fix: emit v-mode-compatible patterns for tool schemas (#1220)

* fix: emit v-mode-compatible patterns for tool schemas

Character classes in emitted patterns carried unescaped '-', '/' and '['. Annex-B and u-mode accept them, so REA's own Zod validation (u) and every u-based validator never saw a defect, but v-mode rejects ClassSetSyntaxCharacters and Ruby rejects [^s[]]. With 6.1.0 tools/list, 117 pattern fields across 11 tools were rejected under v.

The characters are escaped as \x2d, \x2f, \x5b and \x5d rather than \-/\//\[, because oxlint --fix removes an escaped '-' inside a character class as a useless escape, so an escape-only change does not survive the pre-commit hook. \x2d-style escapes are valid under Annex B, u and v, are matched identically, and are not useless-escape candidates.

22 pattern literals in 18 files are touched; no other token changes.

* fix: replace the Zod offset pattern and guard the advertised catalog

The offset datetime site projected Zod's offset class as [ +- ], which no escape in this repository's sources can reach, so evaluate_reconstruction_coverage stayed unreachable for exactly the clients this change targets. It now uses offsetDateTimeSchema in src/domain/offsetDateTime.ts: the same accepted set with the offset class escaped as [ +\x2d ].

Adds tests/boundary/mcp/advertisedPatternsVMode.test.ts, which performs the real tools/list projection, traverses every advertised input and output schema, compiles each pattern under v, and asserts the inventory it walked so a silent traversal collapse fails instead of passing vacuously.

src/browser/CdpSafeMetadata.ts

@@ -92,7 +92,7 @@ const parseCsp = (
   for (const rawDirective of (value ?? "").split(";")) {
     const [rawName, ...tokens] = rawDirective.trim().split(/\s+/u);
     const name = (rawName ?? "").toLowerCase();
-    if (!/^[a-z][a-z0-9-]*$/u.test(name)) continue;
+    if (!/^[a-z][a-z0-9\x2d]*$/u.test(name)) continue;
     const sources: ResponseMetadata["csp"]["directives"][number]["sources"] =
       [];
     for (const token of tokens) {
@@ -249,7 +249,7 @@ const linkParameters = (value: string): ReadonlyMap<string, string> => {
     const name = (separator < 0 ? raw : raw.slice(0, separator))
       .trim()
       .toLowerCase();
-    if (!/^[a-z][a-z0-9-]*$/u.test(name)) continue;
+    if (!/^[a-z][a-z0-9\x2d]*$/u.test(name)) continue;
     const parameter = separator < 0 ? "" : raw.slice(separator + 1).trim();
     parameters.set(name, unquote(parameter));
   }
@@ -285,7 +285,7 @@ const permissionFeatures = (value: string | undefined): string[] =>
       (value ?? "")
         .split(",")
         .map((entry) => entry.split("=", 1)[0]?.trim().toLowerCase() ?? "")
-        .filter((feature) => /^[a-z][a-z0-9-]*$/u.test(feature)),
+        .filter((feature) => /^[a-z][a-z0-9\x2d]*$/u.test(feature)),
     ),
   ].sort();
 

src/contracts/providerSelection.ts

@@ -4,7 +4,7 @@ import { z } from "zod";
 export const analysisProviderIdSchema = z
   .string()
   .min(1)
-  .regex(/^[a-z][a-z0-9-]*$/u)
+  .regex(/^[a-z][a-z0-9\x2d]*$/u)
   .refine((value) => value !== "auto", {
     message: "auto is reserved for provider selection",
   });

src/domain/browserScenarioValues.ts

@@ -10,7 +10,7 @@ import {
 
 export const scenarioIdentifierSchema = z
   .string()
-  .regex(/^[A-Za-z][A-Za-z0-9._-]*$/u);
+  .regex(/^[A-Za-z][A-Za-z0-9._\x2d]*$/u);
 
 const browserExecutablePathSchema = localPathStringSchema.refine(
   isAbsoluteLocalPath,
@@ -108,7 +108,7 @@ export const browserScenarioEnvironmentSchema = z
     timezone: z
       .string()
       .min(1)
-      .regex(/^[A-Za-z0-9_+-]+(?:\/[A-Za-z0-9_+-]+)*$/u)
+      .regex(/^[A-Za-z0-9_+\x2d]+(?:\x2f[A-Za-z0-9_+\x2d]+)*$/u)
       .default("UTC"),
     color_scheme: z.enum(["light", "dark", "no-preference"]).default("light"),
     reduced_motion: z.enum(["reduce", "no-preference"]).default("reduce"),

src/domain/callPathSchemas.ts

@@ -6,10 +6,12 @@ import { prefixedDigestSchema } from "./../domain/digests.js";
 
 const addressSchema = z
   .string()
-  .regex(/^(?:(?:[A-Za-z0-9._~-]|%[0-9A-F]{2})+:)?0x(?:0|[1-9a-f][0-9a-f]*)$/u);
+  .regex(
+    /^(?:(?:[A-Za-z0-9._~\x2d]|%[0-9A-F]{2})+:)?0x(?:0|[1-9a-f][0-9a-f]*)$/u,
+  );
 const inputAddressSchema = z
   .string()
-  .regex(/^(?:(?:[A-Za-z0-9._~-]|%[0-9a-fA-F]{2})+:)?0[xX][0-9a-fA-F]+$/u)
+  .regex(/^(?:(?:[A-Za-z0-9._~\x2d]|%[0-9a-fA-F]{2})+:)?0[xX][0-9a-fA-F]+$/u)
   .transform((address) => {
     const separator = address.lastIndexOf(":") + 1;
     const space = address

src/domain/completionLedgerGeneration.ts

@@ -11,7 +11,7 @@ import {
 import { digestSchema } from "./../domain/digests.js";
 import { prefixedDigestSchema } from "./../domain/digests.js";
 
-const identifierSchema = z.string().regex(/^[A-Za-z][A-Za-z0-9._-]*$/u);
+const identifierSchema = z.string().regex(/^[A-Za-z][A-Za-z0-9._\x2d]*$/u);
 const identitySchema = z.strictObject({
   id: identifierSchema,
   version: z.string().min(1),

src/domain/conformancePackage.ts

@@ -10,7 +10,7 @@ import { prefixedDigestSchema } from "./../domain/digests.js";
 
 const conformancePackageIdSchema = prefixedDigestSchema("cp");
 
-const scenarioIdSchema = z.string().regex(/^[A-Za-z][A-Za-z0-9._-]*$/u);
+const scenarioIdSchema = z.string().regex(/^[A-Za-z][A-Za-z0-9._\x2d]*$/u);
 
 const scenarioManifestSchema = z.strictObject({
   scenario_id: scenarioIdSchema,

src/domain/evidenceCompletionLedger.ts

@@ -5,7 +5,7 @@ import { z } from "zod";
 import { prefixedDigestSchema } from "./../domain/digests.js";
 
 const evidenceIdSchema = prefixedDigestSchema("ev");
-const claimIdSchema = z.string().regex(/^[A-Za-z][A-Za-z0-9._-]*$/u);
+const claimIdSchema = z.string().regex(/^[A-Za-z][A-Za-z0-9._\x2d]*$/u);
 
 /** Terminal outcome recorded for one verifier claim. */
 export const evidenceCompletionStatusSchema = z.enum([

src/domain/native/binaryLayout.ts

@@ -7,7 +7,7 @@ const range = z.strictObject({ offset: unsignedHex, bytes: unsignedHex });
 const canonicalBase64 = z
   .string()
   .regex(
-    /^(?:[A-Za-z0-9+/]{4})*(?:[A-Za-z0-9+/][AQgw]==|[A-Za-z0-9+/]{2}[AEIMQUYcgkosw048]=)?$/,
+    /^(?:[A-Za-z0-9+\x2f]{4})*(?:[A-Za-z0-9+\x2f][AQgw]==|[A-Za-z0-9+\x2f]{2}[AEIMQUYcgkosw048]=)?$/,
   );
 const name = z.strictObject({
   display: z.string(),

src/domain/native/nativeCallObservation.ts

@@ -11,7 +11,7 @@ const nativeString = z
 /** Objective-C class names and selectors never contain spaces or brackets. */
 const objcName = nativeString
   .min(1)
-  .regex(/^[^\s[\]]+$/u, "Use the bare name without spaces or brackets");
+  .regex(/^[^\s\x5b\x5d]+$/u, "Use the bare name without spaces or brackets");
 
 const functionBreakpoint = {
   kind: z.literal("function"),

src/domain/native/recordedCrash.ts

@@ -7,7 +7,7 @@ const signed32 = z.number().int().min(-2147483648).max(2147483647);
 const base64 = z
   .string()
   .regex(
-    /^(?:[A-Za-z0-9+/]{4})*(?:[A-Za-z0-9+/][AQgw]==|[A-Za-z0-9+/]{2}[AEIMQUYcgkosw048]=)?$/,
+    /^(?:[A-Za-z0-9+\x2f]{4})*(?:[A-Za-z0-9+\x2f][AQgw]==|[A-Za-z0-9+\x2f]{2}[AEIMQUYcgkosw048]=)?$/,
   );
 const range = z.strictObject({ offset: hex, bytes: hex });
 const diagnostics = z.strictObject({

src/domain/offsetDateTime.ts

@@ -0,0 +1,21 @@
+import { z } from "zod";
+
+/**
+ * ISO-8601 date-time that requires an explicit UTC offset.
+ *
+ * `z.string().datetime({ offset: true })` emits Zod's offset class as `[+-]`.
+ * Annex B and `u` mode accept that, but ECMA-262 `v` mode rejects the
+ * unescaped `-`, so any tool whose schema carries the pattern is unusable for a
+ * client that compiles advertised patterns — schema compilation happens before
+ * arguments are supplied, so the whole tools request fails rather than one call.
+ *
+ * The accepted set is identical to the Zod validator; only the offset class
+ * escapes `-`, as `\x2d`, because an escaped `-` inside a character class is
+ * removed by `oxlint --fix` as a useless escape.
+ */
+export const offsetDateTimeSchema = z
+  .string()
+  .regex(
+    /^(?:(?:\d\d[2468][048]|\d\d[13579][26]|\d\d0[48]|[02468][048]00|[13579][26]00)-02-29|\d{4}-(?:(?:0[13578]|1[02])-(?:0[1-9]|[12]\d|3[01])|(?:0[469]|11)-(?:0[1-9]|[12]\d|30)|(?:02)-(?:0[1-9]|1\d|2[0-8])))T(?:(?:[01]\d|2[0-3]):[0-5]\d(?::[0-5]\d(?:\.\d+)?)?(?:Z|(?:[+\x2d](?:[01]\d|2[0-3]):[0-5]\d)))$/u,
+    "Use an ISO-8601 date-time with an explicit UTC offset",
+  );

src/domain/process/processTraceEvaluation.ts

@@ -15,7 +15,7 @@ import {
   type ProcessTraceSpecification,
 } from "./processTraceSpecification.js";
 
-const identifierSchema = z.string().regex(/^[A-Za-z][A-Za-z0-9._-]{0,63}$/u);
+const identifierSchema = z.string().regex(/^[A-Za-z][A-Za-z0-9._\x2d]{0,63}$/u);
 
 export type ProcessTraceLocation = ProcessObservationLocation;
 

src/domain/process/processTraceSpecification.ts

@@ -8,7 +8,7 @@ import {
   type ProcessObservationSource,
 } from "./processObservation.js";
 
-const identifierSchema = z.string().regex(/^[A-Za-z][A-Za-z0-9._-]*$/u);
+const identifierSchema = z.string().regex(/^[A-Za-z][A-Za-z0-9._\x2d]*$/u);
 
 /** Process observation families admitted by a declared trace specification. */
 export const processTraceSourceSchema = processObservationSourceSchema;

src/domain/reconstructionCoverage.ts

@@ -14,9 +14,12 @@ import {
 } from "./reconstructionCoverageEvaluation.js";
 import { digestSchema } from "./../domain/digests.js";
 import { prefixedDigestSchema } from "./../domain/digests.js";
+import { offsetDateTimeSchema } from "./offsetDateTime.js";
 
 const evidenceIdSchema = prefixedDigestSchema("ev");
-const stableIdSchema = z.string().regex(/^[A-Za-z][A-Za-z0-9._:/-]{0,199}$/u);
+const stableIdSchema = z
+  .string()
+  .regex(/^[A-Za-z][A-Za-z0-9._:\x2f\x2d]{0,199}$/u);
 const boundedTextSchema = z.string().trim().min(1);
 
 const authoritySchema = z.enum([
@@ -106,7 +109,7 @@ const verifierContractSchema = verifierContractBaseSchema.extend({
 const verifierResultSchema = z.strictObject({
   verifier_id: stableIdSchema,
   contract_sha256: digestSchema,
-  observed_at: z.string().datetime({ offset: true }),
+  observed_at: offsetDateTimeSchema,
   status: z.enum([
     "pass",
     "fail",

src/domain/reconstructionObligationLedgerSchemas.ts

@@ -5,7 +5,7 @@ import { digestSchema } from "./../domain/digests.js";
 import { prefixedDigestSchema } from "./../domain/digests.js";
 
 const evidenceIdSchema = prefixedDigestSchema("ev");
-const stableIdSchema = z.string().regex(/^[A-Za-z][A-Za-z0-9._:/-]*$/u);
+const stableIdSchema = z.string().regex(/^[A-Za-z][A-Za-z0-9._:\x2f\x2d]*$/u);
 const textSchema = z.string().trim().min(1);
 
 export const obligationStatusSchema = z.enum([

src/domain/reconstructionReadinessSchemas.ts

@@ -6,7 +6,7 @@ import { digestSchema } from "./../domain/digests.js";
 import { prefixedDigestSchema } from "./../domain/digests.js";
 
 const evidenceIdSchema = prefixedDigestSchema("ev");
-const stableIdSchema = z.string().regex(/^[A-Za-z][A-Za-z0-9._:/-]*$/u);
+const stableIdSchema = z.string().regex(/^[A-Za-z][A-Za-z0-9._:\x2f\x2d]*$/u);
 const boundedTextSchema = z.string().trim().min(1);
 
 export const readinessStatusSchema = z.enum([

src/domain/reconstructionVerificationSchemas.ts

@@ -5,7 +5,7 @@ import { PROCESS_COMPARISON_DIMENSIONS } from "./process/processComparison.js";
 
 const evidenceIdSchema = prefixedDigestSchema("ev");
 const unknownIdSchema = prefixedDigestSchema("unk");
-const claimIdSchema = z.string().regex(/^[A-Za-z][A-Za-z0-9._-]*$/u);
+const claimIdSchema = z.string().regex(/^[A-Za-z][A-Za-z0-9._\x2d]*$/u);
 const titleSchema = z.string().trim().min(1);
 const verificationStatusSchema = z.enum(["pass", "fail", "unknown"]);
 const commonClaim = {

src/domain/referenceSourceGraph.ts

@@ -177,7 +177,7 @@ const graphShape = {
     caller: z
       .string()
       .min(1)
-      .regex(/^[\w .:@/+-]+$/u),
+      .regex(/^[\w .:@\x2f+\x2d]+$/u),
   }),
   limitations: z.array(boundedTextSchema),
 };

src/ghidra/extensions/GhidraExtensions.ts

@@ -19,7 +19,7 @@ import { nativeAotAdapter } from "./nativeaot/NativeAotAdapter.js";
 const digest = z.string().regex(/^[a-f0-9]{64}$/u);
 /** Exact executable extension artifact committed before the session starts. */
 export const ghidraExtensionSchema = z.strictObject({
-  id: z.string().regex(/^[a-z][a-z0-9-]*$/u),
+  id: z.string().regex(/^[a-z][a-z0-9\x2d]*$/u),
   path: z.string().refine(isAbsolute, "Extension path must be absolute"),
   configured_path: z
     .string()

tests/boundary/mcp/advertisedPatternsVMode.test.ts

@@ -0,0 +1,77 @@
+import { Client, InMemoryTransport } from "@modelcontextprotocol/client";
+import { expect, it, onTestFinished } from "vitest";
+
+import { createServer } from "../../../src/server/createServer.js";
+import { createTestBinarySession } from "../../fixtures/binarySession.js";
+
+interface AdvertisedPattern {
+  readonly tool: string;
+  readonly where: string;
+  readonly pattern: string;
+}
+
+const collectPatterns = (
+  value: unknown,
+  where: string,
+  tool: string,
+  found: AdvertisedPattern[],
+): void => {
+  if (value === null || typeof value !== "object") return;
+  if (Array.isArray(value)) {
+    value.forEach((entry, index) =>
+      collectPatterns(entry, `${where}[${String(index)}]`, tool, found),
+    );
+    return;
+  }
+  for (const [key, entry] of Object.entries(value)) {
+    if (key === "pattern" && typeof entry === "string")
+      found.push({ tool, where: `${where}/pattern`, pattern: entry });
+    collectPatterns(entry, `${where}/${key}`, tool, found);
+  }
+};
+
+it("advertises patterns that compile under every engine a client may use", async () => {
+  const session = createTestBinarySession(() => {
+    throw new Error("No deep provider may start for a schema projection");
+  });
+  const server = createServer(session, session);
+  const client = new Client({ name: "advertised-patterns", version: "1" });
+  onTestFinished(async () => {
+    await client.close();
+    await server.close();
+    await session.close();
+  });
+  const [clientTransport, serverTransport] =
+    InMemoryTransport.createLinkedPair();
+  await server.connect(serverTransport);
+  await client.connect(clientTransport);
+
+  const advertised = (await client.listTools()).tools;
+  const patterns: AdvertisedPattern[] = [];
+  for (const tool of advertised) {
+    collectPatterns(tool.inputSchema, "inputSchema", tool.name, patterns);
+    if (tool.outputSchema !== undefined)
+      collectPatterns(tool.outputSchema, "outputSchema", tool.name, patterns);
+  }
+
+  // A traversal that silently collapses would make the compilation check below
+  // vacuous, so the inventory itself is asserted.
+  expect(advertised.length).toBeGreaterThan(100);
+  expect(patterns.length).toBeGreaterThan(500);
+
+  // Annex B and `u` mode accept an unescaped `-`, `/` or `[` inside a character
+  // class, so REA's own schema validation cannot observe the defect. `v` mode
+  // rejects them, and a client that compiles advertised patterns rejects the
+  // complete tools request rather than the affected call.
+  const uncompilable = patterns.flatMap((entry) => {
+    try {
+      new RegExp(entry.pattern, "v");
+      return [];
+    } catch (cause: unknown) {
+      return [
+        `${entry.tool} ${entry.where} ${entry.pattern}: ${cause instanceof Error ? cause.message : String(cause)}`,
+      ];
+    }
+  });
+  expect(uncompilable).toEqual([]);
+});