
+and provider health remains visible even with `open: false`; that flag alone
+does not establish successful cleanup.
 
 ### Hopper in CI
 
@@ -797,7 +956,7 @@ and `set_addresses_names` name a function entry or any other mapped address
 session metadata edits leave executable bytes unchanged and are discarded on close. GUI controls
 require Hopper; Windows P0 remains read-only.
 
-Windows P0 admits native x86 and x86-64 PE applications and DLLs, plus explicitly
+Windows P0 admits native x86, x86-64, ARMNT (Thumb-2) and ARM64 PE applications and DLLs, plus explicitly
 selected [raw NMOS 6502 images](raw-binary.md#nmos-6502) with a caller-supplied
 entry at file offset 0, on fixed local NTFS volumes. Other raw CPU families remain
 unsupported on Windows.
@@ -815,6 +974,16 @@ rea doctor --json
 rea setup
 ```
 
+On macOS, [Homebrew OpenJDK is keg-only](https://formulae.brew.sh/formula/openjdk)
+and may be unavailable to the system `java` wrapper. Set `JAVA_HOME` to the
+selected formula's `libexec/openjdk.jdk/Contents/Home`, for example
+`/opt/homebrew/opt/openjdk/libexec/openjdk.jdk/Contents/Home` or
+`/usr/local/opt/openjdk/libexec/openjdk.jdk/Contents/Home`. Use a JDK inside the
+Java version bounds declared by your Ghidra installation; REA does not select
+another installed JDK automatically. Doctor distinguishes an absent Java
+command from a command that runs and fails, and retains bounded probe output
+and any parsed properties when the selected runtime cannot be admitted.
+
 On Windows, configure the existing installation in PowerShell:
 
 ```powershell
@@ -846,12 +1015,50 @@ It also checks that the reported JVM `os.arch` and `os.name` map to the same nat
 directory validated for the host; unrecognized reported architectures or a
 non-host operating system are refused. LaunchSupport overrides of either
 property, including through helper-environment JVM option variables, are
-rejected at launch.
+rejected at launch. `JDK_JAVA_OPTIONS` argument files (`@file`) are expanded
+and inspected with that same check using the Java launcher's argfile quoting,
+comments and continuation rules. The reader admits regular UTF-8 files, retains
+failed descriptor cleanup, and bounds all admitted file contents and the prepared
+environment to 64 KiB per launch. Unreadable, oversized or nested `@` references
+are refused before the analysis JVM starts. Validated bytes are snapshotted under
+the owned runtime root, so later helpers and the analysis JVM do not reopen mutable
+caller files. Repeated references share that launch's snapshot; no live state is
+cached across launches. `JAVA_TOOL_OPTIONS`
+and `_JAVA_OPTIONS` do not expand `@file` tokens, so they remain
+literal-token checks.
 When Java is found through `PATH`, setup records its observed JDK home so GUI
 MCP clients do not depend on an incidental shell path. Setup shows every exact
 environment entry in its plan, writes only after approval, and never downloads,
 installs, upgrades, or modifies Ghidra or Java.
 
+Installation inspection derives the private configuration directory identity from
+`application.version` and `application.release.name` in
+`Ghidra/application.properties`. The directory's version spelling follows Ghidra
+(for example, `12.1.0` becomes
+`12.1`), while the reported provider version preserves the original value.
+The launcher uses that inspected coordinate; it does not reread the properties
+or guess from the extracted folder's name.
+If the identity is missing, malformed, or unsafe as a single path component,
+the launcher fails before preparing its home/cache/config/temp directories or
+starting Java. The client may already have allocated private runtime/endpoint
+resources and a target snapshot; existing startup-failure cleanup still applies.
+Restore the selected release's properties instead of renaming its installation folder.
+Inspection follows selected properties symlinks, admits only a regular file,
+and caps the metadata payload at 1 MiB under REA's resource budget. A final
+one-byte probe checks for growth. This is not a Java
+Properties format limit. Reads preserve Ghidra's Latin-1 byte mapping and check
+the descriptor and selected name for changes. Permission failures and resource
+refusals remain explicit; failed descriptor cleanup must succeed before that
+owner admits another inspection.
+Native decompiler compatibility inspection follows the selected executable symlink
+and reads at most the first 4096 bytes through the same descriptor owner. It
+checks the full extent and selected name for changes without reading the remainder.
+An unreadable or unrecognized header leaves compatibility unknown; diagnostics
+retain read failures. Cancellation and incomplete cleanup propagate as failures,
+and failed cleanup retains the exact descriptor for retry.
+This coordinate selects only REA's session-private configuration directory, not
+the installation or an existing user configuration directory.
+
 Each verified session uses an ephemeral temporary project and isolated
 home/cache/config/temp paths. REA passes `-readOnly`, `-deleteProject`, uses
 Ghidra's default analysis settings, and loads its packaged Java
@@ -867,6 +1074,10 @@ ownership inspection, so retaining those wrappers would prevent verified
 process-group cancellation during startup.
 If ownership remains unverifiable, it reports the reason and retains the process
 supervisor and private runtime instead of removing files beneath a live provider.
+Startup preserves an observed launch failure when cancellation arrives during
+rollback. Producer cancellation retains the caller's exact reason separately
+from LaunchSupport warnings and cleanup failures. Failed rollback keeps the
+original process owner for a later cleanup retry.
 The experimental Windows transport uses authenticated IPv4 loopback with a
 private native-owned bearer descriptor and Job Object process ownership.
 
@@ -878,6 +1089,9 @@ bridge, and health readiness. Large binaries can need more: set
 `REA_GHIDRA_STARTUP_TIMEOUT_MS` to an integer between 1 and 2,147,483,647
 milliseconds in the server environment. An absent setting uses the default;
 empty, invalid or out-of-range supplied values fail configuration validation.
+When the deadline elapses, the `provider_timeout` error names
+`REA_GHIDRA_STARTUP_TIMEOUT_MS` with the deadline that was applied and suggests
+`REA_GHIDRA_DISABLED_ANALYZERS` for analyzers that dominate the time.
 REA parses the supplied configuration environment and
 passes the deadline to each provider client; a startup failure is returned by the query that triggered it,
 without exposing partial analysis. This deadline is separate from MCP transport
@@ -886,6 +1100,24 @@ can time out earlier even when Ghidra would complete within its startup deadline
 See [Ghidra first-query deadlines and recovery](mcp-contracts.md#ghidra-first-query-deadlines-and-recovery)
 for client options and the close/reopen recovery flow.
 
+The 330-second default is configurable. Configuration is read when REA starts:
+changing an environment variable in another shell, or closing and reopening a
+binary in the same MCP server, does not change that server's startup allowance.
+For a CLI retry with an example 30-minute allowance:
+
+```sh
+REA_GHIDRA_STARTUP_TIMEOUT_MS=1800000 rea analyze /absolute/path/to/app.exe --provider ghidra --json
+```
+
+For MCP, export useful Evidence before restarting, close the binary and resolve
+any cleanup failure, update the environment in the client's REA server launch
+configuration, then restart REA and reconnect before reopening the target. Allow
+startup plus the query in the client's own tool-call deadline. Every retry uses
+a fresh ephemeral import; smaller result requests do not reduce the prerequisite
+auto-analysis. A larger time allowance does not raise the JVM heap limit or
+guarantee completion. Keep [heap and host CPU controls](#ghidra-heap-and-cpu-controls)
+appropriate for the machine.
+
 One session contains exactly one imported Program; use `provider_id: "ghidra"`, `--provider ghidra`, or
 `REA_ANALYSIS_PROVIDER=ghidra` when both Hopper and Ghidra support the target.
 One persistent decompiler is owned by the Program, and a serial queue keeps
@@ -959,6 +1191,8 @@ export REA_GHIDRA_COMPILER_SPEC_ID="borlandcpp"   # optional; requires a languag
 
 REA passes `-processor`/`-cspec`, checks both against the installation's `.ldefs`, and records them in the analysis profile. DOS imports reject an override.
 
+Language definitions must be well-formed XML encoded as UTF-8 or BOM-marked UTF-16, with any encoding declaration matching the bytes. Literal Unicode replacement characters (U+FFFD) are accepted; malformed source bytes are rejected.
+
 To seed code that auto-analysis cannot reach (computed jumps, pointers in data):
 
 ```bash
@@ -983,12 +1217,22 @@ override in `analyzer_preset` (for example
 open when its committed profile and the setting differ. A name starts with a
 letter or digit and contains only letters, digits, spaces, periods, underscores,
 and hyphens; at most 32 names are admitted, each at most 120 characters.
-Names must identify analyzers registered for the selected program. Unknown names
-or options that remain enabled abort the headless import before auto-analysis
-and the bridge, so an unsuccessful override cannot produce a session claiming
-that the analyzer was disabled. For Windows PE resource analysis, the exact name
-is `WindowsResourceReference`. Disabling an analyzer skips the facts it would
-have produced.
+
+The setting applies to every import in the REA server, but Ghidra registers
+format-specific analyzers, such as `Objective-C Message Analyzer` or
+`Demangler Swift`, only for programs they can analyze. A name that the Ghidra
+installation provides but did not register for the selected program is not
+applicable: the import continues, the profile still commits the configured
+names, and result limitations report the applied preset with that name marked,
+for example `ghidra-default+disabled:Objective-C Message Analyzer(n/a)`. A name
+that no installed analyzer provides, or an option that remains enabled, aborts
+the headless import before auto-analysis and the bridge with an error naming
+`REA_GHIDRA_DISABLED_ANALYZERS` and the analyzer, so a misspelled override
+cannot produce a session claiming that the analyzer was disabled. Changing the
+environment requires restarting the server; reopening a target alone keeps the
+current setting. For Windows PE resource analysis, the exact name is
+`WindowsResourceReference`. Disabling an analyzer skips the facts it would have
+produced.
 
 ### Ghidra heap and CPU controls
 
@@ -1056,8 +1300,14 @@ truncation markers. Darwin ownership-helper compilation failures also retain
 the compiler outcome, exit code or signal, and bounded stderr; a timed-out
 compiler remains distinguishable from a compiler that exited with an error.
 The helper compile deadline defaults to 120,000 ms; set
-`REA_SWIFTC_COMPILE_TIMEOUT_MS` to an integer between 1 and 2,147,483,647 when
-a busy host needs longer.
+`REA_SWIFTC_COMPILE_TIMEOUT_MS` to an integer between 1 and 2,147,483,647 in
+the environment that starts REA when a busy host needs longer. Restart REA to
+apply a change; the next ownership-helper compile uses the new value. A
+timeout names that effective deadline.
+On macOS, the Java prerequisite probe waits for this independently bounded
+preparation before starting its five-second command deadline. Caller cancellation
+still interrupts preparation, and failed helper cleanup remains owned for retry.
+Linux process-table preparation remains inside the command deadline.
 Other rejection values retain their primitive value or an explicit type.
 Shutdown warnings include the failure
 kind, message, and diagnostics, while process and temporary-project cleanup
@@ -1104,7 +1354,7 @@ For a client that requires manual configuration, use:
   "mcpServers": {
     "rea": {
       "command": "npx",
-      "args": ["-y", "rea-agents@6.5.0", "mcp"]
+      "args": ["-y", "rea-agents@6.7.0", "mcp"]
     }
   }
 }
@@ -1149,7 +1399,7 @@ equivalent entry point. Setup run through `npx` registers
 `cmd.exe /d /c` so clients can start the runner's command shim. Neither
 registration points into the runner's temporary package cache.
 
-REA supports Node.js 22.x (>=22.19), 24.x (>=24.11), and 26+. Node.js 23, 25, and prereleases are unsupported. It uses the npm already paired with that runtime and never upgrades Node.js, npm, or Homebrew.
+REA supports Node.js 22.x (>=22.23.4), 24.x (>=24.22.1), and 26.12.1+. Node.js 23, 25, earlier 26.x releases, and prereleases are unsupported. It uses the npm already paired with that runtime and never upgrades Node.js, npm, or Homebrew.
 
 When REA runs on Bun, the `rea doctor` `node` check reports the Bun version and
 the Node.js compatibility version Bun declares, and passes when that
@@ -1169,6 +1419,21 @@ on the shell `PATH`. Use the setup command above for the guided setup
 journey, `npx -y rea-agents@latest` for unattended one-off commands, or install globally
 with `npm install --global rea-agents` for a shell-visible `rea` command.
 
+For many concurrent agent sessions, run setup directly from a global installation
+to avoid keeping a separate npm launcher process for each MCP connection:
+
+```bash
+npm install --global rea-agents
+rea setup
+```
+
+Review the plan, approve the selected client changes, then restart those clients.
+This registers the installed entrypoint directly; installing the global package
+alone does not replace existing `npx` registrations. Each client still has its own
+REA process and analysis state. Direct registrations follow the installed CLI;
+use `rea update` to update it. Running setup through `npx` again selects the pinned
+package-runner command. Do not register paths inside npm's temporary package cache.
+
 The optional curl wrapper installs only the global npm package:
 
 ```bash

docs/javascript-application-graph.md

@@ -68,9 +68,20 @@ The node vocabulary covers:
 - runtime script instance;
 - an explicit `unknown` kind when classification is not justified.
 
+Static network endpoints can have `resolution: "partial"` when the analyzer
+recognizes a request call but cannot resolve its target to one exact value.
+The observation's `value` then preserves the source expression, and its
+evidence points to the call site. Partial and exact endpoints have distinct
+identities even when their displayed values match. Partial identities also retain
+the source site: matching expression text at different sites does not establish
+the same runtime target. This records observed source syntax; it does not
+establish that the call ran or contacted that endpoint.
+Graph coverage describes the analyzed artifact scope and resource limits, not
+runtime execution. A partial target alone does not imply omitted source.
+
 The directed relation vocabulary is `contains`, `loads`, `imports`, `maps_to`,
-`exposes`, `sends`, `invokes`, `handles`, `calls`, `persists_to`, `observed_as`,
-and `changed_from`. Provider names are deliberately absent from both
+`exposes`, `sends`, `invokes`, `handles`, `calls`, `accesses`, `persists_to`,
+`observed_as`, and `changed_from`. Provider names are deliberately absent from both
 vocabularies.
 
 For a `maps_to` edge targeting a source-map node, the edge carries

docs/javascript-application-workflows.md

@@ -6,7 +6,7 @@ are `inspect_analysis_view`, `trace_application_feature`,
 `trace_javascript_semantics`, `compare_application_versions`,
 `compare_source_to_bundle`, and `compare_javascript_export_shapes`; their CLI
 equivalents use the same names with hyphens. `inspect_analysis_view` projects a
-summary, module page, or one module from already completed application
+summary, module page, typed module references, or one module from already completed application
 Evidence; it does not walk relationships.
 
 `inspect_analysis_view` also projects retained `inspect_binary_layout`
@@ -40,6 +40,111 @@ combine its fields with the fact's location. Semantic trace results include the
 canonical context subset referenced by their returned facts, so those facts
 remain self-contained in the response.
 
+Content-deduplicated module references match every retained observation path.
+A path-filtered reference reports the caller-selected path while keeping its
+shared node identity. Item selection also accepts any retained path when it
+identifies one module; if several module kinds share that path, select the
+reported `node_id`. Node-ID selection uses the stable representative path.
+
+## Retained module facts
+
+For canonical identities and retained source-range counts, query the existing
+`inspect_analysis_view` tool with a typed module-reference projection:
+
+```json
+{
+  "source": {
+    "kind": "retained-evidence",
+    "evidence_id": "ev_ACTUAL_PARENT_ID"
+  },
+  "view": {
+    "kind": "javascript",
+    "collection": "modules",
+    "projection": "references",
+    "where": { "path": "main.js", "kind": "javascript-module" },
+    "offset": 0,
+    "limit": 16
+  }
+}
+```
+
+Replace the parent ID with the actual analysis summary's `parent_evidence_id`.
+Filters match exact retained paths and kinds. Omission includes all values;
+`path: null` selects an unavailable path and `path: ""` selects a recorded empty
+path. Multiple identities can share a path. The default offset is 0 and limit
+is 64; follow `coverage.next_offset` for later matching rows in retained order.
+
+Each row contains its `node_id`, kind, path, canonical identity, retained
+source-range observation count, and distinct supporting authority/state/confidence
+combinations in `basis`. These counts do not prove unique reference sites or
+runtime loading. Use the existing module item view with the row's `node_id` for
+exact source ranges and complete observations. The result retains parent/query
+Evidence identity, artifact identity, limitations and unknowns. `source_coverage`
+keeps selected/unselected sources, graph coverage and unknown counts separate
+from page exhaustion; an exhausted or empty page cannot prove absence from an
+incompletely analyzed application. CLI `inspect-analysis-view` reads the same
+request from a JSON file; portable input uses `source.kind: "inline"` with the
+complete parent Evidence.
+
+## Retained file and native facts
+
+Read metadata, native inventory and unavailable processing scopes directly from
+retained application observations:
+
+```json
+{
+  "source": {
+    "kind": "retained-evidence",
+    "evidence_id": "ev_ACTUAL_PARENT_ID"
+  },
+  "view": {
+    "kind": "javascript-facts",
+    "where": { "paths": ["package.json", "build/Release/addon.node"] },
+    "property_keys": [
+      "name",
+      "version",
+      "bytes",
+      "availability",
+      "operation",
+      "analysis_status"
+    ],
+    "offset": 0,
+    "limit": 32
+  }
+}
+```
+
+Each row preserves its node identity, observation ID, source-map reference and
+original Evidence, including location, authority, confidence, coverage and
+limitations. `property_keys` selects exact retained property names; absent keys
+stay omitted and retained null values stay null. An empty list requests only
+identity/provenance. Observations are not merged into a single asserted fact.
+When property names are unknown, opt into
+`include_available_property_keys: true`. Each returned row then lists its
+original `available_property_keys`, without unselected values; combine this with
+`property_keys: []` for discovery only. The default is false, so known-field
+reads do not pay for field inventories. Keep node/observation IDs paired when
+combining rows; location-related observations need not belong to the same node.
+
+`where.paths` compares exact Evidence `artifact-path` values or `source-range`
+sources. Empty query strings remain valid selectors, though current location
+contracts require nonempty paths/sources. It does not inspect display labels or a
+property named `path`, normalize URLs, or reopen files. Omission includes unknown
+locations; an empty list matches none. Optional `where.kinds` filters original
+node kinds, with an empty list likewise matching none. Multiple locations of
+shared content remain separate rows with the same canonical node identity.
+
+These rows select application-graph observations. Semantic relationships and
+semantic unknown records remain in complete parent Evidence and trace tools.
+
+Rows follow retained node/observation order. The default offset/limit is 0/64;
+follow `coverage.next_offset`. Page totals count matching retained observations,
+while `source_coverage` preserves caller selection and application/semantic
+partial or unknown states. An empty or exhausted page does not establish absence
+outside examined coverage. Complete source Evidence remains available through
+export. CLI `inspect-analysis-view` and portable Evidence use the same contract
+and application implementation as MCP.
+
 ## Feature tracing
 
 Select one literal seed kind: node ID, route, string, API, IPC channel, module,
@@ -55,6 +160,14 @@ matches exactly. Otherwise the result recommends provider-neutral follow-up
 tools and reports `requires-provider-analysis`; it never starts or switches a
 provider implicitly.
 
+A handoff requires the add-on's exact content identity. Unavailable unpacked
+companions remain visible in the trace graph with their canonical path,
+availability and container-backed Evidence. They yield no actionable handoff
+until their bytes are available and the application is reanalyzed; the archive
+digest does not identify the missing add-on. An empty handoff list does not prove
+that the trace contains no native components. Source partial coverage remains
+visible independently of handoff eligibility.
+
 ## Semantic relation tracing
 
 `trace_javascript_semantics` queries the authenticated companion graph without
@@ -76,6 +189,33 @@ registrations/removals and dispatch, Node timers and cancellation handles, async
 `node:child_process` creation and ownership, configuration sources and direct
 defaults, request construction and response consumers, parse/coercion/
 validation boundaries, and built-in resource acquisition/release.
+Global timer, `fetch`, `WebSocket`, `JSON.parse`, and coercion calls receive
+these built-in effects only when the file has no visible possible write to that
+global, including a computed write through a global object. The check does not
+evaluate order or reachability, so a later or guarded polyfill write also
+removes the resolved effect.
+
+Local callback candidates do not establish a complete target set when a
+conditional write may leave a parameter, catch binding, or uninitialized
+variable unchanged. These calls retain candidate edges and dynamic uncertainty;
+default traces remain ambiguous with partial coverage. Opt into
+`include_ambiguous_dynamic_edges` to traverse the candidates. Proven preceding
+unconditional assignments, including writes dominating a read in the same
+conditional block, can resolve a target; later writes do not replace an
+earlier read or captured alias. Unproved ordering across branches or function
+activations remains conservative. Returned-reference effects use the same
+read-position facts as call resolution.
+
+Returned-reference effects share one property-projection state per source read.
+When several paths reach that read, differing keys at the same depth become
+wildcards; differing depths become an unknown-depth selection. This retains a
+conservative superset of the represented local targets rather than enumerating
+every alias history or dropping paths at a query cap. It can add possible effects,
+including for finite projection cycles. Static limitations disclose this loss of
+precision; semantic graphs retain `dynamic-property` unknowns at the affected
+callee coordinates, and captured web bundles identify the affected script keys.
+These candidates do not prove runtime calls or mutations. Existing source-worker
+and result-admission limits still apply.
 
 Mutation tracking covers explicit member assignments, updates, deletion, and
 loop assignment targets, including the supported local aliases and shared
@@ -126,8 +266,13 @@ graph and read literal values from `properties.value`, rather than decoding
 role keys or labels. Previously stored graphs and their original IDs remain
 valid inputs.
 
-Endpoint nodes keep the exact endpoint in each observation's `properties.value`;
-semantic request and response nodes keep it in `properties.endpoint`. Endpoint
+Endpoint observations keep their value in `properties.value`. When
+`properties.resolution` is `partial`, that value is the unresolved source
+expression, with an identity distinct from an exact endpoint of the same text.
+Semantic request and response nodes keep their endpoint in `properties.endpoint`. Endpoint
+lookup identities are preserved independently of the primitive-value lattice's
+byte budget. An oversized normalized literal can become unknown while the same
+string remains an exact endpoint for tracing and comparison. Endpoint
 identity keys use `value-sha256:<digest>` over the canonical JSON string when
 that is shorter, or when the original value starts with the reserved digest
 prefix. Other short endpoint keys remain unchanged. Endpoint labels use the
@@ -165,6 +310,27 @@ and the result includes its basis, candidates, changed dimensions, Evidence
 links, limitations, and a `changed_from` graph containing all compared nodes,
 their observations, and relationships.
 
+A matched pair whose only differing dimension is `location` counts as
+`unchanged` when both input graphs have complete coverage and its neighboring
+relationship identities are resolved. This includes an
+identical module shifted by an earlier bundle edit and a named IPC channel
+observed at a new offset. The item still reports `dimensions: ["location"]`
+and both node IDs; the graph retains both original observations and coordinates.
+Content, property, and complete relationship differences still count as
+`changed`. With incomplete coverage, relocation alone remains `unknown` and
+retains the location and coverage dimensions. Matching confidence is unchanged:
+an inferred pair is not promoted to exact identity by this classification.
+
+Complete inventory does not resolve ambiguous or unmatched relationship targets.
+A matched module with identical bytes can therefore remain `unknown` when a
+neighbor, including its containing chunk, has no cross-version match. The item
+retains its exact content match, coordinates, candidate inventory, and an explicit
+relationship-identity limitation. `coverage` is added to its dimensions only for
+incomplete inventory. Definite content, property, or complete topology differences
+remain `changed` while retaining any separate relationship uncertainty.
+An unknown item with no observed content, property, or relationship difference
+does not emit a `changed_from` edge; its pairing and uncertainty remain in the item.
+
 One-sided absence is `added` or `removed` only when the opposite input graph has
 complete coverage. With partial, unavailable, or truncated input, the same
 condition is `unknown`. The comparison retains every classified item and
@@ -208,6 +374,13 @@ Direct return expressions, including expression-bodied arrows, are evaluated
 through an execution-free value lattice. Literal object fields and direct
 return sites are represented in the result. Calls, dynamic spreads, computed
 keys, and parser recovery remain partial or unknown.
+Rebound values are evaluated at their read position: saving an incoming parameter
+before a later assignment preserves caller-dependent uncertainty, while a proven
+preceding assignment can supply a constant. Unresolved ordering remains unknown.
+Supported reads within object and array literals follow left-to-right evaluation;
+an unproved relative order remains unknown. A later write in a separate conditional
+block can still leave an earlier capture unknown because ordering across those
+blocks is not proven.
 Objects and arrays passed to calls, constructors, or tagged templates, used
 as method receivers, or stored through property targets are not assumed
 unchanged afterward. Property stores conservatively retain reference escape
@@ -239,8 +412,23 @@ when the object is an array element, nested inside mixed object/array
 containers, or assigned to a previously declared identifier. Saving the getter
 result in an alias does not make its old field values immutable. Known array
 indices include holes; an element after a spread has only a minimum possible
-index. Unrelated properties and known array prefixes remain precise, and a
-getter read without a write does not invalidate the receiver's fields.
+index. For these returned-reference writes, unrelated properties and known
+array prefixes remain precise. Accessor applicability uses the shared object
+reference owner for literal definition order, own-property shadowing, known
+array elements and explicit literal prototypes. Getter invocation effects
+project write and call operands through the same local alias and `this` owner
+used by returned references. Reads of another selected getter continue that
+effect walk. A receiver reached by these effects becomes uncertain on an
+ordinary or optional read, compound assignment or update. A simple assignment
+or deletion does not invoke the getter. Selected setters invalidate
+the receiver on assignment, including through known array and prototype paths.
+Inert getters and ordinary data slots retain their supported precision.
+
+These are conservative source-derived effects, not execution results. The
+analysis does not establish general accessor purity, model every class or
+reflectively installed descriptor, or cover getter reads introduced by object
+spread/destructuring and all implicit coercion or disposal invocations. Absence of a later explicit write does not establish getter purity;
+unsupported invocation forms can still retain stale receiver-field facts.
 Nested callable returns are not assigned to their parent callable. Projected
 graph observations carry source ranges but never source text.
 

docs/javascript-artifact-reconstruction.md

@@ -50,6 +50,12 @@ The equivalent MCP input is:
 `input_path` must be absolute. `format` accepts `auto`, `asar`, or `directory`.
 `integrity_policy` accepts `fail` or `record-and-continue`; both CLI routes and
 the MCP tool default to `fail`.
+Evidence parameters retain the caller-selected `input_path`, including a
+symlink alias. Both CLI routes anchor relative inputs to their working
+directory without resolving symlinks. The subject and normalized result use
+the resolved path for artifact identity; the selected path remains a separate
+request fact. Older Evidence can omit this parameter; retained records are
+validated without resolving their paths against the live filesystem.
 The result retains the canonical local path, root artifact digest, artifact
 manifest and graph commitments, JavaScript Application Graph, static
 Electron summary, reconstruction statistics, and explicit limitations. It does
@@ -100,6 +106,18 @@ only reduces delivery; source selection reduces text acquisition and analysis.
 Explicit native provider or snapshot routing rejects `--source-path` rather
 than ignoring it.
 
+ASAR header/index admission defaults to 16 MiB / 16777216 bytes. Set
+`REA_ASAR_MAX_HEADER_BYTES` in the environment that starts REA to a decimal
+byte count of at most 268435456 (256 MiB) to raise or lower it; restart REA
+after changing it. Invalid values are configuration errors, never a silent
+fallback. The archive index is read before any member can be selected, so
+`source_paths`, `--source-path`, `detail: "summary"`, `max_heap_mb`, and
+available host memory do not bypass or raise it. When the header exceeds the
+effective allowance, the failure names it, whether it came from the default or
+the setting, and the setting's limits. Alternatively, analyze an extracted
+application directory, which does not admit an ASAR header. Repeating the unchanged
+request fails the same way.
+
 ## Read the analysis
 
 `analyze_javascript_application` returns the complete analysis Evidence by
@@ -166,7 +184,8 @@ for the tool-result contract.
 
 Use MCP `detail: "summary"` for a large application, then read selected views of
 the retained analysis. Summary changes delivery only. To reduce source parsing,
-select exact [source paths](#select-sources-before-analysis).
+select exact [source paths](#select-sources-before-analysis). Source selection
+does not bypass ASAR header/index admission.
 
 If a complete response exceeds the transport budget, use its retained Evidence
 ID for a summary, module page, or item, or export the bundle. See
@@ -182,6 +201,18 @@ partial Evidence. Cancellation after source analysis begins can also retain
 completed facts, including when interrupted during cleanup, graph construction,
 or Evidence sealing. Unexamined relationships remain unknown.
 
+Per-file gaps identify the applicable processing phase: JavaScript source analysis
+uses AST authority, while package/JSON, HTML-reference and source-map processing
+use artifact-byte authority. File-analysis interruption gaps carry `analysis_status`:
+`interrupted` means the phase was entered (including source-worker startup),
+and `not-started` means interruption prevented that phase from beginning.
+Neither state establishes that parsing completed. Remaining parsers are not run
+after interruption just to populate a result.
+
+Native add-ons remain metadata-only and acquire no JavaScript parse or analysis
+gap. This also applies when their paths are outside `source_paths`; their inventory
+and caller selection remain available without implying applicable AST work.
+
 MCP returns a `partial_observation` retained reference after recording succeeds;
 open it with `inspect_analysis_view` on the same connection. If the SDK stopped
 waiting before receiving the reply, follow [MCP cancellation recovery](mcp-contracts.md#progress-and-cancellation)
@@ -221,7 +252,8 @@ Selected bounded text is then parsed as inert data to recover:
 - Electron preload paths and renderer files visible in static syntax;
 - local HTML script entrypoints;
 - Webpack and Rspack chunk registrations and module factories represented as
-  AST literals;
+  AST literals, plus esbuild `__commonJS` module tables; recovered esbuild
+  CommonJS exports use its factory's `exports`-then-`module` parameter order;
 - static imports, dynamic imports, CommonJS `require` calls, workers, and
   service workers (only `new Worker(new URL(path, import.meta.url))` resolves
   relative to the calling module; string worker and service-worker paths use
@@ -299,13 +331,25 @@ suffix. Utility calls such as `useUtils().invoice.list.fetch()`, `useQueries`,
 and links whose callee was renamed by a minifier are not recognized.
 
 Each recovered bundle module retains the exact factory-source digest. A complete
-bounded AST also receives a `babel-ast-v2` structural fingerprint that ignores
-ordinary identifier names while retaining syntax, literals, operators, object
-keys, and member properties. Computed member identifiers retain their access
-syntax without treating a local variable's spelling as a fixed property; template
-literal contents remain part of the fingerprint. Version comparison keeps the
-algorithm in its matching key, so retained `babel-ast-v1` fingerprints cannot match
-these new fingerprints. Re-analyze both snapshots to compare their new structure.
+bounded AST also receives a `babel-ast-v3` structural fingerprint. It ignores
+ordinary identifier spellings while retaining literal and directive values,
+operators, fixed property names, and selected syntax flags such as computed
+keys, declaration kind, function async/generator state, update-expression
+prefix, method kind, and class static state. Local names used in computed keys
+remain rename-resistant; private member names are also ignored while their
+method kind and static state remain represented.
+
+This fingerprint is structural evidence for candidate matching, not proof that
+two modules behave equivalently. Its scope follows the AST facts listed above;
+unrepresented syntax can still collide. The exact source digest remains
+available to show byte-level change, and comparison reports content changes even
+when structural matching succeeds.
+
+Version comparison includes the algorithm in its structural matching key.
+Retained `babel-ast-v1` results can still match other v1 results, and v2 results
+can still match v2 results, but v1/v2 results cannot structurally match v3
+results. Other matching tiers remain available across algorithm versions.
+Re-analyze both snapshots to compare using the v3 rules.
 If fingerprint construction reaches its structural
 bound, the fingerprint is unavailable and its status is `truncated`; a bounded
 prefix is never presented as a complete stable fingerprint.
@@ -326,11 +370,35 @@ substitute version-dependent Electron defaults for omitted `webPreferences`.
 Unresolved option objects, bridge keys, API objects, and IPC channel expressions
 remain unknown and make coverage partial.
 
+BrowserWindow options and `webPreferences` can follow local `const` aliases
+whose initialization precedes the use. The shared lexical and mutation index
+checks the selected `webPreferences` path, so an unrelated `options.icon` write
+does not discard these facts. Reassignment, relevant writes, prior escapes,
+ambiguous initialization, and unresolved spreads remain unknown. The binding
+association is an inference: the window observation retains the constructor
+location and `resolution_sources` identifies the original object literals in
+the same artifact. Inline option syntax keeps its observed Evidence. This
+does not execute code or establish runtime configuration; popup overrides from
+`setWindowOpenHandler` are not covered by this constant-options resolution.
+
 Recognized literal module-table bootstraps also support bounded lexical
 substitution through module exports, import helper bodies, local wrapper calls,
 returned closures, object spreads, and callback parameters. Recovered Electron
 paths retain source ranges, module identities, literal channels and branch
-conditions as static candidates. Unknown spreads, reassigned bindings, accessors,
+conditions as static candidates. A channel the IPC call site itself
+determines, such as a local constant, resolves that call's observation rather
+than adding a second one. A channel or registration handler supplied through
+a wrapper argument is a candidate beside the call's own unresolved
+observation, because callers outside the traced paths may pass other values.
+A wrapper call that is both its own entry and reached through another entry
+is recorded once when both paths find the same facts; calls that supply
+different handlers or channels stay distinct.
+A contextBridge exposure reached through wrapper calls records one API per key
+those invocations supply, each with its own bounded member paths. Invocations
+that supply the same key to one exposure call describe one API covering each
+alternative object. A traced invocation whose key, API object or world stays
+unresolved or guarded keeps the call's own unresolved observation.
+Unknown spreads, reassigned bindings, accessors,
 dynamic channels and unsupported syntax do not become exact implementations.
 This pass neither executes the bundle nor proves that runtime guards pass.
 An ambiguous or unresolved handler retains its channel registration but does
@@ -339,6 +407,29 @@ conditions retain both return branches, including fallthrough. Short-circuit
 selection distinguishes truthiness from nullishness; uncertain `??` paths
 retain a nullish guard rather than a falsy guard.
 
+Recognized module factories retain a separate initialization boundary for each
+reachable return and for fallthrough. A conditional early return does not hide
+later exports on the continuing path; exports and captured initializers after a
+definite exit remain unavailable. Fallthrough candidates retain their branch
+predicates. Exports shared by the complete set of proven exits also retain
+unconditional facts. Their shared view excludes later overwrites and bounds
+captured initializers by the earliest completion; branch-dependent facts keep
+their predicates. Incomplete or additionally guarded exit sets cannot use that
+proof. Completed factory control flow is reused within that analysis only.
+Guarded main-process registrations remain candidate paths even with empty
+callbacks; they do not become unconditional IPC summary facts.
+Factory guards read export writes at their source position. Destructured and
+spread snapshots retain their capture position rather than reading later writes.
+An unproven namespace read boundary remains unknown.
+Completed factory environments preserve their `exports` and `module.exports`
+alias identity when a namespace escapes through an export or returned closure.
+Synchronous helper calls retain the caller's factory activation and read point;
+lexical captures and executing callers remain separate facts.
+Async calls retain their truthy Promise identity until an explicit `await`;
+generator calls produce iterators without evaluating their bodies. Iterator
+advancement and Promise/iterator members remain explicitly unresolved. Awaited
+function values retain their source origins and captured initialization bounds.
+
 These paths appear as contextual `javascript-expression` nodes. An exact `api`
 seed such as `desktop.copy` can trace an exposed member through a unique IPC
 pairing to an external API call. `addListener` is normalized to `on`. Semantic
@@ -402,10 +493,13 @@ For `.node` bindings, member names mean “requested by JavaScript syntax.” A
 resolved add-on path does not convert them into verified native exports. Native
 symbol verification remains a separate deep-analysis claim.
 
-Endpoint observations preserve useful local diagnostics, including query values
-and fragments, while removing only URL username/password credentials. Artifact
-paths, digests, parse locations, and analysis metadata remain actionable
-because REA is local-only.
+Exact endpoint values preserve useful local diagnostics, including query values
+and fragments, while removing URL username/password credentials when the value
+parses as a URL. When a recognized request target cannot be resolved exactly,
+the endpoint's partial resolution retains the original source expression with
+its call-site provenance. This does not establish that the request executed.
+Artifact paths, digests, parse locations, and analysis metadata remain
+actionable because REA is local-only.
 
 ## Safety boundary
 
@@ -420,6 +514,11 @@ inventory digest before parsing, and malformed ASAR operations return typed
 format diagnostics that retain the local container path. Native add-ons are
 represented by metadata only.
 
+POSIX directory names are read as raw bytes and admitted as logical paths only
+when they round-trip through UTF-8. Invalid names are left untouched, retained
+siblings remain inventoried, and traversal coverage is marked unknown with the
+first rejected basename bytes identified in hexadecimal.
+
 ## Bounds and coverage
 
 The reconstruction keeps internal resource safeguards for artifact entries,
@@ -437,7 +536,18 @@ retained candidate strings and temporary canonical-key serialization. This is
 an in-memory semantic allocation bound independent of CLI or MCP transport
 budgets. Source literal bytes remain in the parsed source and artifact evidence;
 when a normalized semantic value exceeds the bound, its value is unknown and
-coverage is partial. Expression
+coverage is partial. Direct string literals, cooked templates without
+substitutions, and derived strings all use this normalization boundary. A
+single exact primitive is validated without allocating deduplication or sort
+state or serializing a canonical key. Configuration defaults use the same
+normalizer; limited defaults retain `value_resource_limit` and an anchored
+resource-limit unknown. Known null, false, zero, and empty-string defaults stay
+exact. This primitive-value bound does not truncate lookup identities such as
+source paths, field names or request endpoints. Full endpoint properties remain
+available for endpoint queries, tracing and comparison. Each analysis records encountered limit classifications independently of
+later projections: a selected healthy sibling can remain exact while whole-file
+coverage reports the encountered bound. Projection preserves limiting values
+from the shared property-slot owner. Expression
 evaluation and provenance walks stop after 256 nested levels. Union products
 and string lengths are checked before allocating combinations or concatenated
 strings. When a boundary is reached, the value stays unknown, semantic coverage
@@ -511,20 +621,68 @@ The default semantic capacities are 100000 nodes across the application and
 `--max-semantic-nodes-per-file`), each from 1 to 1000000. The per-file offer is
 also bounded by the remaining application capacity. Raising these controls does
 not bypass worker heap, execution deadline or parent result admission. Truncated
-coverage reports both effective node ceilings; omitted relationships remain
-unknown.
+coverage lists both effective ceilings, `max_semantic_nodes` and
+`max_semantic_nodes_per_file`. The limitation names only the ceiling that
+stopped retention. Omitted relationships remain unknown.
+
+Unknown records have independent defaults of 100000 across the application and
+20000 per source. Select `max_semantic_unknowns` and
+`max_semantic_unknowns_per_file` (CLI: `--max-semantic-unknowns` and
+`--max-semantic-unknowns-per-file`), each from 1 to 1000000. The parent offers the
+remaining capacity to the worker, and rejects excess records before adoption.
+Unknown admission happens before canonical identity and provenance construction.
+When capacity is exhausted, coverage is partial and truncated. Coverage lists
+both effective unknown ceilings, `max_semantic_unknowns` and
+`max_semantic_unknowns_per_file`. The limitation names only the ceiling that
+stopped retention. `coverage.omitted_unknown_occurrences` counts examined
+emission occurrences by family and reason, including repeated emissions. It does
+not claim distinct discarded identities or count frontiers in unexamined source.
+Omitted unknown provenance and IDs are not retained. Existing source, node,
+execution and memory limits continue to determine examined coverage. Raising only
+a node capacity does not raise the unknown capacity. Graphs without unknown
+omissions omit this optional coverage field, preserving their existing identities.
+
+These node and unknown capacities bound retained graph projection. Parsing,
+static analysis and semantic fact construction precede that projection for an
+admitted source. Lowering the capacities can reduce projection and retention
+work, but does not bound upstream semantic-analysis time or memory. Select fewer
+`source_paths` to reduce the sources analyzed; heap and deadline controls set
+separate resource failure boundaries with the runtime-specific limits above.
 
 While sources remain, the sequential worker reserves one next compact static
 checkpoint at its 8 MiB bound with the same 6x retention expansion used at
 admission. Earlier checkpoints are already charged to retained bytes; the
 remaining inventory does not multiply the reserve. Streamed file transfers keep the default 512 MiB
 allowance, and can grow with the parent's offered retention budget (8x byte
-expansion), up to 2 GiB. Individual records remain bounded to 8 MiB. The parent
+expansion), up to 2 GiB. Individual records remain bounded to 8 MiB. The 8 MiB
+record ceiling and the 2 GiB file maximum have no runtime override, and raising
+`max_heap_mb` does not raise them. The effective file ceiling follows the
+parent's offered result budget. On Node, a larger CLI/MCP process heap
+(`NODE_OPTIONS`, for example `--max-old-space-size=<MiB>`, after restarting that
+process) can raise it up to 2 GiB. Selecting fewer `source_paths` can also
+raise that effective allowance, because the reserve for remaining sources and
+earlier retained results reduce the offered budget. Neither change raises the
+record ceiling or the 2 GiB maximum. The parent
 still independently checks actual bytes, record overhead, digests and source
 identities before adopting results.
 
 A source that exhausts the analyzer heap does not terminate the CLI/MCP
-process. Completed static facts are retained, other files can still be analyzed,
+process. On Node, the failure names `max_heap_mb` (CLI `--max-heap-mb`), the
+effective mebibytes, the inclusive range 128 to 16384, and the 1024 MiB
+default. A changed value applies on the next analysis, which starts a new
+worker. It does not change a worker that is already running and does not
+require restarting the REA server. On Bun, the same heap-exhaustion diagnostic
+is not attributed to that V8 old-space limit, because Bun ignores the flag.
+Raising `max_heap_mb` does not raise the parent result budget,
+the transfer record ceiling, the 2 GiB transfer file maximum, the CLI/MCP
+process heap, or fixed parser limits. The parent result budget follows that
+process heap. On Node, set `NODE_OPTIONS` (for example `--max-old-space-size=<MiB>`)
+in the environment that starts the CLI or MCP server, then restart that
+process, to raise the parent heap and the result budget. A completed static
+checkpoint that does not fit the parent's remaining heap names that parent
+process heap. Increasing `max_heap_mb` does not raise it. On Node, the same
+`NODE_OPTIONS` restart can raise the parent heap and admit the checkpoint.
+Completed static facts are retained, other files can still be analyzed,
 and unexamined semantic relationships remain explicit unknowns. The parent
 also checks transfer expansion and accumulated results against its remaining
 heap before accepting semantic facts. A resource-limited result has partial
@@ -611,6 +769,24 @@ URL targets before selecting a later package-local file; an unsupported bare
 package or recursive alias keeps its unresolved outcome rather than selecting
 a later fallback.
 
+For a proven `createRequire(new URL(relativeFile, import.meta.url))` call,
+relative `require()` edges use the selected filename's directory and retain the
+source-relative `resolution_base_path` on the graph edge. This follows Node's
+`createRequire(filename)` contract; the path is a loader-resolution fact and
+does not change the source file or call-site Evidence identity. If the filename
+cannot be proven, REA keeps the require target unresolved rather than silently
+resolving it from the caller's directory.
+
+Module-reference graph edges, including native-addon `loads` edges, retain
+`properties.operation`: `load` identifies
+an import or loader invocation that may execute the target; `resolve` identifies
+a proven `require.resolve()` lookup that discovers a target without executing it.
+Both retain the exact specifier, resolution mode, selected resolution base, and
+call-site Evidence. Resolve-only edges do not propagate application ownership or
+create executable-import gaps when their targets are missing. A modified
+`.resolve` member remains an uncertain `load`, because its replacement may execute
+code. These static facts do not establish that a call runs at runtime.
+
 ESM relative module paths, legacy package subpaths without exports, and selected
 package exports targets use URL suffix handling and one percent-decoding pass.
 Bare specifiers retain their complete query/fragment text for exports key matching.

docs/javascript-recovery.md

@@ -129,12 +129,20 @@ equivalence remain unknown. No network fetching occurs.
 
 ## Resource policy and verification
 
-Each operation uses one worker, a 1 GiB address-space ceiling and a 120-second
-deadline; it accepts up to 64 MiB of total input across all selected files and
-128 MiB/10000 entries of staged
-output, with 8 MiB per diagnostic/report stream. Exceeding a limit is an explicit
-failure with rollback. These limits apply per operation and do not claim aggregate
-host containment across concurrent callers.
+Each operation uses one worker, a 1 GiB address-space ceiling and a 120000 ms
+deadline. That deadline has no runtime override: raising `analysis_timeout_ms`
+or `max_heap_mb` does not raise it. An unchanged retry may hit the deadline
+again. The operation accepts up to 64 MiB of total
+input across all selected files and 128 MiB/10000 entries of staged output,
+with 8 MiB per diagnostic/report stream. A `provenance.json` over that report
+allowance names the fixed limit: this release has no runtime override, and
+raising `max_heap_mb` or `analysis_timeout_ms` does not raise it. Malformed
+reports stay distinct from that size refusal. Exceeding a limit is an explicit
+failure with rollback. Selected inputs beyond the total input allowance return a
+non-retryable `resource_constraint` with `resource: "file-size"` and
+`reported_limits.maximum_input_bytes`/`remaining_input_bytes`; input that is not
+valid UTF-8 is an `invalid_format` request issue. These limits apply per
+operation and do not claim aggregate host containment across concurrent callers.
 
 See [testing.md](testing.md#javascript-source-recovery) for the focused real
 Wakaru CLI/MCP lane and its source-owned compiler fixtures. Other host platforms

docs/javascript-route-inventory.md

@@ -0,0 +1,113 @@
+# Static JavaScript route inventory
+
+`analyze_javascript_application` and `analyze_web_bundle` retain a shared
+`route_inventory` alongside their existing graph and endpoint observations.
+The inventory answers which route patterns the supplied code suggests, where
+each pattern came from, and which parts remain unknown. It never requests a
+route, executes scripts, expands a sitemap, or establishes reachability.
+
+Select a compact view in either MCP analysis call:
+
+```json
+{
+  "detail": "summary",
+  "view": {
+    "kind": "javascript-routes",
+    "collection": "routes",
+    "where": { "path_prefix": "/reports" },
+    "fields": ["path", "classification", "confidence"],
+    "limit": 20
+  }
+}
+```
+
+Keep the original analysis inputs in that call. For the CLI, pass the same
+selection as `--view '<JSON>' --detail summary --evidence_output routes.json`.
+The complete Evidence bundle is written without overwriting existing files.
+Unselected delivery retains the existing complete result. A web-bundle summary
+requires an explicit route view; application summaries also support the existing
+general summary when `view` is omitted.
+
+Use the returned `normalized_result.parent_evidence_id` with
+`inspect_analysis_view` on the same connection. The CLI's exported bundle can be
+reopened with `inspect-evidence-bundle`; its `records` contain the complete
+Evidence. To use `inspect_analysis_view`'s `evidence-file` source, extract the
+selected parent record into a separate local Evidence JSON file (that source
+accepts one Evidence record, not a bundle). Views do not reopen source files or
+reconnect to a browser. Historical Evidence without `route_inventory` is
+unmeasured, rather than an empty route set.
+
+Collections are `summary`, `routes`, `occurrences`, `sources`, and `unknowns`.
+All row collections support exact batch `where.ids`, field selection, `offset`,
+and `limit`. An explicit empty filter array matches nothing. Routes also support
+exact `paths`, literal `path_prefix`, `classifications`, `mechanisms`, and
+`min_confidence`; occurrences support `route_ids`, `source_ids`, and `mechanisms`;
+sources support `keys` and `statuses`; unknowns support `source_ids` and `kinds`.
+Follow `coverage.next_offset` until `coverage.exhausted`; a page may stop early
+at the 1 MiB serialized view budget. Oversized indivisible facts return a
+resource error carrying the retained parent Evidence ID.
+
+Pattern IDs bind the exact path. Source IDs bind the source key, path, original
+URL, content digest and container digest. Occurrences identify a detector mechanism at a source location; one location
+may support multiple mechanisms. `source_count` counts distinct sources. Lookup IDs use 128 bits of SHA-256;
+full content and inventory digests remain SHA-256. Field selection always keeps
+row identity and necessary provenance. The page's deduplicated `sources` table
+contains the source descriptors referenced by returned rows. Retrieve all
+occurrences by route ID when independent supporting evidence matters. Source
+ranges use one-based lines and zero-based UTF-16 columns with exclusive ends;
+filename-only discoveries have no invented AST range.
+
+The shared parser handles literal router calls, nested route objects, typed
+`_$` trees with explicitly inferred property-name fallbacks, Next App Router
+page paths, and navigation targets. It follows stable lexical bindings and
+keeps unresolved parents, mutations, spreads, and dynamic fields explicit.
+A selected route-builder root owns its configuration even when mutation prevents
+trusting its value. Ordinary wrapper keys do not become route ancestors. Missing
+outer ancestors of unbound typed subtrees remain unknown rather than invented.
+Generic Next runtime path constants require a top-level application segment
+supported by structural or navigation candidates across the supplied sources.
+This limits framework filesystem noise, while leaving unsupported candidates
+unknown. The parser shares the existing AST traversal and lexical index.
+
+[Next dynamic segments](https://nextjs.org/docs/app/api-reference/file-conventions/dynamic-routes)
+map to `:id`, `:slug+`, or `:slug*`; groups and parallel slots do not contribute
+URL segments. Intercepted routes remain unresolved. A prefix before `/_next/`
+is an **inferred** routing base: [assetPrefix](https://nextjs.org/docs/app/api-reference/config/next-config-js/assetPrefix)
+can differ from the routing base. Every affected occurrence keeps
+`unprefixed_path`, `inferred_base_path`, and low confidence. `system` classification
+is explicitly a path-name heuristic; `page` uses page-file or structured/typed configuration support. Name-only
+`route`/`addRoute` calls remain `unknown` because they can also register server APIs.
+
+Next page-source segments keep their filename spelling when converted to route
+patterns. Next's [page-path conversion](https://github.com/vercel/next.js/blob/v16.4.2/packages/next/src/shared/lib/page-path/absolute-path-to-page.ts#L21-L39)
+removes page extensions and the `index` tail without URI-decoding segments, so
+ordinary percent sequences are not decoded: `%61` and `a` remain separate
+candidates and a literal `%` is retained. Its [App pathname normalizer](https://github.com/vercel/next.js/blob/v16.4.2/packages/next/src/server/normalizers/built/app/app-pathname-normalizer.ts#L5-L18)
+specifically applies the [underscore normalizer](https://github.com/vercel/next.js/blob/v16.4.2/packages/next/src/server/normalizers/underscore-normalizer.ts#L2-L9),
+which maps `%5F` to `_`; private-folder checks apply to the source spelling
+before that escape is mapped. When classifying a chunk URL, REA reads its
+pathname, then applies these same Next-specific rules. This is static source-path
+interpretation, not browser request-path decoding or reachability.
+
+Known capture omissions remain inline in `summary.reported_capture_omissions`
+and queryable as `unknowns` of kind `capture-omission`. Each keeps a stable gap ID,
+its script-record/script-source unit, and the reported count (including `null`
+for unknown). They do not invent missing script identities or source ranges.
+
+Counts describe deduplicated patterns in the provided code sources. They are
+not a denominator for complete application or capture coverage. In particular,
+server-only, lazy, inaccessible, or unselected sources may be missing.
+`coverage.application_routes` always remains `unknown`; parsing and candidate
+enumeration have their own completeness. Static work and fact admission share
+the existing 8 MiB per-source transfer ceiling, and the root inventory has an
+8 MiB admission budget. Exhaustion preserves a prefix with explicit capacity;
+the serialization owner also admits route facts within the actual space left by
+legacy facts. `result-bytes` capacity preserves an earlier work-capacity reason;
+if even that envelope cannot fit, the optional route facet remains unmeasured.
+These are accounting bounds, not a claim that total process memory is 8 MiB.
+
+Projection only reduces delivery, not analysis work. Retained immutable Evidence
+can reuse validated tables through a weak cache with the snapshot's lifetime;
+subject identity is checked on every query. No live state or source content is
+cached as current. Use existing source selection and resource controls when
+they match the investigation, keeping their partial coverage visible.

docs/javascript-runtime-observation.md

@@ -105,11 +105,15 @@ The direct observations are:
 - `Debugger.scriptParsed`, establishing that a bounded script location was
   parsed in the observed execution context;
 - `Debugger.scriptFailedToParse`, retained only as a bounded invalid-script
-  count;
+  count in `scripts.parse_failed_total`. A well-formed parse-failure
+  event is distinct from malformed protocol data and does not make loaded
+  script coverage unavailable. Older Evidence may omit this count;
 - `Runtime.executionContextCreated`,
   `Runtime.executionContextDestroyed`, and
   `Runtime.executionContextsCleared`, establishing context lifecycle within
-  the capture window.
+  the capture window. A created or destroyed event whose context ID is
+  missing or not a safe integer adds no context record; it is retained only
+  as a count reported in `unknowns`.
 
 The Inspector protocol does not directly expose require/import caller edges,
 EventEmitter emissions or listener invocation, Electron IPC messages or
@@ -117,6 +121,12 @@ handlers, or script-unload events. Those facts remain explicit unknowns.
 `scriptParsed` proves script presence, not the importing module, feature
 execution, initialization order, or causality.
 
+For an authorized file script, `location.file_path` is the verified filesystem
+identity and optional script-item `reported_url` preserves the exact Inspector
+URL. Script identity includes that reported URL, so query and fragment
+variants of one file remain separate observations. Older Evidence may omit
+`reported_url` and remains readable.
+
 ## Node and Electron roles
 
 The caller declares one role: `node`, `electron-main`, `electron-preload`, or
@@ -150,6 +160,15 @@ and uncollected scripts, but scripts collected before attachment may be
 missing. Absence from the observation window never proves that a script or
 behavior does not occur.
 
+Cancellation stops new target-file authorization after the current filesystem
+operation settles. Cancellation diagnostics retain the examined target prefix;
+unexamined targets remain unknown. Observation cancellation keeps its original
+reason and any completed Evidence. Finalization waits for REA's bounded socket
+close (up to two seconds for graceful and forced confirmation). A failed close
+retains its exact transport owner and blocks replacement connections until
+cleanup succeeds; CLI finalization and MCP shutdown also retry retained cleanup.
+Closing REA's connection does not stop the caller-owned runtime.
+
 ## Static correlation
 
 Pass the resulting observation Evidence and an

docs/javascript-runtime-reconciliation.md

@@ -57,7 +57,11 @@ are complete Evidence objects, not only evidence IDs:
 `static_layers` requires exactly one `application` role and permits additional
 `cache` and `assets` roles. A `file-root` mapping translates a canonical runtime
 filesystem root to an artifact-relative prefix. A `url-prefix` mapping does the
-same for one exact HTTP(S) origin and path prefix. Mappings cannot add CDP
+same for one exact HTTP(S) origin and path prefix. Each URL path segment below
+the prefix is percent-decoded separately: an encoded slash (`%2F`) stays inside
+its segment and never names a nested artifact, so such a location, like one
+with malformed or non-UTF-8 escapes, has no artifact path under that mapping.
+An encoded backslash (`%5C`) remains a literal name byte. Mappings cannot add CDP
 origins, broaden Electron roots, read files, or authorize source-map access;
 they are caller-declared inference inputs applied only to already-authorized
 locations.
@@ -106,7 +110,12 @@ nodes. REA evaluates compatible static candidates conservatively:
    that disagreement. If a mapped candidate's digest is unavailable and no
    mapped content match is known, the comparison remains `unknown`.
 6. Missing authorization mappings and incomplete or truncated input coverage
-   remain unknown or truncated rather than becoming absence claims.
+   remain unknown or truncated rather than becoming absence claims. A runtime
+   location with no automatic artifact-root or caller-declared mapping into
+   any layer, and no content match, is `unknown` with
+   `no-authorized-location-mapping`, with or without captured bytes; several
+   content matches remain `ambiguous` under rule 4. Only a mapped location can
+   be `unmatched` with `no-static-content-match`.
 
 Every successful cross-layer `observed_as` edge has
 `authority: cross-layer-reconciliation` and `state: inferred`. The underlying
@@ -128,6 +137,8 @@ classification:
 - `unknown` covers incomplete captures, layers outside runtime scope, and
   unresolved correspondence. Nodes outside a mapped artifact prefix stay
   unknown even when another prefix in the same static layer was observed.
+  Nodes beneath a mapped artifact prefix also stay unknown when a runtime
+  script or worker under that mapping has no artifact path.
 
 None of these states proves feature execution, initialization order,
 reachability, or causality. In particular, `not-observed-in-capture` never means

docs/jeb-analysis.md

@@ -8,7 +8,16 @@ Start a JEB client that serves MCP and point REA at it:
 
 - Default endpoint: `http://127.0.0.1:8425/mcp`. Override with `REA_JEB_MCP_URL` (absolute `http`/`https` URL with literal host `127.0.0.1` or `[::1]`; redirects, remote hosts, credentials, query strings, and fragments are rejected rather than silently dropped).
 - In the JEB GUI client, start the MCP server from the _File_ menu (or via the VIBRE assistant).
-- Headless, use JEB's documented programmatic route (`JebMcpServerInstance` from the JEB API) with the JEB client on your classpath; JEB 5.48's `-c` headless mode does not accept `--mcp` directly.
+- Headless, the official Linux CE build `5.48.0.202610071539` supports `-c --mcp`. From the JEB installation directory, select JDK 21 or newer through `JAVA_HOME`, choose your target and bind explicitly to loopback:
+
+  ```sh
+  "$JAVA_HOME/bin/java" -jar bin/app/jeb.jar -c --mcp --interface=127.0.0.1 \
+    --port=8425 --infile=/absolute/path/to/target
+  ```
+
+  Complete the vendor's first-use EULA and license-key activation when prompted. The Linux audit used an isolated installation, a 1536 MiB Java heap and one active processor inside a separately bounded process family; JVM heap limits alone do not bound the whole family.
+
+- JEB's documented programmatic route (`JebMcpServerInstance` from the JEB API) remains an alternative. The earlier licensed 5.48 headless verification used that route because its CLI did not accept `--mcp`. Check your installed build's `-c --help` and live tool catalog; the version number alone does not establish the launcher options or supported operations.
 
 For setup registration, set `REA_JEB_MCP_URL` in the REA process environment
 before running `rea setup`. The approved setup plan preserves the exact selected
@@ -19,19 +28,21 @@ JEB service is reachable or that a project is open; use the existing
 MCP `binary_session` probes the live JEB service and its advertised tool catalog.
 Availability is reported per operation: a reachable headless client can support
 inspection and decompilation while refusing `open_jeb_project`. Catalog discovery
-uses one 60-second deadline across connection and pagination; a responsive server
-cannot keep discovery running indefinitely by returning more pages.
+is one deadline across connection and pagination; a responsive server cannot keep
+discovery running indefinitely by returning more pages.
 
-Each upstream tool call has an absolute 60-second request deadline by default.
-Set `REA_JEB_MCP_TIMEOUT_MS` to a positive decimal number of milliseconds to
-select a different per-request budget, for example `REA_JEB_MCP_TIMEOUT_MS=300000`
-for a five-minute decompilation. This budget starts after discovery and applies
-to each individual `tools/call`; it does not extend the separate catalog deadline
-or establish one end-to-end deadline across a whole CLI or MCP operation. A
-request that reaches its budget returns a typed timeout, while caller
-cancellation remains a cancellation.
+Catalog discovery and each `tools/call` share `REA_JEB_MCP_TIMEOUT_MS`. Set it
+in REA's launch environment to an integer from 1 to 2147483647 milliseconds,
+for example `REA_JEB_MCP_TIMEOUT_MS=300000` for a five-minute budget. Omitted
+means the pinned MCP client SDK default of 60000 ms (`@modelcontextprotocol/client`
+2.3.1). Restart REA and reconnect to apply a change. Raising the setting raises
+both phases. It does not make one end-to-end deadline across a whole CLI or MCP
+operation: discovery and each `tools/call` each get that budget. A request that
+reaches its budget returns a typed timeout naming the phase and the effective
+milliseconds, while caller cancellation remains a cancellation. Allow execution,
+cleanup, and delivery headroom in the outer MCP client deadline.
 
-Verified against JEB 5.48.0 (Linux x64, licensed build) serving `jeb-mcp-server` 1.3.0 over streamable HTTP.
+Earlier verification used JEB 5.48.0 (Linux x64, licensed build) serving `jeb-mcp-server` 1.3.0 over streamable HTTP. The official Linux x64 CE build `5.48.0.202610071539` was also verified through real CLI and production stdio MCP native-code workflows; see the [Linux CE audit](https://github.com/morluto/rea/blob/42e1a17d6e6c0487f17be0bae248553c4caba7f7/experiments/analysis-efficiency/jeb-linux-native.md) for measurements and limits. CE verification does not establish licensed/Demo Android decompilation or GUI behavior.
 
 ## Operations
 
@@ -42,16 +53,18 @@ Verified against JEB 5.48.0 (Linux x64, licensed build) serving `jeb-mcp-server`
 | `list_jeb_units` / `list-jeb-units`         | Page units by stable path and type (`filter`, `parent-unit-path`, `index`, `count`)                                             |
 | `decompile_jeb_item` / `decompile-jeb-item` | Decompile one type or method to engine pseudo-code                                                                              |
 
-All four are read-only with respect to the target file; only `open_jeb_project` changes engine-side project state, and Evidence records the endpoint, the caller-selected parameters, and the engine-reported digests. Mutation workflows the engine also offers (renames, comments, `save_project`, `close_project`) are intentionally not exposed yet.
+All four preserve target-file bytes. `open_jeb_project` selects or creates an engine-side project; analysis queries can populate computed analysis within it. Evidence records the endpoint, caller-selected parameters, and available engine-reported digests. Mutation workflows the engine also offers (renames, comments, `save_project`, `close_project`) are intentionally not exposed yet.
 
-**Engine coverage note (verified):** JEB 5.48.0 headless MCP instances do not advertise `open_project` — the target must be selected when the engine client starts (`--infile` on the programmatic launcher, or the GUI's file opening). REA preserves that refusal with the engine's reason instead of failing silently; `open_jeb_project` completes only against engine builds that advertise the tool, such as GUI clients.
+**Engine coverage note:** The earlier licensed 5.48.0 headless verification did not advertise `open_project`; it selected the target when the engine client started. For that build, use `--infile` on the programmatic launcher or open the file in the GUI. The audited CE headless catalog also omitted `open_project`; the verified CLI route above selects its input at startup. Use `binary_session` to inspect the current engine's advertised operations. REA preserves an unavailable tool's refusal; `open_jeb_project` requires an engine that advertises `open_project`. The CE audit does not establish GUI project-opening support.
 
 ## Semantics
 
 - Paths in `open_jeb_project` are resolved by the JEB process, not by REA; REA does not require the artifact to exist locally.
 - The engine caps `list_units` pages at 100; a full page is reported as `coverage: "partial"` because more units may exist. An empty remainder is `complete` only when fewer units than requested were returned.
 - When `decompile_jeb_item` omits `unit_path`, the engine implicitly selects the project's first code unit; REA reports `unit_path: null` and a limitation instead of guessing the effective unit.
+- `item_address` is a JEB code-item reference, such as a value returned by the engine's `list_code_items`. Preserve that representation and select its code unit explicitly. For the audited CE SQLite unit, the exported method reference was `sqlite3_libversion_number`; using its raw ELF symbol address did not produce a successful decompilation. File offsets, ELF virtual addresses and engine item references have different semantics. REA currently exposes unit listing, while code-item discovery remains an upstream JEB operation.
 - Engine refusals (`success: false` envelopes) are preserved as unsupported-target errors carrying the engine's own message, for example a nonexistent item address.
+- Successfully parsed client information is reused for the lifetime of the MCP connection, including when the engine omits its version. Explicit client inspection and availability probes refresh it.
 - An unreachable endpoint is a provider-availability failure with recovery guidance; REA does not retry or spawn the engine.
 
 Real-engine verification lives in the JEB lane described in [testing.md](testing.md); mock transport tests do not establish real-engine compatibility.

docs/kuna-provider.md

@@ -13,39 +13,53 @@ rea doctor --provider kuna
 
 Select it explicitly with `--provider kuna`, MCP's `open_binary.provider_id: "kuna"`, or `REA_ANALYSIS_PROVIDER=kuna`. With Kuna and another provider configured for the same target, `auto` reports `ambiguous` instead of choosing one.
 
+Ordinary readiness cancellation returns `cancelled` (CLI SIGINT/SIGTERM exits
+130/143); an earlier recorded timeout or physical failure keeps its diagnostic.
+Completed version, hash and script observations remain in `kuna_installation`
+with `status: "partial"` and `completed_hash` when available. Typed errors carry
+this in `details.diagnostics`; doctor cleanup checks carry it in their `details`.
+Omitted fields remain unobserved; availability remains unknown until readiness
+completes. Cleanup retries keep both the observations and the exact file owner.
+
 | Setting                       | Default                                                                                                                                 | Meaning                                                                                                                                                                                  |
 | ----------------------------- | --------------------------------------------------------------------------------------------------------------------------------------- | ---------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- |
 | `REA_KUNA_COMMAND`            | none (required)                                                                                                                         | Absolute path of the `kuna` executable.                                                                                                                                                  |
 | `REA_KUNA_CACHE_DIR`          | `~/Library/Caches/rea/kuna` (macOS), `$XDG_CACHE_HOME/rea/kuna` (Linux; `~/.cache/rea/kuna` when `XDG_CACHE_HOME` is unset or relative) | Absolute directory REA owns for project exports. Kuna's own spec cache also lives here, not in your Kuna cache.                                                                          |
 | `REA_KUNA_CACHE_MAX_BYTES`    | 4 GiB                                                                                                                                   | Size cap; least recently used exports are evicted above it.                                                                                                                              |
-| `REA_KUNA_PROJECT_TIMEOUT_MS` | 30 minutes                                                                                                                              | Deadline for one project export.                                                                                                                                                         |
-| `REA_KUNA_TIMEOUT_MS`         | 120 s                                                                                                                                   | Deadline for one single-function decompile.                                                                                                                                              |
+| `REA_KUNA_PROJECT_TIMEOUT_MS` | 30 minutes                                                                                                                              | Deadline for whole-program inventory/export, reference and string scans.                                                                                                                 |
+| `REA_KUNA_TIMEOUT_MS`         | 120 s                                                                                                                                   | Deadline for one single-function decompile, and for the inventory run REA makes when the export ends without one.                                                                        |
 | `REA_KUNA_JOBS`               | 1                                                                                                                                       | `decompile-project --jobs`. 1 is Kuna's reference answer; more jobs finish sooner but lose callee-first prototypes, and change about 40–60% of function bodies on the measured binaries. |
 
+A single-function decompile, inventory, string, graph or xref run that exceeds its deadline fails as `provider_timeout`, naming the setting that owns the deadline and that raising it takes a restart. A project export that exceeds `REA_KUNA_PROJECT_TIMEOUT_MS` is an export failure instead: requests that need the export report `capability_unavailable` with the export's own retry guidance.
+
 Windows is not supported yet. Supported targets are ELF and thin Mach-O executables for x86-64, AArch64 and ARM; PE and universal Mach-O are rejected with a typed reason.
 
 ## What Kuna answers
 
-| Tool                                                        | Source                                                                                                                    |
-| ----------------------------------------------------------- | ------------------------------------------------------------------------------------------------------------------------- |
-| `list_procedures`, `search_procedures`, `procedure_address` | Kuna's callable inventory (`kuna functions --json`), including import slots it does not decompile                         |
-| `address_name`                                              | Function entries, then the data labels Kuna's export assigns                                                              |
-| `procedure_pseudo_code`, `batch_decompile`                  | The project export, or a single-function run while the export builds                                                      |
-| `procedure_assembly`                                        | The project export's labeled listing                                                                                      |
-| `analyze_function`                                          | Pseudocode and assembly from the export; callers, callees, references, comments, strings and CFG are declared unavailable |
+| Tool                                                        | Source                                                                                                                           |
+| ----------------------------------------------------------- | -------------------------------------------------------------------------------------------------------------------------------- |
+| `list_procedures`, `search_procedures`, `procedure_address` | Kuna's callable inventory (`kuna functions --json`), including import slots it does not decompile                                |
+| `address_name`                                              | Function entries, then the data labels Kuna's export assigns                                                                     |
+| `procedure_pseudo_code`, `batch_decompile`                  | The project export, or a single-function run while the export builds                                                             |
+| `procedure_assembly`                                        | The project export's labeled listing                                                                                             |
+| `procedure_callers`, `procedure_callees`                    | Call-kind edges from a lazy, body-free Kuna graph; synthetic import nodes are accepted by address                                |
+| `list_strings`, `search_strings`                            | Typed string candidates from one cached, reference-free scan; exact-address, literal and ECMAScript regex selection              |
+| `xrefs`                                                     | Exact-address incoming reference sites, with forwarding aliases filtered explicitly                                              |
+| `analyze_function`                                          | Pseudocode and assembly from the export, plus call relationships; typed references, comments, strings and CFG remain unavailable |
 
-Cross-references, callers, callees and basic blocks are not available from this provider. Use another provider when you need them.
+Typed procedure references, basic blocks, raw-memory tools, annotations and type recovery are not available from this provider. Use another provider for those operations.
 
 ## How answers are produced
 
-The first analysis request in a session looks for this target's export in the cache. `list_documents` reads the selected path without initializing analysis. The export is keyed by the target's SHA-256, the Kuna executable's SHA-256, its version and `--jobs`.
+The first text or assembly request in a session looks for this target's export in the cache. `list_documents` reads the selected path without initializing analysis. Inventory, function lookup, string queries, call relationships and xrefs do not start a C export. A data-name lookup starts it only when the address is absent from the function inventory. The export is keyed by the target's SHA-256, the Kuna executable's SHA-256, its version and `--jobs`.
 
-- **Warm cache:** analysis reads reuse the verified export. A fresh server verifies its file digests; that cost scales with export size.
+- **Warm cache:** analysis reads reuse the verified export. A fresh server verifies its file digests once, even for concurrent first requests; that cost scales with export size.
 - **Cold cache:**
   - REA starts `kuna decompile-project` in the background on a private copy of the target.
-  - Meanwhile, pseudocode comes from one `kuna decompile --addr` run, and listings from `kuna functions --json`.
-  - Requests for the same function share one run; it stops when its last waiting request is cancelled. Each run analyzes the whole image again in its own process, so one REA process runs at most as many at once as the host has cores beyond `REA_KUNA_JOBS` (at least one). Further requests wait in order, and a waiting request is answered from the export if it finishes first.
-  - Assembly and data names wait for the export; progress notifications report the wait every ten seconds.
+  - Meanwhile, pseudocode comes from one `kuna decompile --addr` run. Name and address lookups, and listings, use the inventory the export takes with `kuna functions --json` before it decompiles. The session runs `kuna functions` itself, under `REA_KUNA_TIMEOUT_MS`, only when no export is building (an inventory-only request does not start one) or the export ends without an inventory.
+  - Requests for the same function share one run; it stops when its last waiting request is cancelled. Each run analyzes the whole image again in its own process, so one REA process runs at most as many at once as the host has cores beyond `REA_KUNA_JOBS` (at least one). Further requests wait in order, and a waiting request is answered from the export if it finishes first. A run whose stop REA cannot prove keeps its slot until `close_binary` proves it stopped.
+  - The session keeps each completed single-function text, up to 16 MiB in all with the least recently used dropped first, and answers repeats from it until the export can. After a failed export that lasts for the session.
+  - Assembly and data names wait for the export; progress notifications report the wait every ten seconds. A session does not rebuild a failed export: requests that need it fail with `retryable: false`, the export's failure, its captured Kuna output and what to change before reopening the target. A setting such as `REA_KUNA_PROJECT_TIMEOUT_MS` or `REA_KUNA_JOBS` is read when REA starts, so changing it also needs a restart. Pseudocode keeps using single-function runs.
   - A single-function answer says so in its limitations. It lacks the export's whole-program prototypes, so a later request can return different text for the same function.
 
 Pseudocode, assembly, `analyze_function` and data-name answers record where their text came from in `raw_result.source`: `single-function`, or `project-export` with the cache directory. `procedure_pseudo_code` and `procedure_assembly` answers from the export also name the file, byte span and the file's SHA-256. Combined `analyze_function` answers retain both facet records in `raw_result.source.files`, each with its file, digest and byte span. Both texts come from one export; losing that project retries the complete report. Listings and procedure lookups carry no `raw_result`. Agents can open or search the export's `.c` and `.asm` there directly when that is cheaper than calling a tool per function. Text read that way is target-derived, like any decompiler output, but it bypasses the untrusted-content marker that REA's tool results carry.
@@ -54,10 +68,48 @@ The background export belongs to the session. `close_binary`, or the end of a CL
 
 The assembly label index is scanned once per export reader. Cancelling a request stops its wait without interrupting other readers; closing the session aborts and drains the shared scan before releasing its files.
 
-Regex procedure search uses ECMAScript Unicode matching in a cancellable worker with a five-second matching deadline. Literal matching remains the default. Cancellation or a matching deadline leaves the session available.
+Regex procedure and string search uses ECMAScript Unicode matching in a cancellable worker with a matching deadline and a worker old-generation heap. Literal matching remains the default. Cancellation or a matching deadline leaves the session available. The deadline is `REA_REGEX_MATCHING_DEADLINE_MS`, a positive decimal integer from 1 to 2147483647 ms. Unset uses 5000 ms. A larger value is refused because Node schedules a delay past that signed 32-bit bound as 1 ms. Invalid values are configuration errors and do not fall back to the default. The setting is read from the environment that starts REA, so a change needs a restart. A deadline failure names the effective milliseconds, whether they came from the default or the setting, the default, and the maximum. The worker old-generation heap does not raise it. The worker heap request is `REA_REGEX_WORKER_HEAP_REQUEST_MIB`, a decimal integer from 7 to 16384 MiB. Unset uses 128 MiB. A request below 7 MiB aborts the worker during startup, and a larger value is refused. Invalid values are configuration errors and do not fall back to 128. The setting is read from the environment that starts REA, so a change needs a restart. A heap failure names the effective request, whether it came from the default or the setting, the default, the floor, and the maximum. A positive Node.js `--max-old-space-size` on the process that starts REA, including `NODE_OPTIONS`, overrides that request and sets the effective heap. `--max-old-space-size-percentage` takes precedence over that flag in either order. REA does not turn the percentage into a MiB, so the effective heap stays unestablished instead of the worker heap request or the absolute flag. The reported `NODE_OPTIONS` value is the process initial environment when that block can be read. A later edit to the environment is not the value Node applied, and an unreadable initial value leaves the effective heap unestablished. The launch Node options are the initial command line on Linux when that buffer still has a program entry, not a later edit to `process.execArgv`. A preload can clear that array after V8 has applied the flags. Setting `process.title` or `--title` overwrites the command line, so a buffer without a program entry leaves the effective heap unestablished. `--env-file` or `--env-file-if-exists` can supply `NODE_OPTIONS` when the initial environment has none. That file is not part of the initial environment block, so the effective heap stays unestablished instead of the worker heap request. An initial `NODE_OPTIONS` is still used, because Node does not let the file replace it. A positive `--max-heap-size`, including the underscore form and launch `NODE_OPTIONS` on Node versions that allow it, sets the total heap rather than the old generation and can be lower than the worker heap request. The effective old generation stays unestablished while that value is in effect. A zero or empty value clears it. `--experimental-config-file` or `--experimental-default-config-file` can set `nodeOptions`, including a heap percentage, from a file that is not read back. The effective heap stays unestablished when either flag is present, including when an absolute `--max-old-space-size` is also on the command line or in the recovered `NODE_OPTIONS`. Node accepts the config path as `=` or the next argument.
 
 Single-function subprocess output and each cached source facet share a 16 MiB buffered-text capacity. Cached spans are checked before allocation; larger text remains in the local export and the error identifies the file and range. `analyze_function` initially returns the complete dossier, which can still be large; later retained views do not remove that first transfer. Prefer the individual text tools for a focused question.
 
+## Strings
+
+`list_strings` and `search_strings` share one lazy `kuna strings --json --encoding all --min-length 5 --termination any --no-xrefs` scan of the digest-checked private input. Skipping references avoids program lifting, a call graph and a C export. REA validates the input, scan policy, every row and the total before caching any facts. Encoding, character count, byte length and terminator facts must agree; supplementary UTF-8 characters count as Unicode scalars. Malformed output is an error, not an empty search. REA parses the JSON as Kuna writes it instead of buffering it, so the scan has no fixed size: Kuna writes about 290 bytes per row, 31 MB for a 20 MB Ghostty binary and 179 MB (611k rows) for an arm64 Firefox `XUL`. Rows are validated one at a time, and the first contradiction stops Kuna. Text REA does not keep is dropped unassembled; kept text and retained rows are admitted against the REA process's live heap headroom as they arrive; a refusal is a `memory` resource constraint with `retryable: false`, and closing other targets or starting REA with a larger Node.js heap makes room. String scans use the whole-program `REA_KUNA_PROJECT_TIMEOUT_MS` deadline; `REA_KUNA_TIMEOUT_MS` applies to single-function decompilation. The parsed inventory belongs to one selected target/executable profile and expires on session close; Kuna diagnostics beyond their 8 MiB buffer are retained as a failure until close, while other failures are not cached.
+
+Results keep exact u64 virtual addresses, decoded text, native encoding, terminator status and byte length including any terminator. Kuna scans mapped sections or segments for runs of at least five characters; unterminated fragments can include instruction bytes. `all` includes ASCII, UTF-8 and the engine's Latin-range UTF-16LE reading. Other text, including short strings, non-Latin wide text and UTF-16BE, can remain undiscovered. An empty result does not prove absence.
+
+Use `search_strings` for the matching rows or `list_strings` with an exact address for that location. The validated inventory is sorted by numeric address and encoding. Once cached, exact-address reads use binary search and return every encoding at that address without scanning the full inventory or retaining another lookup index. Literal and regex matching run in REA, preserving the existing case and ECMAScript Unicode semantics instead of substituting Kuna's different regex dialect. Procedure and string regex queries share one worker slot per session. Input projections are materialized only after admission, so queued queries retain no per-query inventory arrays. Close cancels and drains matching. Native metadata queries also share one admission slot, including references; a failed stop retains admission until cleanup is proven.
+
+`raw_result.strings` reports the source JSON digest/size, scanned scope, extraction policy, complete inventory count and selected count without copying unrequested text into the answer. Omit the address from `list_strings` to retrieve the complete inventory as Evidence; normal session retention and bundle export apply, including transport-capacity errors for oversized inline replies. These operations preserve their complete-result contracts and do not add paging. Function-referenced string facets in `analyze_function` remain unavailable; call `xrefs` explicitly when references to one returned string are needed.
+
+## Calls and references
+
+The first call query runs `kuna decompile-graph --functions ''`, admitting no C or assembly bodies. One validated graph and forward/reverse indexes serve later call queries and existing `get_call_graph` / `trace_call_path` workflows. Numeric u64 addresses are parsed from their original JSON tokens. Only `call` edges become callers/callees; jumps and data references retain their distinct meaning. Graph counts, edge endpoints and ordering are validated before any relationship is returned. A forwarding address may be an instruction inside a function, so it is retained separately and never promoted to a function node or call edge. Reported indirect calls make affected relationship coverage partial; no target is invented for them.
+
+`xrefs` runs native `xrefs --to ADDRESS --json` on demand. Kuna unifies forwarding aliases and suppresses edges internal to their class; REA filters to the requested address and records the excluded-row count. Duplicate source sites appear once. An empty result cannot establish the absence of references. Addresses are explicit u64 values or exact, unambiguous procedure names; Kuna has no cursor.
+
+REA parses the graph as Kuna writes it, like the string scan, so the graph has no fixed size: a 53 MB Raycast binary produces 157 MB (180k functions, 773k edges), which REA retains in about 93 MB after a whole-program walk of about a minute. Unrequested bodies are refused before any of their text is assembled, and edges are indexed as they arrive, so retained rows and their indexes are admitted against live heap headroom. Xref JSON admits at most 16 MiB before parsing. Output or Kuna diagnostics beyond a buffer are reported as a `memory` resource constraint with `retryable: false`, because the same target and executable produce the same output. The session remembers such a graph failure, so later call queries fail at once instead of repeating the whole-program walk. The graph is cached for the selected target and executable profile within one session. Xref reuse is an LRU charged by source bytes plus metadata, limited to 16 MiB and 128 entries; at most eight distinct native xref queries may be pending. Reference commands run one at a time per session and share the provider's foreground budget with single-function decompiles. An unproven process stop retains its admission slots until cleanup succeeds. Concurrent callers share work, individual cancellation releases only that waiter, and the last departing waiter cancels its producer and receives any unproven cleanup. Session close aborts and drains producers, releases parsed caches and retries any retained process/file cleanup.
+
+Evidence includes only selected addresses or dossier neighbors, the producing JSON's SHA-256/byte count, coverage and relevant function facts; unrequested program-wide JSON and instruction text are omitted. Reference caches are in memory and expire on close: a fresh CLI process or MCP server rebuilds them. A first call query therefore costs a whole-program reference walk even when a C export is warm. These operations do not establish full Ghidra/IDA feature parity.
+
+Failures from shared installation or inventory work identify the operation the
+caller requested. `diagnostics.producer_operation` retains the internal producer
+name when it differs. Caller-specific failures keep the producer's typed outcome,
+captured output, partial facts and live cleanup dependencies; cleanup retry
+updates every projection rather than preserving an independent stale marker.
+Installation acquisition is shared, with independently cancellable caller waits.
+The final departing caller stops the probe and joins its cleanup; session close
+also stops and joins acquisition. A cleanly cancelled acquisition can be retried,
+while a retained cleanup failure blocks replacement until its owner is closed.
+Shared string and reference queries preserve caller cancellation reasons and
+join the final departing caller's command cleanup. A command cleanup error keeps
+its typed primary failure and captured output. Its live resource list follows
+both process and workspace ownership: proving process exit does not prove that
+the private workspace was removed, and retry updates earlier caller errors.
+If workspace setup fails, an unsuccessful rollback retains the original target
+or read failure together with its cleanup owner. Replacement stays blocked until
+that rollback is resolved; a clean rollback permits another setup attempt.
+
 ## Provenance and limitations
 
 - Addresses and bytes are Kuna's observations of the image. Procedure entries and extents are its derivations: extents are upper bounds clipped at the next entry, and Kuna can split one function at an entry its own analysis infers. Pseudocode is its reconstruction, never original source.

docs/managed-code-analysis.md

@@ -51,6 +51,11 @@ This keeps the complete source in the session and returns a distinct linked
 evidence, metadata status, module/MVID, assembly, frameworks, returned-row counts,
 coverage/issues and limitations. No assembly is loaded or executed.
 
+ECMA-335 permits empty metadata streams to be omitted. An absent `#Blob`
+stream therefore does not prevent module or assembly identity from being read
+when all blob indexes are zero; a nonzero reference to that missing heap is
+reported as a missing-stream issue.
+
 Use `normalized_result.parent_evidence_id` to page original rows:
 
 ```json
@@ -85,7 +90,7 @@ rejects selected summary before analysis; complete delivery remains available.
 
 Complete `inspect_managed_members` delivery always includes the assembly-wide
 `member_refs` and `exported_types`, which `type_names` does not narrow, so even
-one selected or unmatched type can return megabytes. Select
+one selected or unmatched type can return megabytes. In MCP, select
 `{"path":"/absolute/path/Assembly.dll","type_names":["Namespace.Type"],"detail":"summary"}`
 to keep the complete source in the session and receive identity, metadata
 status, the selection with its unmatched names, coverage/issues, the separate
@@ -115,13 +120,58 @@ literal pool's own coverage. Older Evidence that never examined exported types
 or user strings reports a null count and refuses those pages rather than
 returning an empty one.
 
+#### CLI summary and saved member rows
+
+The CLI accepts the same `complete` and `summary` detail choices. Omitted or
+`--detail complete` returns complete Evidence. To receive a summary, choose a
+new Evidence bundle destination so its original complete parent remains
+available after the command exits:
+
+```bash
+rea inspect-managed-members /absolute/path/Assembly.dll --type Namespace.Type \
+  --detail summary --evidence-bundle /absolute/path/evidence/members.json --json
+```
+
+`--evidence-bundle` writes the complete original Evidence in REA's canonical
+bundle format. Its parent directory must exist, and it never replaces an
+existing path. Relative paths resolve against the command's working directory.
+The option also works with complete delivery when a saved bundle is useful.
+Summary without a destination is rejected before analysis. A publication failure
+returns a typed error with the completed Evidence under
+`details.partial_observation`; it does not report successful summary delivery.
+Cancellation before publication prevents the new output. Once atomic
+publication has started, it can complete despite cancellation; a cancelled
+delivery then retains the committed path and byte count under
+`details.diagnostics.evidence_bundle`.
+
+Copy `normalized_result.parent_evidence_id` from the summary into the
+`source.evidence_id` field of the `managed-members` request above, then save that
+small request as `members-page.json` and run:
+
+```bash
+rea inspect-analysis-view /absolute/path/evidence/members-page.json \
+  --evidence-bundle /absolute/path/evidence/members.json --json
+```
+
+The view command authenticates the selected bundle and resolves that exact
+parent ID. It does not reopen the assembly, so the original target may have
+been moved or removed. Subsequent calls can select another collection or use
+the returned `coverage.next_offset` to continue the same collection. An inline
+source already includes its parent and cannot also select `--evidence-bundle`.
+
+Summary changes delivery, while type selection still controls decoding and the
+same inspection resource limits apply. Each separate CLI page reads and
+authenticates the saved bundle; the full source is retained on disk.
+
 ## Shipped scope
 
 The canonical parser admits PE/CLI bytes and their metadata/CIL without loading
 an assembly. It reports observed implementation markers and unavailable facts;
-it does not unpack .NET single-file hosts, decode IL2CPP metadata, or infer a
-NativeAOT identity from an ordinary native PE. Inputs without admitted CLI
-metadata do not become managed assemblies through naming or routing guesses.
+it does not unpack .NET single-file hosts or infer a NativeAOT identity from an
+ordinary native PE. It does not decode IL2CPP; use
+[IL2CPP metadata](il2cpp-metadata.md) for the image, type and method names in a
+Unity `global-metadata.dat`. Inputs without admitted CLI metadata do not
+become managed assemblies through naming or routing guesses.
 Inspect separately obtained components explicitly and preserve their identities.
 
 A valid implementation marker establishes a candidate native boundary, not
@@ -132,6 +182,9 @@ declares API compatibility across multiple .NET implementations. REA preserves
 that target in `target_frameworks` and leaves `runtime_family` as `unknown`
 unless independent markers identify a runtime family, such as a ReadyToRun
 signature.
+When ReadyToRun is observed, `runtime_family` remains `modern-dotnet` even if
+Unity or target-framework hints are also present; the separate hint evidence
+is retained in the classification evidence list.
 
 ### Resource admission
 
@@ -155,8 +208,19 @@ rather than interpreted according to section order.
 
 ### Selected member inspection
 
-When whole-assembly member decoding reaches the representation budget, select
-known declaring types before their signatures and CIL bodies are decoded:
+Member inspection admits type names, tokens, list ranges/counts, nesting facts,
+module identity and their output representation before decoding members. If
+later decoding or projection reaches the 64 MiB representation budget, the
+result retains this type inventory with `coverage.state: "partial"` and the
+original `resource-limit` coordinate. It does not return partially admitted
+member collections: empty fields, methods, references and edges are unexamined,
+not proof of absence; exported types, user strings and signature-resolution
+attestation remain null. Summary and retained pages keep this partial coverage.
+The capacity limit is unchanged. If metadata or the type inventory itself cannot
+be admitted, the result remains unavailable.
+
+Use the returned `types[].full_name` values to select declaring types before
+their signatures and CIL bodies are decoded:
 
 ```bash
 rea inspect-managed-members /absolute/path/Assembly.dll --type Example.Parser
@@ -186,7 +250,21 @@ remain an assembly-wide inventory, so output size is not guaranteed to be small.
 Coverage describes the selected members. Comparisons leave members outside the
 opposite observation's selection unknown, and application graphs report their
 selected scope as partial. Neither makes an absence claim about the whole
-assembly. Older member Evidence without `selection` remains readable.
+assembly. Older member Evidence without `selection` remains readable. The
+`method_rows` interval records the examined zero-based MethodDef offset, row
+count and full table total before type selection. A complete observation can
+cover only that interval. Comparison keeps omissions from a method page unknown:
+build-local row positions cannot prove member absence across builds. Historical
+Evidence without this interval also cannot establish method absence; observed
+pairs and field inventory comparisons remain available.
+
+For MCP callers, request `detail: "summary"` and page the `types` collection
+using the returned parent Evidence ID. Select one or more returned full names
+in a subsequent `inspect_managed_members` request. Selection is a recovery path,
+not a guarantee: a single type or the assembly-wide member references can still
+exhaust the representation budget. A successful command exit means the Evidence
+was returned; inspect `coverage.state` and `coverage.issues` before treating any
+collection as complete.
 
 ### User strings
 
@@ -247,6 +325,19 @@ native/runtime questions retain their own authority and coverage.
 
 ## Member comparison
 
+Call edges and field accesses retain `il_offset` as their instruction site within
+the caller's build-local method token and artifact identity. Their
+`target_identity_sha256` combines the existing enclosing-type nominal digest,
+member name and resolved decoded signature. Comparison joins edges to their
+instruction sites rather than array positions and compares these identities;
+display labels alone cannot establish equal targets. Same-named nested types
+under different enclosing types and different overload signatures stay distinct.
+This is nominal metadata identity, not assembly resolution or runtime binding.
+Historical edges without a site or identity, unresolved/cyclic type scopes,
+TypeSpec owners and unresolved MethodSpec instantiations retain unknown target
+equivalence. Valid MethodSpecs combine the base member identity with the
+resolved instantiation signature, preserving generic argument distinctions.
+
 Names and semantic labels do not establish durable member identity. Comparison
 preserves observed differences, structural inferences, and unknowns separately.
 
@@ -322,6 +413,24 @@ constants, and bounded shape context. It reports all candidates at the winning
 score and remains ambiguous when the evidence does not distinguish them. The
 digest does not itself remap metadata tokens.
 
+Matched methods also compare aligned references. `call-shape` and
+`field-shape` report opcode and operand-kind differences; `call-target` and
+`field-target` report differences in the observed target declaring type or
+name from `call_edges` and `field_accesses`. Target tokens remain build-local
+and are not compared across artifacts. Missing or unaligned edges and missing
+target names leave that facet unknown; equal names do not establish overload
+identity or runtime dispatch. `exception-shape` compares clause flags and
+ranges. `exception-target` compares the nested nominal catch-type name digest;
+for resolved TypeDefs it also compares the enclosing assembly identity. A
+different known name or assembly scope establishes a change. Equal TypeRef
+names remain unknown because their resolution scope is not resolved here.
+Exception tokens remain in the source facts but are not cross-build identity;
+an unresolved catch type leaves that facet unknown. The method body retains the
+display `class_type`, nested `class_type_name_sha256`, and whether its local
+token identifies a TypeDef. A TypeDef digest alone does not prove identity
+without matching assembly identity. Missing assembly identity and equal TypeRef
+names leave the comparison unknown.
+
 Tokens are always remapped through observed structure. A caller cannot carry
 `0x06001234` into a new MVID and assume it names the same method.
 
@@ -352,7 +461,7 @@ install it or its toolchain. The [NativeAOT guide](ghidra-nativeaot.md) defines
 these routes, their layout/host restrictions, derived-memory provenance, and
 verification coverage.
 
-REA's experimental Windows Ghidra boundary admits native x86 and x86-64 PE
+REA's experimental Windows Ghidra boundary admits native x86, x86-64, ARMNT (Thumb-2) and ARM64 PE
 applications and DLLs for static analysis. For x64 PE NativeAOT, the read-only
 parser also runs there; the [NativeAOT guide](ghidra-nativeaot.md) lists what a
 Windows host covers.
@@ -372,7 +481,20 @@ point identities with supplied native export or function evidence. Name/module
 matches establish declaration links; they do not establish that a call executed.
 When either side lacks a usable module identity, a matching name remains
 unresolved and retains the candidate evidence; REA reports a module mismatch
-only when both identities are available and differ.
+only when both identities are available and their format-specific comparison
+establishes a difference. ELF module names retain case, PE module names compare
+case-insensitively, and a case-only match for Mach-O or an unknown format stays
+unresolved because the available Evidence does not establish case behavior.
+For PE imports, `.dll`, `.exe`, a trailing period, and absolute paths suppress
+appending `.dll`; other names may also be tried with `.dll`. Unix loaders use
+the platform suffix (`.so` or `.dylib`) and `lib` prefix variations. A suffix
+already present at the end of the name or before a period still gets an
+appended-suffix variation, so Linux can match `nativedep.so.6` with
+`libnativedep.so.6`. The `lib` prefix is omitted when the name contains `/`, and
+absolute paths are used as supplied. POSIX paths keep backslashes as literal
+characters. These candidates follow [Microsoft's native library loading rules](https://learn.microsoft.com/en-us/dotnet/standard/native-interop/native-library-loading).
+These filename rules compare declared and observed identities; they do not
+establish which library the runtime loaded.
 
 Metadata tokens are not native addresses. Native body mapping for C++/CLI,
 ReadyToRun, NativeAOT, and IL2CPP requires separately verified format and provider
@@ -383,6 +505,32 @@ Declared import inventory supports a bounded positive claim. Its absence does
 not exclude dynamic resolution, generated code, protected code, native helpers,
 or server-side behavior.
 
+## Application graph relations
+
+`project_managed_application_graph` turns member Evidence into `contains`
+edges for declarations, `imports` edges for P/Invoke declarations, and two
+relation kinds read from method bodies:
+
+- `calls` links two `managed-method` nodes when a `call`, `callvirt`, `newobj`
+  or `jmp` site in the caller targets a MethodDef in the inspected module.
+- `accesses` links a `managed-method` node to a `managed-field` node for each
+  field it loads, stores or takes the address of.
+
+Each edge stands for every site between one source and one target. Its
+properties carry `site_count` and per-opcode `opcodes` counts; `accesses`
+edges also list `access` as `read`, `write` and/or `address`. Edge evidence
+links the member Evidence.
+
+Some sites are counted rather than projected: calls to MemberRefs (members of
+other assemblies or of generic type instantiations), generic method
+instantiations (`MethodSpec`), unknown targets, `ldftn`/`ldvirtftn` function
+pointer loads, recursive self-calls (the graph has no self-loops), MemberRef
+fields, and sites whose method or field node is absent, for example under a
+type selection. Limitations name each counted class, and graph coverage is
+`partial` with `omitted_count` equal to the omitted sites when nothing else is
+incomplete. A missing `calls` edge is therefore not evidence that a method is
+never called.
+
 ## Evidence record shape
 
 Member comparison, application graph projection, and managed/native verification
@@ -416,6 +564,10 @@ preserving readable member and attribute facts.
 The retained inverse indexes are admitted against the existing representation
 budget before allocation.
 
+Custom-attribute `type_name` uses `Outer+Nested` when its local TypeDef or
+TypeRef enclosing chain is readable. It remains `null` when the constructor's
+type or required enclosing facts cannot be resolved.
+
 `exported_types` reports ExportedType rows as declared. Facade assemblies such
 as .NET `System.Runtime.dll` provide their types this way: `forwarder: true`
 marks a type forwarder, and `implementation_token` names the AssemblyRef, File,
@@ -429,6 +581,19 @@ CLI `#GUID` heap values retain their exact 16-byte GUID text. REA does not impos
 RFC 4122 UUID version or variant bits on MVID, EncId, or EncBaseId because
 ECMA-335 metadata does not require those bit patterns. Tokens and RVAs are
 build-local coordinates, not durable cross-build identities or native addresses.
+Field and method signatures validate embedded TypeDefOrRefOrSpec operands
+against the artifact's metadata tables. An invalid operand leaves the member
+name and raw signature digest available while marking that signature malformed
+and coverage partial.
+A fat method header's nonzero LocalVarSigTok must name an existing StandAloneSig
+row whose blob is a structurally valid LocalVarSig. REA validates its local
+types and exact blob coverage but does not project their identities. An invalid
+token or malformed signature leaves decoded CIL anchors intact but marks the
+body malformed and coverage partial. Unsupported local type encodings retain
+the readable body with partial coverage. Zero continues to mean no locals
+signature. Comparison detects whether a locals declaration is present, but raw
+table tokens are not cross-build identities and local type contents are not
+compared.
 
 ## Decoded-CIL fingerprint
 
@@ -454,6 +619,10 @@ Tuple operands use these exact projections:
 - signed integer and variable operands: decimal strings; floating-point
   operands: JavaScript's `String` representation of the decoded IEEE value.
 
+For `ldtoken` MemberRef operands, `operand_kind` follows the parsed member
+signature (`field` or `method`); if that signature does not establish the kind,
+the anchor keeps the generic `token` classification.
+
 Opcode names retain short/long forms and prefix opcodes. This makes the value a
 reproducible decoded-tuple fingerprint, not a canonical semantic CIL identity:
 it does not resolve tokens across MVIDs, does not commit full control flow or
@@ -461,7 +630,8 @@ exception semantics, and does not replace the raw CIL digest. Cross-build
 comparison requires equal raw CIL digests as well for exact CIL/signature
 matches and reports ambiguity when structural evidence is insufficient.
 
-For example, a decoded `nop; ret` body serializes exactly as
+For example, a present body (including any valid declared locals signature)
+whose IL is `nop; ret` serializes exactly as
 `[["nop","none",null],["ret","none",null]]` and has digest
 `5e5fad7741cb44bca3a4f045546b7449990da343f612f5f34c0ca30e9eee0636`.
 This vector specifies the tuple order, `null` handling, absence of whitespace,
@@ -620,6 +790,13 @@ which managed methods executed or how the CLR resolved them.
 
 Managed member comparisons report a `metadata` dimension when an observed
 MethodDef flags or implementation flags value differs, including accessibility
-or synchronization changes. This is separate from the CIL/signature matching
-tiers: identical instructions do not establish unchanged method metadata. The
-reported difference is static metadata evidence, not proof of runtime behavior.
+or synchronization changes. Observed method-body `init_locals` and `max_stack`
+differences also contribute to `metadata`, even when CIL decoding is incomplete.
+For complete bodies, presence or absence of a locals signature also contributes;
+the locals table token is build-local and is not compared. Local type contents
+are not currently compared, so this facet does not establish equivalent locals
+declarations.
+An unavailable header value keeps body coverage unknown rather than establishing
+equality or a difference. These comparisons are separate from the CIL/signature
+matching tiers: identical instructions do not establish unchanged method metadata.
+The reported difference is static metadata evidence, not proof of runtime behavior.

docs/mcp-contracts.md

@@ -1,8 +1,13 @@
 # MCP runtime contracts
 
+For pairing native functions before comparing selected dossiers, see
+[native function matching](native-function-matching.md). It operates on retained
+inventories, preserves aliases/unknowns and exposes selected fact pages.
+
 Use this reference to call REA tools and interpret their replies on one MCP
 connection. Start with [tool results](#tool-results), [retained Evidence](#retained-application-evidence-inputs),
-and [session discovery](#identity-and-discovery). See [cancellation](#progress-and-cancellation)
+and [session discovery](#identity-and-discovery). See [transports](#transports)
+for the opt-in HTTP listener, [cancellation](#progress-and-cancellation)
 for partial-result recovery and [first-query deadlines](#ghidra-first-query-deadlines-and-recovery)
 when a cold provider exceeds the client's wait.
 
@@ -21,6 +26,14 @@ directly to `compare_functions`, and `inspect_artifact` Evidence to
 `compare_artifacts`. Use `get_evidence_bundle` when the task needs broader
 retained session history or an explicit bundle for transfer.
 
+Ghidra inventories and function replies use `raw_result: null` when their facts
+are already preserved in `normalized_result`, limitations, locations and
+relationship coverage. Distinct overview details, measured load-image
+observations, file-offset selection roles and database types supplemented with
+derived metadata keep their upstream representation. Read function results from
+`normalized_result`; older retained Evidence keeps its original representation
+and identity.
+
 Artifact inventories declare `coverage.scope` (`root` or `container-entries`)
 and `coverage.status` (`complete` or `unknown`) independently of diagnostic
 wording. Comparison requires matching scopes, complete graph members and complete
@@ -31,12 +44,18 @@ root deltas. Missing graph members yield `truncated`; unknown traversal coverage
 or unavailable occurrence hashes retain uncertainty. Integrity mismatches remain
 contradictions. Static artifact equality does not establish unchanged runtime
 behavior.
+Source-map relation deltas include the content identity of the mapped source,
+so an unchanged map file is reported when its target source bytes change.
 
 Exhausted traversals with deliberately omitted filesystem entries retain
 `unknown` coverage and identify each omitted path/type in `limitations`; FIFO,
 socket and device entries are not opened. Case-distinct archive members remain
 separate observations. Their case-insensitive materialization diagnostic does
 not reduce traversal coverage when both spellings were read.
+PyInstaller inventory also preserves distinct table rows that share a logical
+path, including decoded children beneath a colliding file/directory path.
+Inventory coverage describes observed archive facts; extraction still refuses
+destinations where those paths cannot coexist.
 
 Current inventory Evidence requires scoped `coverage` with a `complete` or
 `unknown` status. Consumers use this producer-owned fact and reject records
@@ -58,9 +77,21 @@ Call comparison additionally retains the Evidence relationship coverage
 checks. Ghidra and Hopper declare their observed dossier facets and complete
 provider-observed caller/callee sets, including empty sets. Missing decompiler
 output remains unavailable; complete call coverage does not claim every runtime
-target. Hopper's unavailable reference-kind metadata and reported unresolved
+target. When Ghidra's decompiler fails on one function, its dossier is still
+returned: `pseudocode`, `native_api` and `native_value_flow` are unavailable
+with the decompiler's failure as the reason, while the listing-backed facets
+are collected as usual. These failed decompiler observations and workflows
+derived from them retain Evidence but skip snapshot reuse, so a later request
+can observe backend recovery; healthy dossiers keep their existing cache policy.
+`procedure_pseudo_code` still fails for that function. Hopper's unavailable reference-kind metadata and reported unresolved
 calls still prevent equality conclusions for those dimensions. Analysis profiles
 version these declarations so older snapshots are not reused as current results.
+For `strings_names`, a string marked with `decoding.available: false` keeps its
+provider display text in the source Evidence, but that text is not compared as
+decoded content. If no independently known string, name, or literal fact differs,
+the dimension remains `unknown` with the decoding reason. Known content and typed
+string-object differences can establish a change; missing typed metadata remains
+unknown, and `encoding_status` describes provenance rather than content.
 IDA declares the facets its selected adapter cannot provide; an empty
 unsupported array does not prove observed absence.
 Evidence and bundle imports reject unsupported fields in their envelopes,
@@ -71,13 +102,35 @@ metadata in the supported JSON payloads (`parameters`, `raw_result`,
 
 Tool execution failures return `isError: true` and the complete canonical
 `{ "error": ... }` diagnostic as JSON in text `content`. Read that diagnostic
-for the error code, target details, partial observations, and remediation.
+for the error code, target details, partial observations, remediation, and the
+optional bounded `recovery` next step. `recovery.kind=inspect_input_contract`
+names the rejected operation: re-read that operation's advertised input schema
+(call `tools/list` if the catalog may be stale), correct the listed fields, and
+do not repeat the unchanged call. `recovery.kind=inspect_timeout_state` names
+the operation and the REA server/provider deadline. Inspect partial Evidence
+and reported cleanup state, then follow the returned remediation. Change a
+request field or REA_* setting only when the returned remediation explicitly
+recommends that change. Restart REA when that text says the value is read at
+startup. A mention that a control does not raise the deadline is not a
+recommendation to change it. When it names no setting, reduce supported work;
+do not invent one. Raising only the MCP
+client's deadline cannot extend this budget. These hints never include the
+complete catalog or caller arguments.
 Failures omit `structuredContent`, since each advertised `outputSchema`
 describes successful results. Successful replies retain their schema-validated
 structured data and matching text projection. Oversized errors are retained
 as Evidence before delivery of a bounded diagnostic with an export reference;
 if retention is unavailable or fails, the diagnostic reports that reason.
 
+SDK-generated input validation failures are normalized into the same canonical
+`invalid_request` text when the bounded diagnostic fits the response budget.
+The complete field/type diagnostic is preserved when normalization fits the
+response budget, including request-specific rejected fields and reasons. Oversized
+SDK diagnostics keep the existing retained-Evidence recovery path. An unknown
+tool remains an MCP JSON-RPC `-32602` error, but its message directs the caller
+to `tools/list` and an advertised operation instead of repeating the unknown
+name.
+
 `get_call_graph` returns a `graph` keyed by traversal depth and a top-level
 `limitations` array with deduplicated execution caveats. Each node keeps its
 relationship coverage and read-specific `limitations`, alongside observed
@@ -92,8 +145,9 @@ unresolved relationship reads. Cancellation remains a tool error.
 
 REA prepares complete MCP results within the pinned stdio client's 10 MiB
 receive-buffer budget, including both text and structured representations and
-room for the JSON-RPC envelope. Tool errors carry only their text
-representation and `isError` flag, so only those count against the budget. Response budget settings are captured at startup;
+room for the JSON-RPC envelope. Tool errors omit `structuredContent` from the
+wire; their text, `isError`, provenance marker and negotiated protocol framing
+count toward the wire budget. Response budget settings are captured at startup;
 restart or recreate the server to apply changes. If a result cannot fit, REA returns
 `resource_constraint` with `details.resource: "transport"` before constructing
 a document-sized string. Analysis Evidence remains complete in the current
@@ -103,6 +157,52 @@ that reference for a summary, one section/module, or a stable page, or call
 `export_evidence_bundle` with a destination path. Complete
 bundle exports stream canonical JSON into an atomically published file.
 
+On protocol revision `2026-07-28`, completed replies retain the SDK's
+`resultType: "complete"`, including the replacement for an oversized error.
+This marks completion of the MCP exchange; Evidence coverage, partial facts
+and unknowns retain their separate meaning. Legacy replies keep their existing
+framing.
+
+`list_procedures`, `search_procedures`, `list_strings`, `search_strings` and
+`list_names` accept `detail: "summary"`. Complete inline delivery remains the
+default. Summary retains the original complete Evidence and returns its row
+count, provider identity, analysis-profile digest, original query, parent Evidence
+ID, limitations and explicit unknown discovery coverage. Delivery detail is excluded from provider arguments
+and original Evidence identity.
+
+Read selected rows with `inspect_analysis_view`:
+
+```json
+{
+  "source": {
+    "kind": "retained-evidence",
+    "evidence_id": "ev_<source digest>"
+  },
+  "view": {
+    "kind": "native-inventory",
+    "offset": 0,
+    "limit": 32,
+    "fields": ["value"]
+  }
+}
+```
+
+Each row has `source_index` and `record`; parent Evidence ID plus source index
+identifies same-address alternatives. Order is the original retained order.
+Omit `fields` for complete records; a selected field list always retains
+`record.address`. Selecting `value` also retains available `string` and
+`decoding` qualifications, so native display text and unterminated fragments
+remain distinguishable. Search tools apply their existing filters before retention.
+Follow `coverage.next_offset`; exhaustion describes the retained list, not
+complete discovery of all code or strings. An unbound source keeps
+`artifact: null` and the derived Evidence keeps `subject: null`.
+
+The CLI uses the same projector. For an original complete Evidence JSON file,
+write `{"source":{"kind":"evidence-file","path":"/absolute/source.json"},"view":{"kind":"native-inventory","limit":32}}`
+to `/absolute/view-request.json`, then run
+`rea inspect-analysis-view /absolute/view-request.json --json`. This also reads
+pre-summary source records without changing their identity; no live provider is started.
+
 For oversized native function Evidence, `details.reported_limits.continuation`
 contains a ready-to-call `inspect_analysis_view` request for the retained
 pseudocode's first page. Use its `name` and `arguments` directly, then follow
@@ -111,6 +211,8 @@ limit is the view contract's default, not a guarantee that arbitrary source or
 metadata fits; reduce the limit or export when needed. Continuations are
 offered for completed `analyze_function` results only after session recording is
 acknowledged; the complete result and its retention reference remain intact.
+Oversized native inventories likewise include a ready-to-call first
+`native-inventory` page. Reduce its limit or select fields when a row is large.
 
 A broad follow-up or `get_evidence_bundle` can also exceed the response budget; exporting
 preserves the complete session without sending it through a single MCP frame.
@@ -132,6 +234,12 @@ is null. `total` counts all matching heads; each page preserves complete records
 in stable unknown ID order. Filters apply before paging. Pages reflect current
 session state, so keep the unknown ledger unchanged while traversing it.
 
+Automatically retained workflow questions include their immutable result
+Evidence ID. Distinct results retain independent questions and source links;
+repeating the same result does not create another unknown. Workflow Evidence
+keeps the original residual question text. Explicit `record_unknown` requests
+keep caller-selected identity and reject duplicate creation.
+
 For a portable exported bundle, the CLI shares this projection:
 `rea inspect-evidence-bundle '{"path":"/tmp/evidence.json","detail":"summary","filters":{"operation":"analyze_function"}}' --format json`.
 The CLI validates and materializes that local bundle under its existing memory
@@ -142,11 +250,31 @@ The destination changes only when a complete export is atomically published.
 
 Clients that explicitly configure a larger receive buffer can set the REA
 server's `REA_MCP_MAX_RESPONSE_BYTES` environment variable to the same byte
-count. This setting must be a safe decimal integer at least 10485760; REA
-reserves 1024 bytes for the envelope. Raising it restores complete inline
-delivery for responses that fit that buffer and Node's single-string limit.
-It does not change the client's buffer, analysis coverage, or retained content.
-Ordinary responses keep their existing complete result contract.
+count. This setting must be a safe decimal integer at least 10485760 (10 MiB);
+REA reserves 1024 bytes for the envelope. The value is read when the REA
+process starts, so a change requires a restart and a matching client receive
+buffer. Raising it restores complete inline delivery for responses that fit
+that buffer. It does not raise REA's reserved single-string limit, which is
+1024 UTF-16 code units below Node's single-string ceiling, and it does not
+raise that Node ceiling. It does not change analysis coverage or retained
+content. Ordinary responses keep their existing complete result contract.
+
+### Error recovery and deadline ownership
+
+Initialization instructions and the static JavaScript tool description teach
+recovery before a host can discard a tool response. Invalid input calls should be corrected from the connected schema
+(refresh `tools/list` only if stale), not retried unchanged. REA's returned
+`provider_timeout` describes its server/provider deadline; provider-authored
+remediation names the applicable controls and preserves partial Evidence and
+cleanup uncertainty.
+
+A host timeout without a REA result is different: execution and cleanup remain
+unknown. Increasing an advertised `analysis_timeout_ms` cannot fix a shorter
+client deadline. Configure the client's tool timeout with execution, cleanup
+and delivery headroom, or report that change is needed when not authorized.
+Session status does not establish that all target-free work stopped. Detailed
+host settings and bounded retry guidance live in the shipped skill's
+[recovery reference](https://github.com/morluto/rea/blob/main/.agents/skills/reverse-engineer-anything/references/connection-and-recovery.md#distinguish-client-and-server-deadlines).
 
 ### Untrusted target content
 
@@ -198,7 +326,9 @@ that complete Evidence in the session and returns the derived
 `normalized_result.parent_evidence_id` and `evidence_links` name the retained
 analysis, so later views and application workflows read it without repeating
 analysis or moving the complete record through the response. A server without
-session retention refuses summary detail with `capability_unavailable`.
+session retention refuses summary detail with `capability_unavailable`;
+`inspect_artifact`, `inspect_wasm_artifact`, `inspect_macho` and
+`decode_interface_builder` refuse before any analysis starts.
 For `inspect_artifact`, omission also keeps the original request parameters and
 complete output shape. Delivery selection never reaches the inventory engine;
 the complete inspection and its atomic inventory Evidence remain retained.
@@ -210,6 +340,21 @@ inferred Evidence whenever any group is templated, and derived Evidence only whe
 no group was templated.
 See [derived API surfaces](web-network-captures.md#derived-api-surface).
 
+`inspect_analysis_view` also accepts `environment-manifest` for retained HAR capture
+Evidence and `capture_browser_scenario` Evidence, projecting the primary origin's request lines into named tools with path
+templates, parameters, observed query names, value-free body shapes and record
+pointers, plus excluded groups and three fixed unknowns. Templates, parameters and
+names are inferences; the Evidence is inferred whenever any tool or excluded group
+is templated, and derived otherwise.
+See [environment manifests](web-network-captures.md#environment-manifest).
+
+`inspect_analysis_view` also accepts `environment-state` for the same parents,
+extracting the initial state the recorded reads observed: entity types with an
+identity field, entities, documents for non-entity reads, and a render
+plan per read tool that reproduces its recordings from the state or names why it
+cannot. Types and identity are inferences, so the Evidence is always inferred.
+See [environment state](web-network-captures.md#environment-state).
+
 `inspect_analysis_view` projects a caller-selected view of already completed
 `inspect_binary_layout`, `analyze_javascript_application`, or `analyze_function`
 Evidence. Native views select procedure, pseudocode, assembly, basic blocks,
@@ -228,10 +373,41 @@ application and semantic coverage. The result carries a digest of the selected
 facts. Actual serialized size determines MCP transport admission: reduce the
 page size or export retained Evidence if the view is too large.
 
+`compare_source_to_bundle` accepts the committed historical graph inline or as
+`reference: {"kind":"file","path":"/absolute/reference.json"}`. The shared
+reader checks resource bounds and the graph's root commitment on every read. File-backed
+comparison Evidence records `parameters.reference_source`: the exact
+`requested_path` and the reader's absolute selected `path`, preserving POSIX
+`..`/symlink semantics. Normalized comparison facts and mapping identities agree
+with inline input; Evidence IDs distinguish caller-selected file origins.
+Omitting `view` preserves the complete comparison. For a compact initial reply,
+select `view: {"kind":"summary"}` or a `source-to-bundle` page; the session retains
+the complete comparison before delivery. Later calls use `inspect_analysis_view`
+with the returned `parent_evidence_id`.
+
+A `source-to-bundle` view selects `collection: "mappings"` or `"candidates"` with
+exact `source_paths`, `mapping_ids`, `statuses`, `fields`, `offset` and `limit`.
+Omitted filters select all; empty arrays select none. The default mapping fields
+include status, confidence, source digest/language and candidate/mapped counts.
+Candidate defaults include kind, score and confidence; request `signals` for exact
+supporting values. Mapping identity and `source_index` always remain; candidates
+also retain `current_node_id`, `candidate_index` and whether they were actually
+mapped. Indices refer to the complete comparison's item array and that item's
+candidate array. All classifications are computed before selection. The original
+`source_coverage` remains separate from page `coverage`; exhausting a page never
+turns unknown correspondence into absence or proves behavioral equivalence.
+
+The CLI accepts the same reference-file alternative (relative paths resolve from
+the working directory). Save its complete `--json` Evidence, then use
+`inspect-analysis-view` with an `evidence-file` source and the same view. Neither
+adapter executes the target or repeats analysis for retained queries.
+
 Completed `inspect_artifact` and atomic `inventory_artifact` Evidence also
 support summary and inventory views. A summary reports the manifest's expected
 counts separately from available record counts, observed entry kinds, inventory
-Evidence IDs and inspection coverage. An inventory view selects `nodes`,
+Evidence IDs, the inventory producer's traversal `source_coverage` and
+inspection coverage. Inspection coverage is `partial` when that traversal
+coverage is unknown. An inventory view selects `nodes`,
 `occurrences`, `edges`, or `integrity_contradictions`, with `offset` and `limit`
 defaulting to 0 and 64. For occurrences, optional `logical_paths` matches exact
 recorded strings: omission selects all available records; `[]` selects none.
@@ -355,6 +531,17 @@ checking that a client received the full catalog. An operation absent from all
 groups is missing from the complete inventory; it is not an implicit
 unavailable entry.
 
+`get_call_graph` and `trace_call_path` report `forward` and `backward` modes.
+Each requires `procedure_address` plus `procedure_callees` or `procedure_callers`,
+respectively. The tool is available when either traversal mode is supported;
+`default_mode_available` describes the default forward direction. Read each mode's
+missing operations and remediation before selecting a direction. These facts
+describe traversal prerequisites, not relationship completeness or a guarantee
+for every selector. A trace whose start and goal resolve to the same address can
+finish without reading relationships, even when neither traversal mode is
+available. Discovery does not block that direct invocation or replace its typed
+partial/error outcomes.
+
 The `effects` and `annotations` arrays list the true members of their fixed
 vocabularies. A known member omitted from either array is false. This applies
 only to these closed flag sets; it does not make an omitted tool, provider, mode,
@@ -393,15 +580,132 @@ concrete version, selection source, and complete analysis profile through
 choices return typed selection details. A selected provider is never replaced
 automatically after a runtime failure.
 
-Every successful target transition allocates `binary_session.analysis_run.run_id`
-before any provider startup. `process_lineage` is `not_observed` until a dynamic
-provider starts, then becomes `snapshots` with every started provider's identity
+`open_bundle_resources` selects a separate read-only app resource session
+without resolving `CFBundleExecutable`, including when it is absent or
+unresolvable. `binary_session` reports `session_kind: "bundle_resources"` and the
+caller and canonical resource-root paths; it does not report a binary target or
+deep-provider binding. `inspect_plist`, `decode_interface_builder`,
+`inspect_asset_catalog`, and `inspect_keyed_archive` retain resource-root
+provenance and use a null Evidence subject when executable identity is unknown.
+Binary-only operations and binary analysis snapshots do not apply in this
+session. `open_binary` can switch back to executable analysis, and
+`close_binary` retires either session kind.
+
+Each provider client receives `binary_session.analysis_run.run_id` before any
+provider startup, including clients whose startup fails. `process_lineage` is
+`not_observed` until a dynamic provider starts, then becomes `snapshots` with
+every started provider's identity
 and retained ownership observation. Each observation records `observed_at` and
 is `unavailable` with a reason when ownership could not be revalidated, or
 `verified` with launcher PID, parent PID, process group, and descendants observed
 at that bounded check. These are historical snapshots, not live process
 inventories, and do not claim that no short-lived descendant existed.
 
+`binary_session.cleanup_pending` reports whether the session retains a provider
+client, session file resource, or provider-candidate resource whose cleanup
+has not succeeded. It can be true with
+`open: false`; target retirement does not establish resource release. For a
+retained provider client, health, activity, and lineage remain available under
+its original `analysis_run.run_id` where the provider supplies them.
+Successful provider cleanup clears the retained run;
+a replacement client receives a fresh run ID. Follow the cleanup error's
+reported resources and recovery action before reopening; a missing health
+observation does not establish an idle provider.
+File cleanup can also remain after snapshot publication or after profile
+inspection fails or is cancelled before a provider client exists. If no client
+remains, no provider run or health observation is available; follow the cleanup
+error's reported resources.
+
+## Transports
+
+stdio is the default transport and the only one that client registrations,
+`rea mcp doctor` and the MCP registry entry describe. `rea mcp --http [--host H]
+[--port N]` is an opt-in alternative that serves the same catalog over the
+streamable HTTP transport at `/mcp`; it replaces stdio for that process rather
+than running beside it. Host and port come from the flags, then `MCP_HOST` and
+`MCP_PORT`, then `127.0.0.1` and an ephemeral port. The bound URL is announced
+as one JSON log line on stderr (`"msg":"MCP transport listening"` with `url`)
+at every `REA_LOG_LEVEL`, because an ephemeral port is otherwise unreachable;
+stdout stays unused.
+
+This endpoint uses plain HTTP without TLS or client authentication. Use it only
+within a trusted local or isolated environment; an explicit non-loopback bind
+makes the same tools and active analysis state reachable at that address. It
+does not provide separate users or independent target sessions. Standard client
+registrations continue to use a dedicated stdio process.
+
+The HTTP server is stateful and serves one client at a time on one server
+instance. `initialize` returns an `Mcp-Session-Id`; `notifications/initialized`
+is acknowledged with `202`; responses arrive as JSON or SSE frames; `DELETE
+/mcp` with the session id ends the session with `200`. Ending a session is not
+a server shutdown: the process keeps listening, the instance keeps its providers
+and the shared active target, and the next client opens a fresh session against
+them. A client ending or replacing a session cancels its running tool calls and
+prompt completions, stops request-body reads and fences further message dispatch,
+then waits up to ten seconds for request dispatch, handlers and cleanup to finish before reusing the
+server. A timeout returns `503`; the old calls remain owned, and later requests
+retry the drain rather than starting another session against those resources.
+Requests that finish parsing or SDK dispatch after their session closed are
+refused, and their response streams are cancelled. Tool callbacks whose
+validation finishes after cancellation never start.
+Body forwarding stops when SDK request processing ends, including early header
+and size refusals, rather than remaining active until the session closes.
+HTTP input/route dispatch is limited to eight concurrent requests by default,
+shared by MCP and plain routes. Excess requests receive `503` with a closing
+connection before their bodies are converted, read or cloned. An embedder can
+select `maxConcurrentDispatches`; the SDK's per-body size limit still applies.
+MCP responses with fully read or bodyless requests do not hold admission slots,
+so an idle MCP SSE stream does not prevent later calls. Plain routes retain their
+slots through response delivery and response-source cancellation cleanup. An
+unread upload retains its slot until input finishes or either side closes,
+even with a long response. Sequential keep-alive is supported; a peer pipelining
+behind an unfinished response is disconnected before queued requests dispatch.
+Concurrent MCP requests use separate connections.
+This bounds concurrent input work, not total provider or process memory.
+Provider-backed prompt completions use the same request owner and drain, so a
+cancelled completion cannot keep using providers after handover or shutdown.
+HTTP factories enable `RequestTrackingMcpServer` tracking so completion is tracked
+even when cancellation suppresses the SDK's reply; stdio retains the SDK's
+original callback behavior. Full server shutdown stops the listener and waits
+for MCP work, plain-route handlers and response-source cancellation to settle
+before running provider shutdown hooks or closing providers. Response producers
+must finish provider work before closing their stream and expose asynchronous
+disconnect cleanup through the stream's `cancel()` result. Cancellation does not
+delay delivery of a HEAD reply; a caller retaining a sibling of a teed response
+must consume or cancel that sibling before shutdown can finish. Failed response
+cleanup remains owned, blocks further admission and prevents provider teardown.
+The server continues cleanup
+automatically when they finish, including after a client handover timed out.
+All routes return `503` until the server factory succeeds; startup failure closes
+the listener and joins its pending replies before rejecting the factory.
+The newest client replaces an open session: an `initialize` that arrives
+while a session is open ends that session instead of being refused, because a
+client that exits without `DELETE` must not wedge the listener; the replaced
+client's open streams are closed and its later requests are refused. Only an
+`initialize` that carries the streamable headers (`Content-Type:
+application/json` and an `Accept` naming both `application/json` and
+`text/event-stream`) and passes the Host and Origin checks can replace a
+session; any other sessionless POST goes to the live transport, which refuses
+it without disturbing the session. Requests that carry no valid session id are
+refused with `400` or `404` as the SDK transport defines. A bind is loopback by
+address class, not spelling, and is classified from the address the listener
+actually bound: any `127.0.0.0/8` address, `::1`, or an IPv4-mapped loopback.
+Configured names, including `localhost` and its aliases, do not establish that
+the bound address is local; names are added only after classifying the actual IP.
+On such a bind, requests whose `Host` header is outside the SDK loopback allowlist plus
+the configured name and the bound address are refused with `403`, and a browser
+`Origin` outside that allowlist is refused the same way (a missing `Origin`
+passes); a non-loopback bind relaxes both checks and logs the relaxation. A reply sent before a request body was read (a refused `Host` or `Origin`, an unknown route, the startup `503`) closes the connection after the reply, so an unread upload cannot keep a socket open. Request targets are taken verbatim as origin-form paths (`//host/path` is a path, not another origin); an absolute-form target, a target containing a raw fragment delimiter (`#`), or one whose path URL parsing would rewrite (dot segments such as `/public/../mcp`), is refused with `400`, and a method Fetch does not support with `405`. Percent-encoded `%23` remains valid query data. Only a lone `initialize` can replace an open session; a batch that carries one is left to the transport's refusal. Scoped
+IPv6 literals (`fe80::1%eth0`) are refused before binding, because no URL can
+name them. Listener errors reach the runtime's connection-loss handler only
+after a successful bind; a bind failure prints the listen-failure message alone. Request bodies are bounded by the SDK's
+4 MiB limit, and the response budget keeps its stdio-derived floor. The
+listener ends on SIGINT or SIGTERM only, not on stdin EOF, and exits `130` or
+`143` after cleanup settles. A signal that arrives while startup is still
+running is recorded and acted on as soon as the listener and its providers are
+owned, so startup never ends without cleanup; a second signal during startup
+exits immediately.
+
 ## Progress and cancellation
 
 REA accepts ordinary `tools/call` progress tokens. The numeric `progress` counts
@@ -419,6 +723,9 @@ Provider calls receive the request cancellation signal. Artifact traversal,
 hashing, version comparisons, Hopper requests, and process capture
 check the same signal. Cancellation is distinct from timeout. A cleanup failure
 uses `cleanup_incomplete` and lists only the owned resource kinds that remain.
+Once those owners confirm cleanup, composed failures recover their original
+typed primary outcome and recovery guidance. Historical composition diagnostics
+remain in `details.cleanup_history`, separate from the primary diagnostics.
 An atomic Evidence bundle or snapshot write can publish its complete output
 before staging cleanup fails. In that case, `details.published: true`,
 `details.published_path`, and `details.bytes` retain the publication receipt.
@@ -428,12 +735,25 @@ Native call tracing and process capture retain available observations in
 `details.partial_observation` on failure, including when cleanup succeeds.
 The observation reports its partial coverage; cleanup details describe host
 state separately from the execution failure.
+For these tools, cancellation observed during result finalization or owned
+cleanup also returns typed cancellation with the available observations.
+A normal target exit does not establish request completion. Incomplete cleanup
+still takes precedence and retains its resource diagnostics.
 A process capture that fails during a finalization interval keeps every signal
 attempt and its delivery result in `details.partial_observation`, with
 `elapsed_ms` and `exit_code` null when no exit was observed.
 Derived comparisons and reconstruction verification yield before computation
 and before publication, so cancellation cannot race with successful Evidence.
 
+Bundle recovery requires a recorded Evidence ID. Some provider errors carry
+typed raw observations in `details.partial_observation` instead of Evidence.
+Those raw observations are not recorded in the Evidence bundle. If the SDK
+suppresses the error reply after client cancellation, Evidence tools cannot
+recover them. An empty bundle does not prove that no partial work or side
+effect occurred; output receipts and cleanup observations remain distinct.
+The broader retention decision is tracked in
+[#4276](https://github.com/morluto/rea/issues/4276).
+
 `analyze_javascript_application` also yields between reconstruction phases and
 during graph/result sealing, cross-graph binding checks, Evidence JSON validation,
 and canonical hashing. Its final result validation reuses exact graphs whose
@@ -492,12 +812,40 @@ analysis runtime. These failures are reported at the operation that needs
 them. A configured provider or an endpoint alone does not establish that a
 target is supported.
 
+Composite Hopper regex searches report failures against the accepted
+`search_strings` or `search_procedures` operation. The underlying inventory
+method and request ID remain in `details.diagnostics.provider_request`, with
+the bridge error code and diagnostic type when available. Reattribution
+preserves the failure class, captured output, and live cleanup ownership.
+
 Evidence bundles, snapshots, and extraction use the paths and output behavior
 declared by their tools. Artifact extraction materializes the selected regular
 files into a fresh REA-chosen temporary directory and reports its path.
 
 `analyze_javascript_application` reads the supplied local directory or ASAR path
-and returns its result and Evidence inline.
+and returns its result and Evidence inline. Its owned analysis worker and private
+temporary source files are declared as `launchesProcess` and `writesFilesystem`,
+including for a single JavaScript source. This internal analysis does not execute
+the supplied application code.
+
+## Static Ghidra bytes and source mappings
+
+`read_bytes` returns initialized analysis memory, including loader fixups, rather
+than live process memory. Check `returned_bytes` and `complete` before decoding:
+reads stop at unmapped or uninitialized bytes. Missing bytes are unknown; do not
+pad an incomplete read with zeros and treat the result as an observation.
+
+`list_segments` reports memory-block initialization and address-space facts.
+`address_to_file_offset` maps one address through the provider's source-byte
+mapping. A successful start-address mapping does not establish that an entire
+range is contiguous in the original file, or that its bytes equal analysis
+memory after loader fixups. Uninitialized, missing, and ambiguous mappings fail
+explicitly; preserve the reported failure rather than guessing an offset.
+
+For constructor-populated pointer or resource slots, a static zero is not proof
+of a later null value. Inspect the slot's `xrefs` and the writer's assembly or
+pseudocode, retaining uncertainty about execution order and runtime values.
+These operations inspect static evidence and do not execute the target.
 
 ## Ghidra first-query deadlines and recovery
 
@@ -509,13 +857,13 @@ auto-analysis, bridge connection, and health readiness before returning analysis
 
 These deadlines have different owners:
 
-| Deadline             | Owner and effect                                                                                                                                                                                                                    |
-| -------------------- | ----------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- |
-| MCP initialize       | The client bounds transport/REA connection startup, before any target query.                                                                                                                                                        |
-| Hopper startup       | REA allows 120,000 ms by default; `REA_HOPPER_STARTUP_TIMEOUT_MS` accepts decimal integers from 1–2,147,483,647 ms. Invalid values fail configuration validation. Startup failure is reported by the first provider query.          |
-| Ghidra startup       | REA allows 330,000 ms by default; `REA_GHIDRA_STARTUP_TIMEOUT_MS` accepts decimal integers from 1–2,147,483,647 ms. Invalid supplied values fail configuration validation. Startup failure is reported by the first provider query. |
-| Darwin swiftc helper | REA allows 120,000 ms by default for ownership-helper compilation; `REA_SWIFTC_COMPILE_TIMEOUT_MS` accepts decimal integers from 1–2,147,483,647 ms. Invalid values fail configuration validation.                                  |
-| Individual tool call | The client bounds its wait, including cold engine startup. The pinned client SDK 2.3.1 defaults to 60,000 ms and can cancel earlier than REA's startup deadline.                                                                    |
+| Deadline             | Owner and effect                                                                                                                                                                                                                                                                                                    |
+| -------------------- | ------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- |
+| MCP initialize       | The client bounds transport/REA connection startup, before any target query.                                                                                                                                                                                                                                        |
+| Hopper startup       | REA allows 120,000 ms by default; `REA_HOPPER_STARTUP_TIMEOUT_MS` accepts decimal integers from 1–2,147,483,647 ms. Invalid values fail configuration validation. Startup failure is reported by the first provider query.                                                                                          |
+| Ghidra startup       | REA allows 330,000 ms by default; `REA_GHIDRA_STARTUP_TIMEOUT_MS` accepts decimal integers from 1–2,147,483,647 ms. Invalid supplied values fail configuration validation. Startup failure is reported by the first provider query.                                                                                 |
+| Darwin swiftc helper | REA allows 120,000 ms by default for ownership-helper compilation; `REA_SWIFTC_COMPILE_TIMEOUT_MS` accepts decimal integers from 1–2,147,483,647 ms. Invalid values fail configuration validation. Set it in the environment that starts REA and restart REA; the next ownership-helper compile uses the new value. |
+| Individual tool call | The client bounds its wait, including cold engine startup. The pinned client SDK 2.3.1 defaults to 60,000 ms and can cancel earlier than REA's startup deadline.                                                                                                                                                    |
 
 For [Codex and Kimi registrations](client-configuration.md), setup supplies
 missing tool deadlines from the known provider allowance plus 120 seconds of
@@ -523,7 +871,8 @@ operation headroom: 450 seconds with the default Ghidra allowance, or 2820
 seconds with `REA_GHIDRA_STARTUP_TIMEOUT_MS=2700000`. Positive caller-selected
 deadlines remain unchanged. Doctor reports an informational recommendation
 when a configured deadline is shorter than the known provider allowance. An
-explicit JEB upstream request budget also contributes to that allowance.
+explicit JEB upstream deadline (`REA_JEB_MCP_TIMEOUT_MS`, shared by catalog
+discovery and each tools/call) also contributes to that allowance.
 
 For an already connected client using the pinned SDK, request options are the
 **second** argument of `callTool`:
@@ -592,6 +941,13 @@ the current client, so it is not a fresh-start recovery. On the same connection:
    `provider_id: "ghidra"`, then retry the first query with an appropriate client
    deadline. The connection and selected provider do not need to change.
 
+If REA's provider startup allowance also needs raising, preserve Evidence and
+close as above, then update `REA_GHIDRA_STARTUP_TIMEOUT_MS` in REA's launch
+environment and restart the server. Reconnect the MCP client before reopening
+the target. Configuration is read at process startup; close/reopen on the same
+connection cannot apply a changed environment. See the
+[CLI retry example](installation.md#ghidra).
+
 This close/reopen flow was exercised on one real Linux stdio connection after a
 controlled startup timeout, followed by a successful overview and function
 analysis. A separate Windows x64 stdio run with Ghidra 12.1.4, JDK 21 and Node
@@ -653,6 +1009,16 @@ bytes and returns the `integrity_contradictions` records for the extracted
 files. An active ASAR unpacked entry whose companion file is absent cannot be
 materialized; extraction fails as `unavailable` with that logical path.
 
+Extraction materializes regular files exposed by the active archive reader.
+When that reader exposes a nested archive as a file, extraction copies the
+archive file; inspect and extract that file separately to unpack its contents.
+Extraction retains inventory and materialization traversal limitations in its
+result and outer Evidence. If either traversal has unknown coverage, the
+observed regular files can still be copied, but the result does not establish
+a complete source copy. For example, a POSIX filename that is not valid UTF-8
+remains omitted with its raw-name explanation; it is not replacement-decoded
+into an output filename.
+
 Packaging tools commonly sign or strip `.asar.unpacked` native binaries after
 writing the archive header, so their declared hashes no longer match. Electron
 does not check unpacked companions against those hashes at runtime. Integrity
@@ -691,18 +1057,39 @@ after SDK conversion and check advertised validation and actual calls.
 Individual model APIs can impose additional nesting limits; complete producer
 captures can exceed ten structural levels.
 
+`search_bytes` advertises its pattern and mask as descriptive strings. Its
+shared canonical validator scans byte pairs, whitespace, wildcards and aligned
+masks without a repeating regular expression or temporary byte arrays. The
+server rejects malformed or unconstrained patterns before provider dispatch;
+the advertised strings alone do not express those rules.
+
+Boolean literal constraints in input schemas are relaxed only in this
+advertisement. For example, a boolean `const` is presented as `type: boolean`;
+an affected `not` is omitted, conditional predicates and their branches are
+omitted together, pure boolean-literal `oneOf` branches collapse to a boolean,
+and other affected `oneOf` branches are presented as `anyOf` to retain branch
+guidance without making overlapping widened branches reject inputs.
+This can make the advertised schema accept inputs the canonical contract
+rejects. The MCP handler still validates every call with the canonical
+contract. The Gemini CLI verifier checks the outgoing
+`parametersJsonSchema` and canonical contract examples against that declaration
+using a loopback model fixture; it does not establish live Google API or
+OpenCode client behavior. Set `REA_VERIFY_GEMINI_SCHEMA_PROFILE=compact` to
+verify the compact advertisement through the same client boundary.
+
 ### Compact input schema profile
 
 Some function-calling providers reject a request while any single tool's
 `parameters` schema exceeds a literal serialized size cap; Moonshot clients
-such as kimi-cli fail every request while the default catalog is attached
+such as kimi-cli have reported failures with the full catalog attached
 ([#1484](https://github.com/morluto/rea/issues/1484)). Setting the server's
-`REA_MCP_INPUT_SCHEMA_PROFILE` environment variable to `compact` renders every
-advertised input schema within a 13,312-byte budget, measured to sit below
+`REA_MCP_INPUT_SCHEMA_PROFILE` environment variable to `compact` (the default)
+renders every advertised input schema within a 13,312-byte budget, measured to sit below
 Moonshot's observed cap (community measurements: a 13.7 KB schema is accepted
 while a ~15 KB one is rejected). The setting is read at startup; restart the
-server to apply changes, as with `REA_MCP_MAX_RESPONSE_BYTES`. The default
-value `full` keeps the complete advertisement unchanged. Compact presentation
+server to apply changes, as with `REA_MCP_MAX_RESPONSE_BYTES`. Explicitly select
+`full` to retain complete advertised input guidance and nested validation detail.
+Existing explicit profile choices are preserved. Compact presentation
 also reduces schemas whose expanded JSON tree exceeds ten levels, the limit
 reported by strict-depth MCP clients in [#2813](https://github.com/morluto/rea/issues/2813).
 WorkBuddy/CodeBuddy gateways have also rejected full REA catalogs with an
@@ -783,6 +1170,44 @@ includes all dimension limitations. Equal CFGs establish only equal observed
 block sizes and successor relationships. Text from different provider
 identities remains unknown.
 
+The `calls` and `references` dimensions share endpoint identity and name
+ambiguity rules. Calls retain procedure entry coordinates and direction;
+references retain each edge's raw source and target coordinates, direction and
+available kind facts. Resolving a reference endpoint to a containing procedure
+does not replace its raw coordinate. Hexadecimal case and zero padding do not
+change coordinates; address spaces remain distinct.
+
+Coordinate equality requires equal artifact digests and formats, a known equal
+architecture, and matching provider-owned native coordinate commitments. Those
+commitments retain the exact selected Mach-O image where available and a digest
+of the loader's mapping inputs. They exclude unrelated analysis settings, so
+changing an analyzer option alone does not invalidate coordinate correspondence.
+Equal coordinates in different slices or mappings cannot establish equality.
+Hopper's default executable profile and Ghidra's supported deterministic imports
+publish these commitments. Hopper loader overrides and attached databases, IDA
+profiles, Ghidra universal-image imports without an exact slice commitment, and
+older or imported Evidence without a commitment leave coordinate identity
+unknown. Unparseable coordinates also remain unknown.
+
+Across different artifacts, endpoints use nonempty stable procedure names.
+Missing or address-generated names remain unknown. A name observed at multiple
+procedure entries across either dossier's call and reference facets retains its
+label constraint but cannot establish equal endpoints. Same-container records
+without a common coordinate context likewise cannot establish equality from
+matching labels. Procedure labels do not carry authoritative symbol-origin
+facts, so recognition of generated names remains conservative and cannot prove
+whether every provider label was generated or explicitly named.
+
+Pairing preserves duplicate counts and considers alternative compatible
+pairings; array order never identifies an edge. Compatible records with unknown
+endpoint identity or reference kind facts remain `unknown`. Independent known
+facts that contradict every possible pairing still establish `changed`.
+Source Evidence retains the original addresses, names and unavailable facts.
+Reconstruction verification uses the selected dimension's status: a same-name
+call retargeting in a common context fails a calls claim; an unresolved slice or
+mapping cannot pass it. Empty function labels leave `function_match` ambiguous
+and the identity dimension unknown.
+
 ### Comparison input shapes
 
 `compare_web_captures` accepts exactly one of two input shapes:
@@ -802,6 +1227,14 @@ invoking the handler. Domain validation additionally checks
 relationships such as event sequence references and retained counts; JSON
 Schema does not express those cross-field invariants.
 
+Comparison also preserves the capture's coverage. A partial HTTP inventory can
+establish an observed difference against a complete peer, but an omitted request
+does not establish removal. Partial WebSocket streams do not establish frame
+alignment. WebMCP frame IDs and tool keys are local to a capture; semantic owner
+facts and known counts determine correspondence. Retained script digests can
+establish source changes. See [browser comparison semantics](browser-observation.md)
+for value uncertainty and storage-fingerprint recapture requirements.
+
 ## Generated catalog
 
 Custom clients upgrading from 6.1 should follow the