Merge remote-tracking branch 'origin/main' into HEAD

docs(skill): route browser actions to scenario capture
Route user browser actions to passive observation and agent actions to declared scenarios. Clarify requested artifacts, step outcomes, connect navigation and cleanup, and the limits of listener and request-initiator evidence. Distinguish discarded raw headers from retained response metadata. Update the bundled skill to version 42 with matching package metadata. Validation on the synchronized revision: docs (5 tests), typecheck/lint, formatting, 61 focused tests, and real Linux Chromium launch/connect and passive CLI probes. Verified declared steps, requested artifacts, digests, external page preservation, temporary form reset and cleanup. Native Windows/macOS and separate follow-up tools were not rerun. Package install/update and CLI/MCP verification passed earlier on 738d02a, before this main sync. Co-authored-by: 0xRafie <155859586+0xRafie@users.noreply.github.com> Co-authored-by: Kaoru0822 <208281322+kaoru0822-kitauji@users.noreply.github.com> Co-authored-by: factory-droid[bot] <138933559+factory-droid[bot]@users.noreply.github.com>
2026-10-14 00:31:27 +00:00 · 2026-10-14 08:31:16 +08:00 · 2026-10-14 08:28:23 +08:00 · 2026-10-14 08:27:37 +08:00 · 2026-10-14 08:22:50 +08:00 · 2026-10-14 08:12:33 +08:00
482 changed files with 94685 additions and 34984 deletions
@@ -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: "41"
+  version: "42"
 ---

 # REA
@@ -31,6 +31,49 @@ 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.

+1. Diagnose without changing REA or client configuration. `npx` may download
+   the package and write to npm's cache. For Codex, run
+   `npx -y rea-agents@latest doctor --client codex --json`. Substitute the current
+   supported client: `claude_code`, `claude_desktop`, `codex`, `cursor`,
+   `gemini_cli`, `windsurf`, `devin`, `opencode`, `antigravity`, `copilot_cli`,
+   `commandcode`, `vscode`, `grok_build`, `omp`, or `grok_bot`. `grok_bot` has no local
+   file registration: doctor reports the Grok Bot chat step, and setup does not
+   write `mcp.json`. If the client is unknown, use `doctor --json` and
+   inspect its registration results before choosing a setup scope.
+2. Distinguish the reason. Missing, malformed, or stale registration needs a
+   scoped configuration repair. An aligned registration with no tools in the
+   active session needs a restart/reconnection; doctor cannot prove that the
+   current agent has connected. A missing provider affects only tasks requiring
+   that provider: static JavaScript inspection needs neither Hopper nor Ghidra,
+   and Android inspection has separate bring-your-own JADX/Java prerequisites.
+3. For a configuration repair, prepare the read-only plan:
+   `npx -y rea-agents@latest setup --client codex --dry-run --json`.
+   Use the current client's ID, show its exact proposed paths, backups, and
+   changes, and obtain approval before setup writes configuration or installs
+   Hopper. Setup normally installs the matching bundled skill too; include that
+   replacement in the reviewed plan. After approval, apply the same scope with
+   `npx -y rea-agents@latest setup --client codex --yes`. Add `--install-hopper`
+   only if that separate installation was needed and explicitly approved.
+4. Restart/reconnect the affected agent. Verify that REA tools actually appear
+   in the session, then resume the original investigation. If they remain
+   absent, inspect the client's MCP launch error rather than repeating setup.
+
+REA setup never installs or upgrades Node.js, npm, Homebrew, Java, Ghidra, IDA,
+JADX, Binwalk, or Unblob. Use existing prerequisites; do not install unrelated
+software to repair
+MCP registration. For an unsupported client, use manual stdio registration
+with a version-pinned `rea-agents` package or continue through the CLI.
+
+A concrete CLI fallback for an operator-supplied JavaScript tree or ASAR is:
+
+```bash
+npx -y rea-agents@latest analyze-javascript-application /absolute/path/to/app --json
+```
+
+No MCP registration or native engine is required. Read the returned Evidence,
+graph, limitations, and unknowns with the same care as an MCP result; the CLI
+returns the Evidence record directly. This fallback does not establish that
+MCP is configured. Native CLI tasks still require their selected engine.
 Prefer an existing `rea` CLI. When none is installed and REA package acquisition
 is within the authorized scope, use `npx -y rea-agents@latest` in its place.
 Configuration repair starts with a read-only setup plan; show paths, backups and
@@ -61,6 +104,10 @@ explicit path or endpoint and do not need it.
  consult the NativeAOT workflow in `references/native-and-artifacts.md` for
  optional metadata recovery and its supported host/layout boundary.
 - User-owned browser page already open: `list_browser_targets`.
+- Browser action the agent must perform (click, type, navigate):
+  `capture_browser_scenario` with declared actions. See
+  [runtime observation](references/runtime-observation.md) for connect/launch
+  and the focused follow-up browser tools.
 - Fiddler Everywhere traffic or a native `.saz` capture: read the
  [Fiddler workflow](references/runtime-observation.md#fiddler-everywhere-traffic-debugging).
  Inspect saved SAZ with `inspect_web_network_capture` and `format: "saz"`;
@@ -280,7 +327,8 @@ as complete while required questions remain open.
  [references/javascript-applications.md](references/javascript-applications.md)
 - Android APK declarations, classes, methods, and static references:
  [references/android-applications.md](references/android-applications.md)
- Passive browser/Electron observation and static/runtime reconciliation:
+- Browser/Electron observation (passive and scenario) and static/runtime
+  reconciliation:
  [references/runtime-observation.md](references/runtime-observation.md)
 - Evidence paging, comparisons, residual unknowns, and verification:
  [references/evidence-workflows.md](references/evidence-workflows.md)
@@ -65,3 +65,7 @@ Use `inspect_analysis_view` with `view: {kind: "api-surface"}` on retained
 method, origin and path template. Format matches are inferences, not handler
 identity or schemas. Group pages cite parent record ordinals; consult record
 coverage and exclusions before claiming completeness. No requests are replayed.
+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.
@@ -103,6 +103,11 @@ architectures, build_metadata, load_commands, dependencies, imports, exports,
 segments and provenance. `source_exhaustive: false` means unreturned symbols
 may exist.

+`decode_interface_builder` graphs are often several megabytes and can exceed the
+MCP response budget. Select `detail: "summary"` for every decoded document and
+node/edge counts, then page `view: {kind: "interface-builder", collection:
+"edges", document: "<documents[].relative_path>", offset: 0, limit: 64}`.
+
 To see which functions or Objective-C methods a Mach-O actually calls in one
 run, use `observe_native_calls` with explicit breakpoints and a bounded
 `duration_ms`/`max_events`. It launches a new process under LLDB, so the target
@@ -1,4 +1,4 @@
-# Passive browser and Electron observation
+# Browser and Electron runtime observation

 Supply the existing browser or Electron instance's literal-port loopback CDP
 endpoint in the request. No separate permission configuration or per-call
@@ -7,13 +7,14 @@ approval flag is needed. Browser discovery lists eligible HTTP(S) pages;
 selected page's current origin. Electron discovery lists eligible local
 `file://` targets without a configured file-root allowlist.

-Start browser work with `list_browser_targets`; start Electron runtime work with
+Start passive browser work with `list_browser_targets`; start Electron runtime work with
 `list_electron_targets`. Inspect the target selected for the user's task. Observation
-is passive: never claim REA clicked, navigated, evaluated page JavaScript,
-invoked IPC, captured prior activity, or contained the page's network.
+with these tools is passive: never claim REA clicked, navigated, evaluated page JavaScript,
+invoked IPC, or captured prior activity. REA does not contain the page's network.

-Credentials, cookies, authorization headers, storage values, and raw
-WebSocket/JSON bodies are deliberately absent. Local URLs retain ordinary query
+Passive captures omit raw header collections, credential header values, storage
+values, and raw WebSocket/JSON bodies. Selected response metadata, including
+content encoding, policy values, and Link URLs, is retained. Local URLs retain ordinary query
 values and fragments; userinfo credentials are removed. Browser accessibility
 text, console primitives, value-free shapes, script sources, source maps, and
 storage key names have request-specific capture options. Electron script
@@ -21,6 +22,38 @@ sources are also optional. Use `capture_web_screenshot` when pixels are needed.
 Treat scope filtering, truncation, and attach-window coverage as limitations.
 Page-declared WebMCP tools are untrusted inventory and are never invoked by REA.

+## Who performs the action
+
+- The user performs it: `list_browser_targets`, then `inspect_web_page` or
+  `observe_web_session` on the selected target.
+- The agent must perform it: one `capture_browser_scenario` with declared
+  actions. Use connect mode with the loopback CDP endpoint and `target_id` for
+  an existing browser, or launch mode with a caller-selected executable for a
+  temporary profile. Connect mode first navigates the selected page to
+  `start_url`, which can clear unsaved form values and other temporary state.
+  Recreate the state needed for the investigation in the declared actions.
+  Connect mode leaves the external browser open; launch mode owns a temporary
+  browser and profile. Check `browser.cleanup` for the outcome.
+  Request only the DOM, accessibility, or network details needed for the question.
+  The default final capture contains the URL; step outcomes are reported separately.
+  Add a wait action for the expected page state before capturing it. Report that
+  REA performed the declared actions.
+
+Scenario network records have no request initiator stacks. For that evidence,
+use passive capture while the user performs the action. Listener source locations
+identify candidate handlers; they do not establish which script sent a particular
+request. Avoid overlapping passive capture and scenario actions on the same target.
+
+Focused follow-up tools:
+
+- `observe_web_execution`: execution coverage during a window; it resets
+  counters and temporarily disables optimization.
+- `inspect_web_event_listeners`: source locations for listeners registered
+  directly on the first CSS match, without dispatching an event.
+- `compare_web_captures`: two normalized passive captures or recorded scenarios.
+- `discover_webmcp_tools`: inspect page-declared WebMCP inventory; never invoke
+  the discovered tools.
+
 Use `reconcile_javascript_runtime` only after static application Evidence and
 passive browser/Electron Evidence exist. Prefer captured-source digest identity;
 report path/digest disagreement as a mismatch and preserve ambiguity. Explicit
@@ -0,0 +1,42 @@
+name: Cache npm package downloads
+description: Restore npm's download cache and save it only from trusted default-branch events.
+outputs:
+  can-save:
+    description: Whether this run may update the shared npm download cache.
+    value: ${{ steps.context.outputs.can-save }}
+  path:
+    description: The npm download-cache directory.
+    value: ${{ steps.context.outputs.path }}
+runs:
+  using: composite
+  steps:
+    - id: context
+      shell: bash
+      env:
+        DEFAULT_BRANCH: ${{ github.event.repository.default_branch }}
+        LOCKFILES_HASH: ${{ hashFiles('package-lock.json', 'npm-shrinkwrap.json', '.npmrc') }}
+      run: |
+        cache_path="$(npm config get cache)"
+        runtime_arch="$(node -p 'process.arch')"
+        can_save=false
+        if [[ -n "${DEFAULT_BRANCH}" && "${GITHUB_REF}" == "refs/heads/${DEFAULT_BRANCH}" && ( "${GITHUB_EVENT_NAME}" == "push" || "${GITHUB_EVENT_NAME}" == "schedule" || "${GITHUB_EVENT_NAME}" == "workflow_dispatch" ) ]]; then
+          can_save=true
+        fi
+        cache_key="npm-download-${RUNNER_OS}-${runtime_arch}-${LOCKFILES_HASH}"
+        {
+          printf 'can-save=%s\n' "${can_save}"
+          printf 'path=%s\n' "${cache_path}"
+          printf 'key=%s\n' "${cache_key}"
+        } >> "${GITHUB_OUTPUT}"
+    - name: Restore and save npm package downloads
+      if: steps.context.outputs.can-save == 'true'
+      uses: actions/cache@0057852bfaa89a56745cba8c7296529d2fc39830 # v4.3.0
+      with:
+        path: ${{ steps.context.outputs.path }}
+        key: ${{ steps.context.outputs.key }}
+    - name: Restore npm package downloads without saving
+      if: steps.context.outputs.can-save != 'true'
+      uses: actions/cache/restore@0057852bfaa89a56745cba8c7296529d2fc39830 # v4.3.0
+      with:
+        path: ${{ steps.context.outputs.path }}
+        key: ${{ steps.context.outputs.key }}
@@ -12,12 +12,13 @@ concurrency:
 permissions:
  contents: read
  actions: read
+  pull-requests: read

 jobs:
  report:
    name: Collect recent completed workflow attempts
    runs-on: ubuntu-latest
-    timeout-minutes: 5
+    timeout-minutes: 10
    steps:
      # Consume only API data and execute batch code from the trusted default branch.
      - uses: actions/checkout@11d5960a326750d5838078e36cf38b85af677262 # v4.4.0
@@ -27,6 +28,18 @@ jobs:
      - uses: actions/setup-node@49933ea5288caeca8642d1e84afbd3f7d6820020 # v4.4.0
        with:
          node-version-file: .nvmrc
+      - name: Cache npm package downloads
+        id: npm-cache
+        continue-on-error: true
+        uses: ./.github/actions/npm-cache
+      - name: Diagnose missing and startup-failed PR CI
+        continue-on-error: true
+        env:
+          GH_TOKEN: ${{ github.token }}
+          CI_REPOSITORY: ${{ github.repository }}
+          CI_REPOSITORY_ID: ${{ github.repository_id }}
+          CI_DEFAULT_BRANCH: ${{ github.event.repository.default_branch }}
+        run: node scripts/ci/pr-ci-health.mjs
      - name: Record queue, execution, and matched step timing
        continue-on-error: true
        env:
@@ -44,3 +57,28 @@ jobs:
          if-no-files-found: error
          overwrite: true
          retention-days: 14
+      - name: Install dependencies for static failure attribution
+        id: failure-deps
+        if: ${{ !cancelled() && hashFiles('.cache/ci-job-timings/batch-manifest.json') != '' }}
+        continue-on-error: true
+        run: npm ci
+      - name: Attribute supported static failures against main
+        if: ${{ !cancelled() && steps.failure-deps.outcome == 'success' && hashFiles('.cache/ci-job-timings/batch-manifest.json') != '' }}
+        continue-on-error: true
+        env:
+          GH_TOKEN: ${{ github.token }}
+          CI_REPOSITORY: ${{ github.repository }}
+          CI_DEFAULT_BRANCH: ${{ github.event.repository.default_branch }}
+        run: node scripts/ci/failure-report.mjs
+      - name: Upload CI health and failure diagnostics
+        if: ${{ !cancelled() }}
+        continue-on-error: true
+        uses: actions/upload-artifact@ea165f8d65b6e75b540449e92b4886f43607fa02 # v4.6.2
+        with:
+          name: ci-health-diagnostics
+          path: |
+            .cache/ci-pr-health/
+            .cache/ci-failures/
+          include-hidden-files: true
+          if-no-files-found: warn
+          retention-days: 14
@@ -42,6 +42,7 @@ jobs:
      tier: ${{ steps.plan.outputs.tier }}
      qualification_run_id: ${{ steps.qualification.outputs.run_id }}
      code: ${{ steps.plan.outputs.code }}
+      reverse_tools: ${{ steps.plan.outputs.reverse_tools }}
      sqlite: ${{ steps.plan.outputs.sqlite }}
      runtime: ${{ steps.plan.outputs.runtime }}
      package: ${{ steps.plan.outputs.package }}
@@ -122,11 +123,28 @@ jobs:
      - uses: actions/setup-node@49933ea5288caeca8642d1e84afbd3f7d6820020 # v4.4.0
        with:
          node-version-file: .nvmrc
-          cache: npm
+      - name: Cache npm package downloads
+        id: npm-cache
+        uses: ./.github/actions/npm-cache
      - run: npm ci
+      - name: Restore static task cache
+        if: needs.changes.outputs.code == 'true'
+        id: static-cache
+        uses: actions/cache/restore@0057852bfaa89a56745cba8c7296529d2fc39830 # v4.3.0
+        with:
+          path: .turbo/cache/
+          key: rea-static-v1-${{ runner.os }}-${{ runner.arch }}-${{ hashFiles('.nvmrc', '.npmrc', 'package-lock.json') }}-${{ github.sha }}
+          # Turbo validates each task's inputs after restoring this snapshot.
+          restore-keys: rea-static-v1-${{ runner.os }}-${{ runner.arch }}-${{ hashFiles('.nvmrc', '.npmrc', 'package-lock.json') }}-
      - 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=
+      - 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 }}
+        run: npm run ci:static-context
      - if: needs.changes.outputs.code == 'true'
        run: npm run check:ci
      - if: needs.changes.outputs.code == 'true'
@@ -136,6 +154,15 @@ jobs:
        run: npm run test:python
      - if: needs.changes.outputs.code != 'true'
        run: npm run format:check
+      - name: Save validated static task cache
+        if: >-
+          success() && needs.changes.outputs.code == 'true' &&
+          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: Upload complete static diagnostics
        if: always() && needs.changes.outputs.code == 'true'
        uses: actions/upload-artifact@ea165f8d65b6e75b540449e92b4886f43607fa02 # v4.6.2
@@ -146,6 +173,34 @@ jobs:
          if-no-files-found: ignore
          retention-days: 7

+  test-source:
+    name: Source tests
+    needs: changes
+    if: needs.changes.outputs.code == 'true' && needs.changes.outputs.tier == 'fast'
+    runs-on: ubuntu-latest
+    timeout-minutes: 10
+    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: Run source tests without waiting for build
+        run: npm run test:fast -- --reporter=default --reporter=blob
+        env:
+          VITEST_BLOB_LABEL: source
+      - uses: actions/upload-artifact@ea165f8d65b6e75b540449e92b4886f43607fa02 # v4.6.2
+        if: ${{ !cancelled() }}
+        with:
+          name: vitest-report-source
+          path: .vitest/blob/*
+          include-hidden-files: true
+          if-no-files-found: error
+          retention-days: 7
+
  test-shard:
    name: Test shard (${{ matrix.shard }})
    needs: [changes, build]
@@ -153,15 +208,21 @@ jobs:
    runs-on: ubuntu-latest
    timeout-minutes: 30
    strategy:
-      fail-fast: false
+      # Failed PRs release the remaining matrix slots; qualification retains
+      # 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:
-        shard: [1/6, 2/6, 3/6, 4/6, 5/6, 6/6]
+        # Fast feedback spends the same six test jobs on one source group and
+        # five built groups; full qualification retains six complete shards.
+        shard: ${{ fromJSON(needs.changes.outputs.tier == 'fast' && '["1/5","2/5","3/5","4/5","5/5"]' || '["1/6","2/6","3/6","4/6","5/6","6/6"]') }}
    steps:
      - uses: actions/checkout@11d5960a326750d5838078e36cf38b85af677262 # v4.4.0
      - uses: actions/setup-node@49933ea5288caeca8642d1e84afbd3f7d6820020 # v4.4.0
        with:
          node-version-file: .nvmrc
-          cache: npm
+      - name: Cache npm package downloads
+        id: npm-cache
+        uses: ./.github/actions/npm-cache
      - run: npm ci
      - uses: actions/download-artifact@d3f86a106a0bac45b974a628896c90dbdf5c8093 # v4.3.0
        with:
@@ -172,21 +233,42 @@ jobs:
        run: npm run test:ci:shard:run -- --shard=${{ matrix.shard }}
        env:
          REA_CI_TEST_HISTORY: .cache/ci-test-history/test-results.json
-      - name: Run complete portable suite for fast feedback
+      - name: Run built-dependent tests for fast feedback
        if: needs.changes.outputs.tier == 'fast'
-        run: npm run test:run -- --reporter=default --reporter=blob --shard=${{ matrix.shard }}
+        run: npm run test:built -- --reporter=default --reporter=blob --shard=${{ matrix.shard }}
        env:
+          VITEST_BLOB_LABEL: built
          REA_CI_TEST_HISTORY: .cache/ci-test-history/test-results.json
      - uses: actions/upload-artifact@ea165f8d65b6e75b540449e92b4886f43607fa02 # v4.6.2
        if: ${{ !cancelled() }}
        with:
-          name: vitest-report-${{ strategy.job-index }}
+          name: vitest-report-built-${{ strategy.job-index }}
          path: .vitest/blob/*
          include-hidden-files: true
          if-no-files-found: error
          overwrite: true
          retention-days: 7

+  reverse-tools-quality:
+    name: Reverse tools coverage and mutation
+    needs: [changes, static]
+    if: needs.changes.outputs.reverse_tools == 'true'
+    runs-on: ubuntu-latest
+    timeout-minutes: 60
+    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: Run reverse tools tests with coverage
+        run: npm run test:coverage:reverse-tools
+      - name: Report reverse tools mutation results
+        run: npm run test:mutation:reverse-tools
+
  build:
    name: Build package
    needs: changes
@@ -212,7 +294,9 @@ jobs:
      - uses: actions/setup-node@49933ea5288caeca8642d1e84afbd3f7d6820020 # v4.4.0
        with:
          node-version-file: .nvmrc
-          cache: npm
+      - name: Cache npm package downloads
+        id: npm-cache
+        uses: ./.github/actions/npm-cache
      - run: npm ci
      - name: Reuse portable build outputs across runs
        uses: ./.github/actions/portable-build-cache
@@ -271,7 +355,7 @@ jobs:

  test:
    name: Portable tests
-    needs: [changes, test-shard]
+    needs: [changes, test-source, test-shard]
    if: ${{ !cancelled() && needs.changes.outputs.tier != 'reused' }}
    runs-on: ubuntu-latest
    timeout-minutes: 10
@@ -286,13 +370,18 @@ jobs:
        if: needs.changes.outputs.code == 'true' && success()
        with:
          node-version-file: .nvmrc
-          cache: npm
+      - name: Cache npm package downloads
+        if: needs.changes.outputs.code == 'true' && success()
+        id: npm-cache
+        uses: ./.github/actions/npm-cache
      - if: needs.changes.outputs.code == 'true' && success()
        run: npm ci
      - uses: actions/download-artifact@d3f86a106a0bac45b974a628896c90dbdf5c8093 # v4.3.0
        if: >-
          !cancelled() && needs.changes.outputs.code == 'true' &&
-          (needs.test-shard.result == 'success' || needs.test-shard.result == 'failure')
+          (needs.test-shard.result == 'success' || needs.test-shard.result == 'failure' ||
+           (needs.changes.outputs.tier == 'fast' &&
+            (needs.test-source.result == 'success' || needs.test-source.result == 'failure')))
        with:
          path: .vitest/blob
          pattern: vitest-report-*
@@ -320,7 +409,8 @@ jobs:
        if: >-
          !cancelled() && needs.changes.outputs.code == 'true' &&
          needs.changes.outputs.tier == 'fast' &&
-          (needs.test-shard.result == 'success' || needs.test-shard.result == 'failure')
+          (needs.test-shard.result == 'success' || needs.test-shard.result == 'failure' ||
+           needs.test-source.result == 'success' || needs.test-source.result == 'failure')
        run: >-
          npm exec -- vitest --merge-reports
          --reporter=default
@@ -333,6 +423,8 @@ jobs:
        if: needs.changes.outputs.code == 'true' && !cancelled() && hashFiles('test-timings.json') != ''
        env:
          SHARD_RESULT: ${{ needs.test-shard.result }}
+          SOURCE_REQUIRED: ${{ needs.changes.outputs.tier == 'fast' }}
+          SOURCE_RESULT: ${{ needs.test-source.result }}
          MERGE_RESULT: ${{ needs.changes.outputs.tier == 'full' && steps.merge.outcome || steps.merge-fast.outcome }}
          CI_VERIFY_TEST_INVENTORY: "true"
          CI_TEST_SELECTION: ${{ needs.changes.outputs.test_selection }}
@@ -365,7 +457,13 @@ jobs:
        env:
          CODE_REQUIRED: ${{ needs.changes.outputs.code }}
          SHARD_RESULT: ${{ needs.test-shard.result }}
-        run: test "${CODE_REQUIRED}" != "true" || test "${SHARD_RESULT}" = "success"
+          SOURCE_REQUIRED: ${{ needs.changes.outputs.tier == 'fast' }}
+          SOURCE_RESULT: ${{ needs.test-source.result }}
+        run: |
+          if [ "${CODE_REQUIRED}" = "true" ]; then
+            test "${SHARD_RESULT}" = "success" || exit 1
+            test "${SOURCE_REQUIRED}" != "true" || test "${SOURCE_RESULT}" = "success"
+          fi

  conformance-fixtures:
    name: Source-built conformance fixtures
@@ -450,7 +548,9 @@ jobs:
      - uses: actions/setup-node@49933ea5288caeca8642d1e84afbd3f7d6820020 # v4.4.0
        with:
          node-version-file: .nvmrc
-          cache: npm
+      - name: Cache npm package downloads
+        id: npm-cache
+        uses: ./.github/actions/npm-cache
      - name: Verify native runtime identity
        env:
          EXPECTED_CI_PLATFORM: ${{ matrix.platform }}
@@ -472,11 +572,6 @@ jobs:
      - run: npm run verify:windows-native:artifact
      - name: Verify NativeAOT diagnostic applicability on this host
        run: npm run test:focused -- tests/evaluation/package/nativeAotDiagnostics.test.ts
-      - name: Resolve npm cache before verifier HOME isolation
-        id: npm-cache
-        run: |
-          cache_path="$(npm config get cache)"
-          echo "path=${cache_path}" >> "${GITHUB_OUTPUT}"
      - run: npm run verify:package
        env:
          npm_config_cache: ${{ steps.npm-cache.outputs.path }}
@@ -506,7 +601,9 @@ jobs:
      - uses: actions/setup-node@49933ea5288caeca8642d1e84afbd3f7d6820020 # v4.4.0
        with:
          node-version-file: .nvmrc
-          cache: npm
+      - name: Cache npm package downloads
+        id: npm-cache
+        uses: ./.github/actions/npm-cache
      - run: npm ci
      - uses: actions/download-artifact@d3f86a106a0bac45b974a628896c90dbdf5c8093 # v4.3.0
        with:
@@ -554,7 +651,9 @@ jobs:
      - uses: actions/setup-node@49933ea5288caeca8642d1e84afbd3f7d6820020 # v4.4.0
        with:
          node-version-file: .nvmrc
-          cache: npm
+      - name: Cache npm package downloads
+        id: npm-cache
+        uses: ./.github/actions/npm-cache
      - run: npm ci
      - name: Run the public CLI/MCP reconstruction journey
        env:
@@ -583,7 +682,9 @@ jobs:
      - uses: actions/setup-node@49933ea5288caeca8642d1e84afbd3f7d6820020 # v4.4.0
        with:
          node-version-file: .nvmrc
-          cache: npm
+      - name: Cache npm package downloads
+        id: npm-cache
+        uses: ./.github/actions/npm-cache
      - run: npm ci
      - uses: actions/download-artifact@d3f86a106a0bac45b974a628896c90dbdf5c8093 # v4.3.0
        with:
@@ -703,7 +804,9 @@ jobs:
      - uses: actions/setup-node@49933ea5288caeca8642d1e84afbd3f7d6820020 # v4.4.0
        with:
          node-version-file: .nvmrc
-          cache: npm
+      - name: Cache npm package downloads
+        id: npm-cache
+        uses: ./.github/actions/npm-cache
      - run: npm ci
      - run: npm run docs:check
      - name: Generation leaves authored files unchanged
@@ -792,7 +895,9 @@ jobs:
        changes,
        static,
        build,
+        reverse-tools-quality,
        test-shard,
+        test-source,
        sqlite-runtime,
        test,
        conformance-fixtures,
@@ -23,6 +23,8 @@ jobs:
      - uses: actions/setup-node@49933ea5288caeca8642d1e84afbd3f7d6820020 # v4.4.0
        with:
          node-version-file: .nvmrc
-          cache: npm
+      - name: Cache npm package downloads
+        id: npm-cache
+        uses: ./.github/actions/npm-cache
      - run: npm ci
      - run: npm run docs:build
@@ -32,7 +32,9 @@ jobs:
      - uses: actions/setup-node@49933ea5288caeca8642d1e84afbd3f7d6820020 # v4.4.0
        with:
          node-version-file: .nvmrc
-          cache: npm
+      - name: Cache npm package downloads
+        id: npm-cache
+        uses: ./.github/actions/npm-cache
      - run: npm ci
      - uses: ./.github/actions/portable-build-cache
        with:
@@ -25,7 +25,9 @@ jobs:
      - uses: actions/setup-node@49933ea5288caeca8642d1e84afbd3f7d6820020 # v4.4.0
        with:
          node-version-file: .nvmrc
-          cache: npm
+      - name: Cache npm package downloads
+        id: npm-cache
+        uses: ./.github/actions/npm-cache
      - run: npm ci
      - uses: ./.github/actions/portable-build-cache
        with:
@@ -0,0 +1,76 @@
+name: Real upstream Cutter verification
+
+on:
+  workflow_dispatch:
+  repository_dispatch:
+    types: [real-cutter-upstream]
+
+permissions:
+  contents: read
+
+jobs:
+  windows-native-artifact:
+    name: Build Windows bridge security boundary
+    runs-on: ubuntu-latest
+    steps:
+      - uses: actions/checkout@11d5960a326750d5838078e36cf38b85af677262 # v4.4.0
+      - uses: actions/setup-node@49933ea5288caeca8642d1e84afbd3f7d6820020 # v4.4.0
+        with:
+          node-version-file: .nvmrc
+      - name: Install the cross compiler used by the Windows native lane
+        run: |
+          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
+        with:
+          name: cutter-windows-native
+          path: native/windows/build/
+          if-no-files-found: error
+
+  cutter-upstream:
+    name: Cutter upstream smoke (${{ matrix.platform }})
+    needs: windows-native-artifact
+    runs-on: ${{ matrix.runner }}
+    timeout-minutes: 20
+    strategy:
+      fail-fast: false
+      matrix:
+        include:
+          - platform: linux
+            runner: [self-hosted, cutter-upstream, linux]
+          - platform: macOS
+            runner: [self-hosted, cutter-upstream, macOS]
+          - platform: Windows
+            runner: [self-hosted, cutter-upstream, Windows]
+    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
+      - if: matrix.platform == 'Windows'
+        uses: actions/download-artifact@v4
+        with:
+          name: cutter-windows-native
+          path: native/windows/build/
+      - if: matrix.platform == 'Windows'
+        run: npm run verify:windows-native:artifact
+      - if: matrix.platform == 'Windows'
+        run: npm run build:cached
+      - if: matrix.platform == 'Windows'
+        run: npm run verify:windows-native
+      - name: Verify host and real Cutter bridge
+        env:
+          EXPECTED_PLATFORM: ${{ matrix.platform }}
+        run: npm run verify:cutter
+      - uses: actions/upload-artifact@v4
+        if: always()
+        with:
+          name: cutter-upstream-${{ matrix.platform }}
+          path: cutter-smoke.json
+          if-no-files-found: ignore
+          retention-days: 30
@@ -26,7 +26,9 @@ jobs:
      - uses: actions/setup-node@49933ea5288caeca8642d1e84afbd3f7d6820020 # v4.4.0
        with:
          node-version-file: .nvmrc
-          cache: npm
+      - name: Cache npm package downloads
+        id: npm-cache
+        uses: ./.github/actions/npm-cache
      - run: npm ci
      - uses: ./.github/actions/portable-build-cache
        with:
@@ -48,7 +48,9 @@ jobs:
      - uses: actions/setup-node@49933ea5288caeca8642d1e84afbd3f7d6820020 # v4.4.0
        with:
          node-version-file: .nvmrc
-          cache: npm
+      - name: Cache npm package downloads
+        id: npm-cache
+        uses: ./.github/actions/npm-cache
      - uses: actions/setup-go@40f1582b2485089dde7abd97c1529aa768e1baff # v5.6.0
        with:
          go-version: "1.26.1"
@@ -19,7 +19,9 @@ jobs:
      - uses: actions/setup-node@49933ea5288caeca8642d1e84afbd3f7d6820020 # v4.4.0
        with:
          node-version-file: .nvmrc
-          cache: npm
+      - name: Cache npm package downloads
+        id: npm-cache
+        uses: ./.github/actions/npm-cache
      - run: npm ci
      - name: Pack CLI for setup verification
        shell: bash
@@ -40,7 +42,9 @@ jobs:
      - uses: actions/setup-node@49933ea5288caeca8642d1e84afbd3f7d6820020 # v4.4.0
        with:
          node-version-file: .nvmrc
-          cache: npm
+      - name: Cache npm package downloads
+        id: npm-cache
+        uses: ./.github/actions/npm-cache
      - run: npm ci
      - name: Install packed CLI in isolation
        shell: bash
@@ -19,7 +19,9 @@ jobs:
      - uses: actions/setup-node@49933ea5288caeca8642d1e84afbd3f7d6820020 # v4.4.0
        with:
          node-version-file: .nvmrc
-          cache: npm
+      - name: Cache npm package downloads
+        id: npm-cache
+        uses: ./.github/actions/npm-cache
      - run: npm ci
      - name: Install packed CLI in isolation
        shell: bash
@@ -27,7 +27,9 @@ jobs:
      - uses: actions/setup-node@49933ea5288caeca8642d1e84afbd3f7d6820020 # v4.4.0
        with:
          node-version-file: .nvmrc
-          cache: npm
+      - name: Cache npm package downloads
+        id: npm-cache
+        uses: ./.github/actions/npm-cache
      - run: npm ci
      - uses: ./.github/actions/portable-build-cache
        with:
@@ -47,7 +47,9 @@ jobs:
      - uses: actions/setup-node@49933ea5288caeca8642d1e84afbd3f7d6820020 # v4.4.0
        with:
          node-version-file: .nvmrc
-          cache: npm
+      - name: Cache npm package downloads
+        id: npm-cache
+        uses: ./.github/actions/npm-cache
      - 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
@@ -29,7 +29,9 @@ jobs:
      - uses: actions/setup-node@49933ea5288caeca8642d1e84afbd3f7d6820020 # v4.4.0
        with:
          node-version-file: .nvmrc
-          cache: npm
+      - name: Cache npm package downloads
+        id: npm-cache
+        uses: ./.github/actions/npm-cache
      - run: npm ci
      - uses: ./.github/actions/portable-build-cache
        with:
@@ -40,7 +40,9 @@ jobs:
      - uses: actions/setup-node@49933ea5288caeca8642d1e84afbd3f7d6820020 # v4.4.0
        with:
          node-version-file: .nvmrc
-          cache: npm
+      - name: Cache npm package downloads
+        id: npm-cache
+        uses: ./.github/actions/npm-cache
      - run: sudo apt-get update && sudo apt-get install -y sigrok-cli util-linux python3
      - run: npm ci
      - run: npm run build:cached
@@ -26,7 +26,9 @@ jobs:
      - uses: actions/setup-node@49933ea5288caeca8642d1e84afbd3f7d6820020 # v4.4.0
        with:
          node-version-file: .nvmrc
-          cache: npm
+      - name: Cache npm package downloads
+        id: npm-cache
+        uses: ./.github/actions/npm-cache
      - name: Use pinned package manager
        run: npm install --global npm@11.16.0
      - run: npm ci
@@ -21,12 +21,14 @@ jobs:
      - uses: actions/setup-node@49933ea5288caeca8642d1e84afbd3f7d6820020 # v4.4.0
        with:
          node-version-file: .nvmrc
-          cache: npm
+      - 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 }}
-      - run: npm run verify:browser:runtime
+      - run: env -u DBUS_SESSION_BUS_ADDRESS npm run verify:browser:runtime
        env:
          REA_BROWSER_EXECUTABLE: /usr/bin/google-chrome
          REA_BROWSER_NO_SANDBOX: "true"
@@ -21,7 +21,9 @@ jobs:
      - uses: actions/setup-node@49933ea5288caeca8642d1e84afbd3f7d6820020 # v4.4.0
        with:
          node-version-file: .nvmrc
-          cache: npm
+      - name: Cache npm package downloads
+        id: npm-cache
+        uses: ./.github/actions/npm-cache
      - run: npm ci
      - uses: ./.github/actions/portable-build-cache
        with:
@@ -27,7 +27,9 @@ jobs:
        with:
          node-version: ${{ matrix.node }}
          check-latest: true
-          cache: npm
+      - name: Cache npm package downloads
+        id: npm-cache
+        uses: ./.github/actions/npm-cache
      - run: npm ci
      - uses: actions/download-artifact@d3f86a106a0bac45b974a628896c90dbdf5c8093 # v4.3.0
        with:
@@ -33,6 +33,8 @@ codedb.snapshot
 htmlcov/
 .coverage
 coverage/
+.stryker-tmp/
+reports/mutation/
 .vitest/
 .vitest-reports/
 test-results.xml
@@ -1,6 +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)

 # REA: Reverse Engineer Anything

@@ -23,6 +24,12 @@

 <code>npx rea-agents setup</code>

+> [!IMPORTANT]
+> Get REA from [morluto/rea](https://github.com/morluto/rea),
+> [rea.tools](https://rea.tools/), or the [rea-agents npm package](https://www.npmjs.com/package/rea-agents).
+> Follow the installation and release links from these official sources. A similar
+> repository name or copied README does not establish that a download is official.
+
 <br />

 <img src="docs/assets/rea-hopper-analysis.png" alt="REA launching its analysis bridge inside Hopper while inspecting a native binary" width="1200" />
@@ -180,6 +187,11 @@ Static JavaScript and .NET inspection read the supplied files without running
 the application. Runtime capture runs or interacts with the selected target
 using your user permissions; each runtime guide describes its effects.

+For offline native-binary inspection, REA also supports caller-supplied GNU
+objdump and Rizin. See the [reverse-engineering tools guide](docs/reverse-engineering-tools.md)
+for provider setup, CLI commands, and Evidence limitations. REA does not install
+or update either provider.
+
 <a id="choosing-a-deep-analysis-provider"></a>

 Native formats and host support vary by provider. See
@@ -0,0 +1,816 @@
+"""REA's experimental local bridge plugin for Cutter.
+
+Install this file in Cutter's user plugin directory. The bridge is REA-owned
+IPC, not an upstream Cutter transport. It binds to loopback and requires a
+per-process random token stored in a user-only descriptor file.
+"""
+
+from __future__ import annotations
+
+import json
+import hashlib
+import hmac
+import os
+import secrets
+import socket
+import stat
+import tempfile
+import threading
+import time
+import uuid
+from pathlib import Path
+from typing import Any
+
+if os.name == "nt":
+    import ctypes
+    from ctypes import wintypes
+
+import cutter
+
+try:
+    from PySide6.QtCore import QObject, Signal
+except ImportError:
+    from PySide2.QtCore import QObject, Signal
+
+
+MAX_FRAME_BYTES = 16 * 1024 * 1024
+SOCKET_TIMEOUT_SECONDS = 20
+AUTH_FRAME_TIMEOUT_SECONDS = 1
+MAX_DESCRIPTOR_BYTES = 64 * 1024
+MAX_COMPLETION_RECEIPTS = 64
+
+
+def _windows_api() -> tuple[Any, Any]:
+    kernel32 = ctypes.WinDLL("kernel32", use_last_error=True)
+    advapi32 = ctypes.WinDLL("advapi32", use_last_error=True)
+    kernel32.GetCurrentProcess.restype = wintypes.HANDLE
+    kernel32.LocalFree.argtypes = [wintypes.HLOCAL]
+    kernel32.LocalFree.restype = wintypes.HLOCAL
+    kernel32.CloseHandle.argtypes = [wintypes.HANDLE]
+    kernel32.CloseHandle.restype = wintypes.BOOL
+    kernel32.GetFileAttributesW.argtypes = [wintypes.LPCWSTR]
+    kernel32.GetFileAttributesW.restype = wintypes.DWORD
+    kernel32.CreateDirectoryW.argtypes = [wintypes.LPCWSTR, wintypes.LPVOID]
+    kernel32.CreateDirectoryW.restype = wintypes.BOOL
+    kernel32.CreateFileW.argtypes = [wintypes.LPCWSTR, wintypes.DWORD, wintypes.DWORD,
+                                     wintypes.LPVOID, wintypes.DWORD, wintypes.DWORD, wintypes.HANDLE]
+    kernel32.CreateFileW.restype = wintypes.HANDLE
+    kernel32.WriteFile.argtypes = [wintypes.HANDLE, wintypes.LPVOID, wintypes.DWORD,
+                                   ctypes.POINTER(wintypes.DWORD), wintypes.LPVOID]
+    kernel32.WriteFile.restype = wintypes.BOOL
+    kernel32.FlushFileBuffers.argtypes = [wintypes.HANDLE]
+    kernel32.FlushFileBuffers.restype = wintypes.BOOL
+    kernel32.MoveFileExW.argtypes = [wintypes.LPCWSTR, wintypes.LPCWSTR, wintypes.DWORD]
+    kernel32.MoveFileExW.restype = wintypes.BOOL
+    kernel32.DeleteFileW.argtypes = [wintypes.LPCWSTR]
+    kernel32.DeleteFileW.restype = wintypes.BOOL
+    advapi32.OpenProcessToken.argtypes = [wintypes.HANDLE, wintypes.DWORD, ctypes.POINTER(wintypes.HANDLE)]
+    advapi32.OpenProcessToken.restype = wintypes.BOOL
+    advapi32.GetTokenInformation.argtypes = [wintypes.HANDLE, ctypes.c_int, wintypes.LPVOID,
+                                              wintypes.DWORD, ctypes.POINTER(wintypes.DWORD)]
+    advapi32.GetTokenInformation.restype = wintypes.BOOL
+    advapi32.ConvertSidToStringSidW.argtypes = [wintypes.LPVOID, ctypes.POINTER(wintypes.LPWSTR)]
+    advapi32.ConvertSidToStringSidW.restype = wintypes.BOOL
+    advapi32.GetNamedSecurityInfoW.argtypes = [wintypes.LPWSTR, ctypes.c_int, wintypes.DWORD,
+                                                ctypes.POINTER(wintypes.LPVOID), ctypes.POINTER(wintypes.LPVOID),
+                                                ctypes.POINTER(wintypes.LPVOID), ctypes.POINTER(wintypes.LPVOID),
+                                                ctypes.POINTER(wintypes.LPVOID)]
+    advapi32.GetNamedSecurityInfoW.restype = wintypes.DWORD
+    advapi32.GetSecurityDescriptorControl.argtypes = [wintypes.LPVOID, ctypes.POINTER(wintypes.WORD),
+                                                       ctypes.POINTER(wintypes.DWORD)]
+    advapi32.GetSecurityDescriptorControl.restype = wintypes.BOOL
+    advapi32.EqualSid.argtypes = [wintypes.LPVOID, wintypes.LPVOID]
+    advapi32.EqualSid.restype = wintypes.BOOL
+    advapi32.GetAclInformation.argtypes = [wintypes.LPVOID, wintypes.LPVOID, wintypes.DWORD, ctypes.c_int]
+    advapi32.GetAclInformation.restype = wintypes.BOOL
+    advapi32.GetAce.argtypes = [wintypes.LPVOID, wintypes.DWORD, ctypes.POINTER(wintypes.LPVOID)]
+    advapi32.GetAce.restype = wintypes.BOOL
+    advapi32.IsValidSid.argtypes = [wintypes.LPVOID]
+    advapi32.IsValidSid.restype = wintypes.BOOL
+    advapi32.GetLengthSid.argtypes = [wintypes.LPVOID]
+    advapi32.GetLengthSid.restype = wintypes.DWORD
+    advapi32.ConvertStringSecurityDescriptorToSecurityDescriptorW.argtypes = [
+        wintypes.LPCWSTR, wintypes.DWORD, ctypes.POINTER(wintypes.LPVOID), ctypes.POINTER(wintypes.DWORD),
+    ]
+    advapi32.ConvertStringSecurityDescriptorToSecurityDescriptorW.restype = wintypes.BOOL
+    return kernel32, advapi32
+
+
+def _windows_user_sid_text() -> str:
+    kernel32, advapi32 = _windows_api()
+    token = wintypes.HANDLE()
+    if not advapi32.OpenProcessToken(kernel32.GetCurrentProcess(), 0x0008, ctypes.byref(token)):
+        raise ctypes.WinError(ctypes.get_last_error())
+    try:
+        required = wintypes.DWORD()
+        advapi32.GetTokenInformation(token, 1, None, 0, ctypes.byref(required))
+        if required.value == 0:
+            raise ctypes.WinError(ctypes.get_last_error())
+        buffer = ctypes.create_string_buffer(required.value)
+        if not advapi32.GetTokenInformation(token, 1, buffer, required, ctypes.byref(required)):
+            raise ctypes.WinError(ctypes.get_last_error())
+
+        class SID_AND_ATTRIBUTES(ctypes.Structure):
+            _fields_ = [("Sid", wintypes.LPVOID), ("Attributes", wintypes.DWORD)]
+
+        class TOKEN_USER(ctypes.Structure):
+            _fields_ = [("User", SID_AND_ATTRIBUTES)]
+
+        sid = ctypes.cast(buffer, ctypes.POINTER(TOKEN_USER)).contents.User.Sid
+        text = wintypes.LPWSTR()
+        if not advapi32.ConvertSidToStringSidW(sid, ctypes.byref(text)):
+            raise ctypes.WinError(ctypes.get_last_error())
+        try:
+            return text.value
+        finally:
+            kernel32.LocalFree(text)
+    finally:
+        kernel32.CloseHandle(token)
+
+
+def _verify_windows_private_path(path: Path, directory: bool) -> None:
+    kernel32, advapi32 = _windows_api()
+    attributes = kernel32.GetFileAttributesW(str(path))
+    if attributes == 0xFFFFFFFF:
+        raise ctypes.WinError(ctypes.get_last_error())
+    if attributes & 0x400 or bool(attributes & 0x10) != directory:
+        raise RuntimeError("Cutter bridge path is a reparse point or has the wrong object type")
+
+    class ACL_SIZE_INFORMATION(ctypes.Structure):
+        _fields_ = [("AceCount", wintypes.DWORD), ("AclBytesInUse", wintypes.DWORD), ("AclBytesFree", wintypes.DWORD)]
+
+    class ACE_HEADER(ctypes.Structure):
+        _fields_ = [("AceType", wintypes.BYTE), ("AceFlags", wintypes.BYTE), ("AceSize", wintypes.WORD)]
+
+    class SID_HEADER(ctypes.Structure):
+        _fields_ = [("Revision", wintypes.BYTE), ("SubAuthorityCount", wintypes.BYTE),
+                    ("IdentifierAuthority", wintypes.BYTE * 6)]
+
+    class ACCESS_ALLOWED_ACE(ctypes.Structure):
+        _fields_ = [("Header", ACE_HEADER), ("Mask", wintypes.DWORD), ("SidStart", wintypes.DWORD)]
+
+    owner = wintypes.LPVOID()
+    dacl = wintypes.LPVOID()
+    descriptor = wintypes.LPVOID()
+    result = advapi32.GetNamedSecurityInfoW(
+        str(path), 1, 0x1 | 0x4, ctypes.byref(owner), None, ctypes.byref(dacl), None, ctypes.byref(descriptor),
+    )
+    if result != 0:
+        raise ctypes.WinError(result)
+    try:
+        control = wintypes.WORD()
+        revision = wintypes.DWORD()
+        if not advapi32.GetSecurityDescriptorControl(descriptor, ctypes.byref(control), ctypes.byref(revision)):
+            raise ctypes.WinError(ctypes.get_last_error())
+        if not control.value & 0x1000 or not dacl:
+            raise RuntimeError("Cutter bridge ACL must be present and protected from inheritance")
+
+        token = wintypes.HANDLE()
+        if not advapi32.OpenProcessToken(kernel32.GetCurrentProcess(), 0x0008, ctypes.byref(token)):
+            raise ctypes.WinError(ctypes.get_last_error())
+        try:
+            required = wintypes.DWORD()
+            advapi32.GetTokenInformation(token, 1, None, 0, ctypes.byref(required))
+            token_user = ctypes.create_string_buffer(required.value)
+            if not advapi32.GetTokenInformation(token, 1, token_user, required, ctypes.byref(required)):
+                raise ctypes.WinError(ctypes.get_last_error())
+
+            class SID_AND_ATTRIBUTES(ctypes.Structure):
+                _fields_ = [("Sid", wintypes.LPVOID), ("Attributes", wintypes.DWORD)]
+
+            class TOKEN_USER(ctypes.Structure):
+                _fields_ = [("User", SID_AND_ATTRIBUTES)]
+
+            current_sid = ctypes.cast(token_user, ctypes.POINTER(TOKEN_USER)).contents.User.Sid
+            if not advapi32.EqualSid(owner, current_sid):
+                raise RuntimeError("Cutter bridge path owner does not match the current user")
+        finally:
+            kernel32.CloseHandle(token)
+
+        info = ACL_SIZE_INFORMATION()
+        if not advapi32.GetAclInformation(dacl, ctypes.byref(info), ctypes.sizeof(info), 2) or info.AceCount != 2:
+            raise RuntimeError("Cutter bridge ACL must allow only the current user and SYSTEM")
+        expected_user = _windows_user_sid_text()
+        user_allowed = False
+        system_allowed = False
+        for index in range(info.AceCount):
+            ace_pointer = wintypes.LPVOID()
+            if not advapi32.GetAce(dacl, index, ctypes.byref(ace_pointer)):
+                raise ctypes.WinError(ctypes.get_last_error())
+            header = ctypes.cast(ace_pointer, ctypes.POINTER(ACE_HEADER)).contents
+            if header.AceType != 0 or header.AceSize < ctypes.sizeof(ACE_HEADER) + ctypes.sizeof(wintypes.DWORD) + ctypes.sizeof(SID_HEADER):
+                raise RuntimeError("Cutter bridge ACL contains a malformed or unsupported ACE")
+            ace_address = ctypes.cast(ace_pointer, ctypes.c_void_p).value
+            if ace_address is None:
+                raise RuntimeError("Cutter bridge ACL returned a null ACE")
+            sid_pointer = ctypes.c_void_p(ace_address + ctypes.sizeof(ACE_HEADER) + ctypes.sizeof(wintypes.DWORD))
+            sid_header = ctypes.cast(sid_pointer, ctypes.POINTER(SID_HEADER)).contents
+            sid_size = ctypes.sizeof(SID_HEADER) + sid_header.SubAuthorityCount * ctypes.sizeof(wintypes.DWORD)
+            if sid_size > header.AceSize - ctypes.sizeof(ACE_HEADER) - ctypes.sizeof(wintypes.DWORD) or not advapi32.IsValidSid(sid_pointer) or advapi32.GetLengthSid(sid_pointer) != sid_size:
+                raise RuntimeError("Cutter bridge ACL contains an invalid SID")
+            ace = ctypes.cast(ace_pointer, ctypes.POINTER(ACCESS_ALLOWED_ACE)).contents
+            sid_text = wintypes.LPWSTR()
+            if ace.Mask != 0x001F01FF or ace.Header.AceFlags != 0x03:
+                raise RuntimeError("Cutter bridge ACL contains an unsupported ACE")
+            if not advapi32.ConvertSidToStringSidW(sid_pointer, ctypes.byref(sid_text)):
+                raise ctypes.WinError(ctypes.get_last_error())
+            try:
+                if sid_text.value == expected_user:
+                    user_allowed = True
+                elif sid_text.value == "S-1-5-18":
+                    system_allowed = True
+                else:
+                    raise RuntimeError("Cutter bridge ACL allows an unrelated principal")
+            finally:
+                kernel32.LocalFree(sid_text)
+        if not user_allowed or not system_allowed:
+            raise RuntimeError("Cutter bridge ACL must allow only the current user and SYSTEM")
+    finally:
+        kernel32.LocalFree(descriptor)
+
+
+def _windows_security_attributes() -> tuple[Any, Any]:
+    kernel32, advapi32 = _windows_api()
+    sid = _windows_user_sid_text()
+    sddl = f"O:{sid}G:{sid}D:P(A;OICI;FA;;;{sid})(A;OICI;FA;;;SY)"
+    descriptor = wintypes.LPVOID()
+    if not advapi32.ConvertStringSecurityDescriptorToSecurityDescriptorW(sddl, 1, ctypes.byref(descriptor), None):
+        raise ctypes.WinError(ctypes.get_last_error())
+
+    class SECURITY_ATTRIBUTES(ctypes.Structure):
+        _fields_ = [("nLength", wintypes.DWORD), ("lpSecurityDescriptor", wintypes.LPVOID), ("bInheritHandle", wintypes.BOOL)]
+
+    attributes = SECURITY_ATTRIBUTES(ctypes.sizeof(SECURITY_ATTRIBUTES), descriptor, False)
+    return kernel32, (descriptor, attributes)
+
+
+def _ensure_windows_bridge_directory(directory: Path) -> None:
+    directory.parent.mkdir(parents=True, exist_ok=True)
+    if not directory.exists():
+        kernel32, (descriptor, attributes) = _windows_security_attributes()
+        try:
+            if not kernel32.CreateDirectoryW(str(directory), ctypes.byref(attributes)):
+                error = ctypes.get_last_error()
+                if error != 183:
+                    raise ctypes.WinError(error)
+        finally:
+            kernel32.LocalFree(descriptor)
+    _verify_windows_private_path(directory, True)
+
+
+def _ensure_posix_bridge_directory(directory: Path) -> None:
+    directory.mkdir(mode=0o700, parents=True, exist_ok=True)
+    if not hasattr(os, "O_NOFOLLOW") or not hasattr(os, "O_DIRECTORY"):
+        raise RuntimeError("Cutter bridge directory requires no-follow directory handles")
+    descriptor = os.open(directory, os.O_RDONLY | os.O_DIRECTORY | os.O_NOFOLLOW)
+    try:
+        directory_info = os.fstat(descriptor)
+        if directory_info.st_uid != os.getuid():
+            raise RuntimeError("Cutter bridge directory must be owned by the current user")
+        os.fchmod(descriptor, 0o700)
+        directory_info = os.fstat(descriptor)
+        if stat.S_IMODE(directory_info.st_mode) != 0o700:
+            raise RuntimeError("Cutter bridge directory must have mode 0700")
+    finally:
+        os.close(descriptor)
+
+
+def _write_windows_descriptor(directory: Path, destination: Path, payload: bytes) -> None:
+    if len(payload) > MAX_DESCRIPTOR_BYTES:
+        raise RuntimeError("Cutter bridge descriptor exceeds its byte limit")
+    kernel32, (security_descriptor, attributes) = _windows_security_attributes()
+    temporary = directory / f".{destination.name}.{uuid.uuid4().hex}.tmp"
+    handle = wintypes.HANDLE()
+    try:
+        handle = kernel32.CreateFileW(str(temporary), 0x40000000 | 0x00020000 | 0x00040000,
+                                      0x1 | 0x2 | 0x4, ctypes.byref(attributes), 1, 0x80, None)
+        if handle == wintypes.HANDLE(-1).value:
+            raise ctypes.WinError(ctypes.get_last_error())
+        _verify_windows_private_path(temporary, False)
+        if payload:
+            data = ctypes.create_string_buffer(payload)
+            written = wintypes.DWORD()
+            if not kernel32.WriteFile(handle, data, len(payload), ctypes.byref(written), None) or written.value != len(payload):
+                raise ctypes.WinError(ctypes.get_last_error())
+        if not kernel32.FlushFileBuffers(handle):
+            raise ctypes.WinError(ctypes.get_last_error())
+        kernel32.CloseHandle(handle)
+        handle = wintypes.HANDLE()
+        if not kernel32.MoveFileExW(str(temporary), str(destination), 0x1 | 0x8):
+            raise ctypes.WinError(ctypes.get_last_error())
+    finally:
+        if handle and handle != wintypes.HANDLE(-1).value:
+            kernel32.CloseHandle(handle)
+        kernel32.DeleteFileW(str(temporary))
+        kernel32.LocalFree(security_descriptor)
+
+
+def _bridge_directory() -> Path:
+    configured = os.environ.get("REA_CUTTER_BRIDGE_DIR")
+    if configured:
+        return Path(configured).expanduser()
+    local_app_data = os.environ.get("LOCALAPPDATA")
+    if os.name == "nt" and local_app_data:
+        return Path(local_app_data) / "REA" / "CutterBridge"
+    return Path.home() / ".cache" / "rea" / "cutter-bridge"
+
+
+class _RequestDispatcher(QObject):
+    dispatch = Signal(object, object)
+
+    def __init__(self) -> None:
+        super().__init__()
+        self.dispatch.connect(self._handle)
+        self.generation = 0
+        self.completed_commands: dict[str, str] = {}
+        self._receipt_lock = threading.Lock()
+        self.current_file = self._read_current_file()
+        try:
+            self.cutter_version = str(cutter.core().getVersionInformation())
+        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
+        except Exception:
+            return None
+
+    def _handle(self, request: dict[str, Any], reply: dict[str, Any]) -> None:
+        kind = request.get("kind")
+        request_id = request.get("request_id")
+        if kind in ("command", "ack"):
+            try:
+                valid_request_id = (
+                    isinstance(request_id, str)
+                    and str(uuid.UUID(request_id)) == request_id.lower()
+                )
+            except (ValueError, AttributeError):
+                valid_request_id = False
+            if not valid_request_id:
+                reply.update({
+                    "ok": False,
+                    "error": "invalid-command-request-id",
+                    "execution_state": "not_started",
+                })
+                request["_event"].set()
+                return
+        if kind == "ack":
+            with self._receipt_lock:
+                acknowledged = self.completed_commands.pop(request_id, None) is not None
+            reply.update({"ok": True, "acknowledged": acknowledged})
+            request["_event"].set()
+            return
+        if request["_cancelled"].is_set():
+            reply.update({
+                "ok": False,
+                "error": "request-cancelled-before-execution",
+                "execution_state": "not_started",
+            })
+            if kind == "command":
+                self._record_command_completion(request_id, "not_started")
+            request["_event"].set()
+            return
+        self._refresh_document_identity()
+        if kind == "status":
+            with self._receipt_lock:
+                completed_commands = [
+                    {"request_id": receipt_id, "execution_state": state}
+                    for receipt_id, state in self.completed_commands.items()
+                ]
+            reply.update({
+                "ok": True,
+                "session_id": request.get("session_id"),
+                "current_file": self.current_file,
+                "document_generation": self.generation,
+                "cutter_version": self.cutter_version,
+                "identity_status": "partial",
+                "completed_commands": completed_commands,
+            })
+            request["_event"].set()
+            return
+        if kind != "command":
+            reply.update({"ok": False, "error": "unsupported-request-kind"})
+            request["_event"].set()
+            return
+        if request.get("expected_generation") != self.generation:
+            reply.update({
+                "ok": False,
+                "error": "active-document-generation-changed",
+                "execution_state": "not_started",
+            })
+            self._record_command_completion(request_id, "not_started")
+            request["_event"].set()
+            return
+        command = request.get("command")
+        if not isinstance(command, str) or not command.strip():
+            reply.update({
+                "ok": False,
+                "error": "invalid-command",
+                "execution_state": "not_started",
+            })
+            self._record_command_completion(request_id, "not_started")
+            request["_event"].set()
+            return
+        reply["execution_state"] = "running"
+        try:
+            output = cutter.cmdj(command) if request.get("json") is True else cutter.cmd(command)
+            command_error = None
+        except Exception as error:
+            output = None
+            command_error = str(error)
+        self._refresh_document_identity()
+        if command_error is None:
+            reply.update({
+                "ok": True,
+                "output": output,
+                "current_file": self.current_file,
+                "document_generation": self.generation,
+                "cutter_version": self.cutter_version,
+                "identity_status": "partial",
+                "execution_state": "complete",
+            })
+            self._record_command_completion(request_id, "complete")
+        else:
+            reply.update({
+                "ok": False,
+                "error": command_error,
+                "current_file": self.current_file,
+                "document_generation": self.generation,
+                "cutter_version": self.cutter_version,
+                "identity_status": "partial",
+                "execution_state": "unknown",
+                "message": "The command raised an exception; it may have produced partial effects. Do not retry automatically.",
+            })
+            self._record_command_completion(request_id, "failed")
+        request["_event"].set()
+
+    def _record_command_completion(
+        self, request_id: str, execution_state: str
+    ) -> None:
+        with self._receipt_lock:
+            if request_id in self.completed_commands or len(self.completed_commands) < MAX_COMPLETION_RECEIPTS:
+                self.completed_commands[request_id] = execution_state
+
+    def can_accept_command(self) -> bool:
+        with self._receipt_lock:
+            return len(self.completed_commands) < MAX_COMPLETION_RECEIPTS
+
+    def _refresh_document_identity(self) -> None:
+        current_file = self._read_current_file()
+        if current_file != self.current_file:
+            self.current_file = current_file
+            self.generation += 1
+
+
+class _BridgeServer(threading.Thread):
+    def __init__(self, dispatcher: _RequestDispatcher) -> None:
+        super().__init__(name="rea-cutter-bridge", daemon=True)
+        self.dispatcher = dispatcher
+        self.server = socket.socket(socket.AF_INET, socket.SOCK_STREAM)
+        self.server.setsockopt(socket.SOL_SOCKET, socket.SO_REUSEADDR, 1)
+        self.server.bind(("127.0.0.1", 0))
+        self.server.listen(8)
+        self.server.settimeout(0.5)
+        self.port = self.server.getsockname()[1]
+        self.session_id = str(uuid.uuid4())
+        self.token = secrets.token_urlsafe(32)
+        self.poisoned = False
+        self.connection_ordinal = 0
+        self.directory = _bridge_directory()
+        self.descriptor = self.directory / f"cutter-{os.getpid()}-{self.session_id}.json"
+
+    def publish(self) -> None:
+        if os.name == "nt":
+            _ensure_windows_bridge_directory(self.directory)
+        else:
+            _ensure_posix_bridge_directory(self.directory)
+        if os.name == "nt" and (self.directory.is_symlink() or not self.directory.is_dir()):
+            raise RuntimeError("Cutter bridge directory must be a real directory")
+        payload = {
+            "session_id": self.session_id,
+            "pid": os.getpid(),
+            "host": "127.0.0.1",
+            "port": self.port,
+            "token": self.token,
+            "document_generation": self.dispatcher.generation,
+            "current_file": self.dispatcher.current_file,
+            "cutter_version": self.dispatcher.cutter_version,
+            "identity_status": "partial",
+        }
+        if os.name == "nt":
+            _write_windows_descriptor(self.directory, self.descriptor, json.dumps(payload).encode("utf-8"))
+            return
+        descriptor, temporary_path = tempfile.mkstemp(prefix=f".{self.descriptor.name}.", dir=self.directory)
+        try:
+            os.fchmod(descriptor, 0o600)
+            descriptor_info = os.fstat(descriptor)
+            if descriptor_info.st_uid != os.getuid() or stat.S_IMODE(descriptor_info.st_mode) & 0o077:
+                raise RuntimeError("Cutter bridge descriptor must be owned by the current user with mode 0600")
+            with os.fdopen(descriptor, "w", encoding="utf-8") as stream:
+                stream.write(json.dumps(payload))
+                stream.flush()
+                os.fsync(stream.fileno())
+            os.replace(temporary_path, self.descriptor)
+        except Exception:
+            try:
+                os.close(descriptor)
+            except OSError:
+                pass
+            try:
+                os.unlink(temporary_path)
+            except OSError:
+                pass
+            raise
+
+    def run(self) -> None:
+        try:
+            self.publish()
+        except Exception:
+            self.server.close()
+            return
+        while True:
+            try:
+                connection, _ = self.server.accept()
+            except socket.timeout:
+                if getattr(self, "stopping", False):
+                    break
+                continue
+            except OSError:
+                break
+            self.connection_ordinal += 1
+            connection_ordinal = self.connection_ordinal
+            with connection:
+                connection.settimeout(AUTH_FRAME_TIMEOUT_SECONDS)
+                authenticated = False
+                try:
+                    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
+                    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)
+                    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):
+                        self._send(connection, {"ok": False, "error": "authentication-failed"})
+                        continue
+                    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
+                    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)
+                except socket.timeout:
+                    if authenticated:
+                        self._send_authenticated(connection, {
+                            "ok": False,
+                            "error": "request-timeout",
+                            "execution_state": "unknown",
+                            "message": "The command may have produced partial or persistent effects. Do not retry automatically.",
+                        }, 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
+
+    @staticmethod
+    def _receive_frame(connection: socket.socket) -> 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")
+            remaining = deadline - time.monotonic()
+            if remaining <= 0:
+                raise socket.timeout("authentication frame timed out")
+            connection.settimeout(remaining)
+            part = connection.recv(min(65536, MAX_FRAME_BYTES + 1 - len(data)))
+            if not part:
+                raise ValueError("incomplete authentication frame")
+            data.extend(part)
+        line, remainder = bytes(data).split(b"\n", 1)
+        if len(line) > MAX_FRAME_BYTES or remainder:
+            raise ValueError("invalid framed request")
+        value = json.loads(line)
+        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:
+            return
+        try:
+            connection.sendall(encoded)
+        except OSError:
+            pass
+
+    def _encode_authenticated_response(self, value: dict[str, Any], client_nonce: str, server_nonce: str) -> bytes | 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)
+            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'
+
+    @staticmethod
+    def _truncated_response(value: dict[str, Any]) -> dict[str, Any]:
+        response: dict[str, Any] = {
+            "ok": value.get("ok") is True,
+            "output": None,
+            "execution_state": value.get("execution_state", "unknown"),
+            "output_truncated": True,
+            "message": "The command result exceeded the bridge frame limit; completion state is preserved and output was omitted.",
+        }
+        for key in ("current_file", "document_generation", "cutter_version", "identity_status"):
+            if key in value:
+                response[key] = value[key]
+        if value.get("ok") is not True and isinstance(value.get("error"), str):
+            response["error"] = value["error"]
+        return response
+
+    @staticmethod
+    def _send(connection: socket.socket, value: dict[str, Any]) -> None:
+        if not _within_response_limit(value):
+            value = {
+                "ok": False,
+                "error": "response-limit-exceeded",
+                "execution_state": value.get("execution_state", "unknown"),
+                "output_truncated": True,
+                "message": "The command completed or may have completed, but its result exceeded the bridge response limit. Do not retry automatically.",
+            }
+        parts: list[bytes] = []
+        encoded_size = 0
+        for part in json.JSONEncoder(ensure_ascii=False).iterencode(value):
+            encoded_part = part.encode("utf-8")
+            encoded_size += len(encoded_part)
+            if encoded_size > MAX_FRAME_BYTES:
+                fallback = {
+                    "ok": False,
+                    "error": "response-limit-exceeded",
+                    "execution_state": value.get("execution_state", "unknown"),
+                    "output_truncated": True,
+                    "message": "The command completed or may have completed, but its result exceeded the bridge response limit. Do not retry automatically.",
+                }
+                parts = [json.dumps(fallback, separators=(",", ":")).encode("utf-8")]
+                break
+            parts.append(encoded_part)
+        encoded = b"".join(parts)
+        try:
+            connection.sendall(encoded + b"\n")
+        except OSError:
+            # A disconnected caller must not terminate the bridge thread or
+            # prevent later commands from being served.
+            pass
+
+    def stop(self) -> None:
+        self.stopping = True
+        try:
+            self.server.close()
+        except OSError:
+            pass
+
+
+class ReaCutterPlugin(cutter.CutterPlugin):
+    name = "REA Local Bridge"
+    description = "Local REA analysis bridge for the active Cutter session"
+    version = "1.0"
+    author = "REA"
+
+    def __init__(self) -> None:
+        super().__init__()
+        self._dispatcher: _RequestDispatcher | None = None
+        self._bridge: _BridgeServer | None = None
+
+    def setupPlugin(self) -> None:
+        return
+
+    def setupInterface(self, main: Any) -> None:
+        self._dispatcher = _RequestDispatcher()
+        self._bridge = _BridgeServer(self._dispatcher)
+        self._bridge.start()
+
+    def terminate(self) -> None:
+        if self._bridge is not None:
+            self._bridge.stop()
+            self._bridge.join(timeout=2)
+            self._bridge = None
+        self._dispatcher = None
+
+
+def create_cutter_plugin() -> ReaCutterPlugin:
+    return ReaCutterPlugin()
+
+
+def _within_response_limit(value: Any) -> bool:
+    pending = [value]
+    seen: set[int] = set()
+    size = 0
+    while pending:
+        current = pending.pop()
+        if isinstance(current, str):
+            size += 2
+            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)
+        elif isinstance(current, dict):
+            identity = id(current)
+            if identity in seen:
+                return False
+            seen.add(identity)
+            size += 2 + 4 * len(current)
+            if size > MAX_FRAME_BYTES:
+                return False
+            pending.extend(current.keys())
+            pending.extend(current.values())
+        elif isinstance(current, (list, tuple)):
+            identity = id(current)
+            if identity in seen:
+                return False
+            seen.add(identity)
+            size += 2 + 2 * len(current)
+            if size > MAX_FRAME_BYTES:
+                return False
+            pending.extend(current)
+        else:
+            size += 16
+        if size > MAX_FRAME_BYTES:
+            return False
+    return True
@@ -81,6 +81,7 @@ import ghidra.program.model.listing.Function;
 import ghidra.program.model.listing.FunctionIterator;
 import ghidra.program.model.listing.Instruction;
 import ghidra.program.model.listing.InstructionIterator;
+import ghidra.program.model.listing.Listing;
 import ghidra.program.model.listing.Variable;
 import ghidra.program.model.mem.MemoryBlock;
 import ghidra.program.model.mem.MemoryBlockSourceInfo;
@@ -405,7 +406,9 @@ public final class ReaGhidraBridge extends HeadlessScript {
            case "xrefs" -> xrefs(request.params);
            case "analyze_function" -> {
                FunctionAnalysisObservation observation = analyzeFunctionObservation(request.params);
-                yield functionObservation(observation.dossier(), observation.limitations());
+                JsonObject result = functionObservation(observation.dossier(), observation.limitations());
+                result.add("relationship_coverage", observation.relationshipCoverage());
+                yield result;
            }
            case "set_address_name", "set_addresses_names" -> {
                if (!descriptor.transport.equals("unix-socket"))
@@ -1016,12 +1019,19 @@ public final class ReaGhidraBridge extends HeadlessScript {
        for (Function item : ordered) {
            result.add(canonicalAddress(item.getEntryPoint()));
        }
-        return functionObservation(
-            result,
-            noReturnCallLimitations(
+        if (callers) {
+            JsonObject observation = functionObservation(result, noReturnCallLimitations(
                currentProgram.getListing().getInstructions(function.getBody(), true)
-            )
+            ));
+            observation.add("relationship_coverage", callRelationshipCoverage(null,
+                unrepresentedIncomingCalls(function, observed)));
+            monitor.checkCancelled();
+            return observation;
+        }
+        CallBoundaryObservation calls = callBoundaryObservation(
+            function, currentProgram.getListing().getInstructions(function.getBody(), true), observed, false, null
        );
+        return callRelationshipObservation(result, calls);
    }

    private JsonObject procedureInfo(JsonObject params) throws Exception {
@@ -1071,18 +1081,10 @@ public final class ReaGhidraBridge extends HeadlessScript {
        result.addProperty("direction", direction);
        result.add("references", edges);
        result.addProperty("reference_kinds_available", true);
-        JsonArray unresolvedCalls = new JsonArray();
-        if (direction.equals("outgoing")) for (Instruction instruction : scanInstructions(function, Integer.MAX_VALUE).instructions) {
-            if (!instruction.getFlowType().isCall()) continue;
-            boolean resolved = false;
-            for (Reference reference : instruction.getReferencesFrom()) if (reference.getReferenceType().isCall()) resolved = true;
-            if (!resolved) {
-                JsonObject site = new JsonObject();
-                site.addProperty("address", canonicalAddress(instruction.getAddress()));
-                site.addProperty("reason", "Instruction decoder identifies a call but ReferenceManager supplies no call destination");
-                unresolvedCalls.add(site);
-            }
-        }
+        JsonArray unresolvedCalls = direction.equals("outgoing")
+            ? callBoundaryObservation(
+                function, currentProgram.getListing().getInstructions(function.getBody(), true), null, true, references
+            ).unresolvedCalls() : new JsonArray();
        result.add("unresolved_calls", unresolvedCalls);
        return result;
    }
@@ -1490,19 +1492,28 @@ public final class ReaGhidraBridge extends HeadlessScript {
        InstructionScan scan = scanInstructions(function, Integer.MAX_VALUE);
        DecompileResults decompilation = decompile(function);
        String pseudocode = decompiledText(decompilation);
+        JsonObject availability = new JsonObject();
+        availability.add("pseudocode", facetAvailability(
+            pseudocode != null,
+            function.isExternal() ? "Ghidra does not decompile external functions."
+                : function.getBody().isEmpty() ? "Ghidra has no function body to decompile."
+                : "Ghidra completed decompilation without returning pseudocode."
+        ));
        if (pseudocode == null) {
            pseudocode = "";
        }
-        List<Reference> incomingReferences = collectReferences(function, "incoming")
-            .stream()
+        List<Reference> bodyIncoming = collectReferences(function, "incoming");
+        List<Reference> incomingReferences = bodyIncoming.stream()
            .filter(reference -> !function.getBody().contains(reference.getFromAddress()))
            .toList();
        List<Reference> outgoingReferences = collectReferences(function, "outgoing");
        JsonArray incoming = referenceEdges(incomingReferences);
        JsonArray outgoing = referenceEdges(outgoingReferences);
-        JsonArray comments = comments(scan.instructions);
-        JsonArray callers = procedureIdentities(function.getCallingFunctions(monitor));
-        JsonArray callees = procedureIdentities(function.getCalledFunctions(monitor));
+        JsonArray comments = comments(function);
+        Set<Function> callerFunctions = function.getCallingFunctions(monitor);
+        JsonArray callers = procedureIdentities(callerFunctions);
+        Set<Function> calleeFunctions = function.getCalledFunctions(monitor);
+        JsonArray callees = procedureIdentities(calleeFunctions);
        JsonArray referencedStrings = referencedStrings(outgoingReferences);
        JsonArray referencedNames = referencedNames(outgoingReferences);
        JsonArray blocks = basicBlockValues(function);
@@ -1536,6 +1547,15 @@ public final class ReaGhidraBridge extends HeadlessScript {
            nativeApiBoundary(function, decompilation)
        );
        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")) {
+            availability.add(facet, facetAvailability(true, null));
+        }
+        availability.add("assembly", facetAvailability(
+            !scan.truncated(), "The function instruction scan was truncated."
+        ));
+        result.add("facet_availability", availability);
        JsonArray limitations = new JsonArray();
        limitations.add(
            "Unresolved computed or indirect flows without target addresses are not represented as reference edges."
@@ -1563,8 +1583,10 @@ public final class ReaGhidraBridge extends HeadlessScript {
                "Swift String literal decoding is implemented for AArch64 register pairs only; no Swift literals were decoded for this architecture."
            );
        }
-        List<String> boundaryLimitations = noReturnCallLimitations(scan.instructions);
-        for (String limitation : boundaryLimitations) {
+        CallBoundaryObservation calls = callBoundaryObservation(
+            function, scan.instructions, calleeFunctions, true, outgoingReferences);
+        result.add("unresolved_calls", calls.unresolvedCalls());
+        for (String limitation : calls.limitations()) {
            limitations.add(limitation);
        }
        if (decompilation != null && decompilation.getErrorMessage() != null &&
@@ -1572,7 +1594,14 @@ public final class ReaGhidraBridge extends HeadlessScript {
            limitations.add("Ghidra decompiler diagnostic: " + decompilation.getErrorMessage());
        }
        result.add("limitations", limitations);
-        return new FunctionAnalysisObservation(result, boundaryLimitations);
+        monitor.checkCancelled();
+        int missingCallers = 0;
+        for (Reference reference : bodyIncoming) {
+            monitor.checkCancelled();
+            if (unrepresentedCaller(reference, callerFunctions)) missingCallers++;
+        }
+        return new FunctionAnalysisObservation(result, calls.limitations(),
+            callRelationshipCoverage(calls, missingCallers));
    }

    private JsonObject nativeValueFlow(DecompileResults decompilation) throws Exception {
@@ -3276,54 +3305,202 @@ public final class ReaGhidraBridge extends HeadlessScript {
        return result;
    }

+    private static JsonObject facetAvailability(boolean available, String reason) {
+        JsonObject result = new JsonObject();
+        result.addProperty("available", available);
+        if (!available) result.addProperty("reason", reason);
+        return result;
+    }
+
+    private JsonObject callRelationshipObservation(JsonElement value, CallBoundaryObservation calls)
+            throws Exception {
+        monitor.checkCancelled();
+        JsonObject result = functionObservation(value, calls.limitations());
+        result.add("relationship_coverage", callRelationshipCoverage(calls, null));
+        return result;
+    }
+
+    private static JsonObject callRelationshipCoverage(
+            CallBoundaryObservation calls, Integer missingCallers) {
+        JsonObject coverage = completeResolvedCallCoverage();
+        Integer omitted = calls == null ? null : calls.unrepresentedCalleeCalls();
+        JsonArray directions = new JsonArray();
+        if (missingCallers != null && missingCallers > 0) directions.add("callers");
+        String callerLimitation = missingCallers != null && missingCallers > 0
+            ? "Ghidra's caller inventory omits sources of " + missingCallers +
+                " call reference(s) into this function's body." : null;
+        if ((calls != null && omitted == null) || (calls == null && missingCallers == null)) {
+            coverage.addProperty("status", "unknown");
+            String reason = calls == null ? "No call inventory was examined."
+                : "The callee inventory was not examined alongside decoded call targets.";
+            coverage.addProperty("reason", reason + (callerLimitation == null ? "" : " " + callerLimitation));
+            if (calls == null && missingCallers == null) directions.add("callers");
+            directions.add("callees");
+            coverage.add("limited_directions", directions);
+            return coverage;
+        }
+        JsonArray limitations = new JsonArray();
+        if (callerLimitation != null) limitations.add(callerLimitation);
+        if (omitted != null && omitted > 0) {
+            limitations.add("Ghidra's callee inventory omits targets of " + omitted +
+                " call reference(s) or decoded call(s) without destinations.");
+            directions.add("callees");
+        }
+        if (!limitations.isEmpty()) {
+            coverage.addProperty("status", "partial");
+            coverage.add("limitations", limitations);
+            coverage.add("limited_directions", directions);
+        }
+        return coverage;
+    }
+
+    private static JsonObject completeResolvedCallCoverage() {
+        JsonObject coverage = new JsonObject();
+        coverage.addProperty("status", "complete");
+        JsonArray diagnostics = new JsonArray();
+        diagnostics.add(
+            "Only resolved Ghidra call references are returned; unresolved computed or indirect calls remain unknown, while function classifications distinguish thunks and externals."
+        );
+        coverage.add("diagnostics", diagnostics);
+        return coverage;
+    }
+
+    private boolean unrepresentedCaller(Reference reference, Set<Function> callers) {
+        if (!reference.getReferenceType().isCall()) return false;
+        Function caller = containingFunction(reference.getFromAddress());
+        return caller == null || !callers.contains(caller);
+    }
+
+    private int unrepresentedIncomingCalls(Function function, Set<Function> callers)
+            throws Exception {
+        int[] omitted = { 0 };
+        visitReferences(function, "incoming", reference -> {
+            if (unrepresentedCaller(reference, callers)) omitted[0]++;
+        });
+        return omitted[0];
+    }
+
    private List<String> noReturnCallLimitations(Iterable<Instruction> instructions)
            throws Exception {
        List<String> limitations = new ArrayList<>();
        for (Instruction instruction : instructions) {
            monitor.checkCancelled();
-            if (!instruction.getFlowType().isCall() || !instruction.getFlowType().isTerminal()) {
-                continue;
-            }
-            Address[] destinations = instruction.getFlows();
-            StringBuilder limitation = new StringBuilder();
-            limitation.append("Ghidra Listing reports a terminal call at ")
-                .append(canonicalAddress(instruction.getAddress()));
-            if (destinations.length == 0) {
-                limitation.append(" with no resolved callee; its no-return basis is unknown.");
-            }
-            else {
-                limitation.append(" to ");
-                for (int index = 0; index < destinations.length; index++) {
-                    if (index > 0) limitation.append(", ");
-                    Address destination = destinations[index];
-                    limitation.append(canonicalAddress(destination));
-                    Function target = currentProgram.getFunctionManager().getFunctionAt(destination);
-                    if (target == null) {
-                        limitation.append(" (callee identity unknown)");
-                    }
-                    else {
-                        limitation.append(" ").append(procedureName(target));
-                        limitation.append(" (external=").append(target.isExternal());
-                        limitation.append(", thunk=").append(target.isThunk());
-                        limitation.append(", hasNoReturn=").append(target.hasNoReturn()).append(")");
-                    }
-                }
-                limitation.append("; fallthrough is excluded by Ghidra's flow model.");
-                Function target = destinations.length == 1
-                    ? currentProgram.getFunctionManager().getFunctionAt(destinations[0])
-                    : null;
-                if (target != null && target.hasNoReturn()) {
-                    limitation.append(" This records Ghidra's FunctionManager flag, not an independent verification that the callee cannot return.");
-                }
-                else {
-                    limitation.append(" The reason for the terminal flow is not established by this observation.");
-                }
-            }
-            limitations.add(limitation.toString());
+            String limitation = terminalCallLimitation(instruction);
+            if (limitation != null) limitations.add(limitation);
        }
        return limitations;
    }

+    private CallBoundaryObservation callBoundaryObservation(
+            Function function, Iterable<Instruction> instructions, Set<Function> returnedCallees,
+            boolean includeUnresolvedSites, List<Reference> observedReferences)
+            throws Exception {
+        List<String> limitations = new ArrayList<>();
+        CallBoundaryReferences calls = new CallBoundaryReferences(returnedCallees, includeUnresolvedSites);
+        if (observedReferences == null) visitReferences(function, "outgoing", calls::observe);
+        else for (Reference reference : observedReferences) {
+            monitor.checkCancelled();
+            calls.observe(reference);
+        }
+        for (Instruction instruction : instructions) {
+            monitor.checkCancelled();
+            String limitation = terminalCallLimitation(instruction);
+            if (limitation != null) limitations.add(limitation);
+            if (!instruction.getFlowType().isCall()) continue;
+            boolean destination = false;
+            for (Reference reference : instruction.getReferencesFrom()) {
+                if (!reference.getReferenceType().isCall()) continue;
+                destination = true;
+                break;
+            }
+            if (!destination) {
+                calls.unrepresented++;
+                calls.unresolved(instruction.getAddress(),
+                    "Instruction decoder identifies a call but ReferenceManager supplies no call destination");
+            }
+        }
+        return new CallBoundaryObservation(List.copyOf(limitations),
+            returnedCallees == null ? null : calls.unrepresented, calls.unresolvedCalls());
+    }
+
+    private final class CallBoundaryReferences {
+        private final Set<Function> returnedCallees;
+        private final TreeMap<Address, String> unresolved;
+        private int unrepresented;
+
+        CallBoundaryReferences(Set<Function> returnedCallees, boolean includeUnresolvedSites) {
+            this.returnedCallees = returnedCallees;
+            unresolved = includeUnresolvedSites ? new TreeMap<>() : null;
+        }
+
+        void observe(Reference reference) {
+            if (!reference.getReferenceType().isCall()) return;
+            // Reference identity follows the same containing-function owner as referenceEdge.
+            Function target = containingFunction(reference.getToAddress());
+            if (returnedCallees != null && (target == null || !returnedCallees.contains(target)))
+                unrepresented++;
+            if (target == null) unresolved(reference.getFromAddress(),
+                "A call destination has no known containing Ghidra procedure");
+        }
+
+        void unresolved(Address address, String reason) {
+            if (unresolved != null) unresolved.put(address, reason);
+        }
+
+        JsonArray unresolvedCalls() {
+            if (unresolved == null) return null;
+            JsonArray result = new JsonArray();
+            for (var item : unresolved.entrySet()) {
+                JsonObject site = new JsonObject();
+                site.addProperty("address", canonicalAddress(item.getKey()));
+                site.addProperty("reason", item.getValue());
+                result.add(site);
+            }
+            return result;
+        }
+    }
+
+    private String terminalCallLimitation(Instruction instruction) {
+        if (!instruction.getFlowType().isCall() || !instruction.getFlowType().isTerminal())
+            return null;
+        Address[] destinations = instruction.getFlows();
+        StringBuilder limitation = new StringBuilder();
+        limitation.append("Ghidra Listing reports a terminal call at ")
+            .append(canonicalAddress(instruction.getAddress()));
+        if (destinations.length == 0) {
+            limitation.append(" with no resolved callee; its no-return basis is unknown.");
+        }
+        else {
+            limitation.append(" to ");
+            for (int index = 0; index < destinations.length; index++) {
+                if (index > 0) limitation.append(", ");
+                Address destination = destinations[index];
+                limitation.append(canonicalAddress(destination));
+                Function target = currentProgram.getFunctionManager().getFunctionAt(destination);
+                if (target == null) {
+                    limitation.append(" (callee identity unknown)");
+                }
+                else {
+                    limitation.append(" ").append(procedureName(target));
+                    limitation.append(" (external=").append(target.isExternal());
+                    limitation.append(", thunk=").append(target.isThunk());
+                    limitation.append(", hasNoReturn=").append(target.hasNoReturn()).append(")");
+                }
+            }
+            limitation.append("; fallthrough is excluded by Ghidra's flow model.");
+            Function target = destinations.length == 1
+                ? currentProgram.getFunctionManager().getFunctionAt(destinations[0])
+                : null;
+            if (target != null && target.hasNoReturn()) {
+                limitation.append(" This records Ghidra's FunctionManager flag, not an independent verification that the callee cannot return.");
+            }
+            else {
+                limitation.append(" The reason for the terminal flow is not established by this observation.");
+            }
+        }
+        return limitation.toString();
+    }
+
    private static String renderAssembly(List<Instruction> instructions) {
        return String.join("\n", renderAssemblyLines(instructions));
    }
@@ -3347,15 +3524,27 @@ public final class ReaGhidraBridge extends HeadlessScript {
    private List<Reference> collectReferences(
            Function function,
            String direction) throws Exception {
+        TreeMap<String, Reference> observed = new TreeMap<>();
+        visitReferences(function, direction, reference -> addReference(observed, reference));
+        List<Reference> result = new ArrayList<>(observed.values());
+        result.sort(REFERENCE_ORDER);
+        return result;
+    }
+
+    @FunctionalInterface
+    private interface ReferenceVisitor {
+        void accept(Reference reference) throws Exception;
+    }
+
+    private void visitReferences(Function function, String direction, ReferenceVisitor visitor)
+            throws Exception {
        // References can target instruction interiors or originate in embedded
        // data. Use actual body ownership, including an external's exact entry.
-        AddressSet scope = new AddressSet(function.getBody());
-        scope.add(function.getEntryPoint());
+        AddressSet scope = functionObservationScope(function);
        boolean outgoing = direction.equals("outgoing");
        AddressIterator addresses = outgoing
            ? currentProgram.getReferenceManager().getReferenceSourceIterator(scope, true)
            : currentProgram.getReferenceManager().getReferenceDestinationIterator(scope, true);
-        TreeMap<String, Reference> observed = new TreeMap<>();
        while (addresses.hasNext()) {
            monitor.checkCancelled();
            Address address = addresses.next();
@@ -3363,7 +3552,7 @@ public final class ReaGhidraBridge extends HeadlessScript {
                for (Reference reference : currentProgram.getReferenceManager()
                        .getReferencesFrom(address)) {
                    monitor.checkCancelled();
-                    addReference(observed, reference);
+                    if (!reference.isEntryPointReference()) visitor.accept(reference);
                }
            }
            else {
@@ -3371,19 +3560,14 @@ public final class ReaGhidraBridge extends HeadlessScript {
                    .getReferencesTo(address);
                while (iterator.hasNext()) {
                    monitor.checkCancelled();
-                    addReference(observed, iterator.next());
+                    Reference reference = iterator.next();
+                    if (!reference.isEntryPointReference()) visitor.accept(reference);
                }
            }
        }
-        List<Reference> result = new ArrayList<>(observed.values());
-        result.sort(REFERENCE_ORDER);
-        return result;
    }

    private static void addReference(TreeMap<String, Reference> observed, Reference reference) {
-        if (reference.isEntryPointReference()) {
-            return;
-        }
        String key = canonicalAddress(reference.getFromAddress()) + "\u0000" +
            canonicalAddress(reference.getToAddress()) + "\u0000" +
            reference.getReferenceType().getName() + "\u0000" +
@@ -3392,6 +3576,12 @@ public final class ReaGhidraBridge extends HeadlessScript {
        observed.putIfAbsent(key, reference);
    }

+    private static AddressSet functionObservationScope(Function function) {
+        AddressSet scope = new AddressSet(function.getBody());
+        scope.add(function.getEntryPoint());
+        return scope;
+    }
+
    private JsonArray referenceEdges(List<Reference> references) {
        JsonArray result = new JsonArray();
        for (Reference reference : references) {
@@ -3545,23 +3735,28 @@ public final class ReaGhidraBridge extends HeadlessScript {
        return result;
    }

-    private static JsonArray comments(List<Instruction> instructions) {
+    private JsonArray comments(Function function) throws Exception {
+        Listing listing = currentProgram.getListing();
+        AddressSet scope = functionObservationScope(function);
+        AddressIterator addresses = listing.getCommentAddressIterator(scope, true);
        TreeMap<String, JsonObject> observed = new TreeMap<>();
-        for (Instruction instruction : instructions) {
+        while (addresses.hasNext()) {
+            monitor.checkCancelled();
+            Address address = addresses.next();
            for (CommentType type : CommentType.values()) {
-                String text = instruction.getComment(type);
+                String text = listing.getComment(type, address);
                if (text == null || text.isEmpty()) {
                    continue;
                }
                String kind = type == CommentType.EOL ? "inline" : "comment";
                JsonObject item = new JsonObject();
-                item.addProperty("address", canonicalAddress(instruction.getAddress()));
+                item.addProperty("address", canonicalAddress(address));
                item.addProperty("kind", kind);
                item.addProperty("text", text);
-                String key = canonicalAddress(instruction.getAddress()) + "\u0000" +
+                String key = canonicalAddress(address) + "\u0000" +
                    kind + "\u0000" + type.name() + "\u0000" + text;
                observed.put(key, item);
-                    }
+            }
        }
        JsonArray result = new JsonArray();
        observed.values().forEach(result::add);
@@ -4497,7 +4692,8 @@ 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) {}
+    private record FunctionAnalysisObservation(JsonObject dossier, List<String> limitations, JsonObject relationshipCoverage) {}
+    private record CallBoundaryObservation(List<String> limitations, Integer unrepresentedCalleeCalls, JsonArray unresolvedCalls) {}
    private record SwiftStringLiteralDecode(JsonArray literals, JsonArray omissions) {}
    private record SwiftLiteralLayout(
        String representation,
@@ -644,7 +644,7 @@ def _analyze_function(document, params):
            "end": _hex(_basic_block_end(procedure, block)),
            "successors": sorted(set(successors), key=lambda value: int(value, 16)),
        })
-    pseudo = _pseudocode(document, procedure) or ""
+    pseudo = _pseudocode(document, procedure)
    assembly_lines = _assembly(procedure).splitlines()
    callers = sorted((_procedure_identity(item) for item in procedure.getAllCallerProcedures()), key=lambda item: int(item["address"], 16))
    callees = sorted((_procedure_identity(item) for item in procedure.getAllCalleeProcedures()), key=lambda item: int(item["address"], 16))
@@ -706,7 +706,7 @@ def _analyze_function(document, params):
            "signature": procedure.signatureString(),
            "locals": _procedure_locals(procedure),
        },
-        "pseudocode": pseudo,
+        "pseudocode": pseudo or "",
        "assembly": assembly_lines,
        "comments": comments,
        "callers": callers, "callees": callees,
@@ -718,6 +718,12 @@ def _analyze_function(document, params):
        "basic_blocks": blocks,
        "native_api": None,
        "native_value_flow": None,
+        "facet_availability": {
+            "pseudocode": {"available": True} if pseudo is not None else _unavailable("Hopper did not supply pseudocode."),
+            **{facet: {"available": True} for facet in (
+                "assembly", "comments", "callers", "callees", "references", "strings_names", "cfg"
+            )},
+        },
        "limitations": [
            "Native CallReference classifications are observed where available; detailed reference flags remain unknown.",
            "Hopper's public Python API does not expose equivalent external or thunk classification in this dossier.",
@@ -63,6 +63,7 @@ export default defineConfig({
          { text: "Ghidra DOS analysis", link: "/ghidra-dos" },
          { text: "Ghidra raw binary", link: "/raw-binary" },
          { text: "Ghidra NativeAOT", link: "/ghidra-nativeaot" },
+          { text: "Ghidra PowerPC", link: "/ghidra-ppc" },
          { text: "Ghidra semantics", link: "/ghidra-provider" },
        ],
      },
@@ -33,10 +33,40 @@ unverified.
 ## Supply tools

 Use a **full JDK 17 or newer** (including its compiler module) and
-**jadx-headless-mcp 0.7.1**. REA runs its packaged Java metadata bridge in
-source-file mode against the unmodified engine JAR. REA does not install Java,
-an Android SDK, JADX, a device service or an emulator. Set the path in the shell
-or MCP server environment:
+[**jadx-headless-mcp 0.7.1**](https://github.com/1013503897/jadx-headless-mcp/releases/download/v0.7.1/jadx-headless-mcp-0.7.1-all.jar).
+Verify the downloaded JAR's SHA-256 against
+[REA's pinned release metadata](https://github.com/morluto/rea/blob/main/src/android/JadxRelease.ts)
+before using it. The following download and verification commands work for npm
+package consumers as well as repository checkouts; the repository-only
+`npm run fixtures:android` helper is not included in the npm package.
+
+On macOS or Linux, choose a destination and verify the pinned digest:
+
+```sh
+mkdir -p "$HOME/.local/share/rea/android" &&
+cd "$HOME/.local/share/rea/android" &&
+curl -fL -o jadx-headless-mcp-0.7.1-all.jar https://github.com/1013503897/jadx-headless-mcp/releases/download/v0.7.1/jadx-headless-mcp-0.7.1-all.jar &&
+if command -v sha256sum >/dev/null; then
+  echo '6e5eacf500b64292bfb73c49797c1958f6ee44646e43e868039ae7feb573ff75  jadx-headless-mcp-0.7.1-all.jar' | sha256sum -c -
+else
+  echo '6e5eacf500b64292bfb73c49797c1958f6ee44646e43e868039ae7feb573ff75  jadx-headless-mcp-0.7.1-all.jar' | shasum -a 256 -c -
+fi &&
+export REA_JADX_MCP_JAR="$PWD/jadx-headless-mcp-0.7.1-all.jar"
+```
+
+On Windows, run the equivalent in PowerShell:
+
+```powershell
+$jar = Join-Path $HOME '.local\share\rea\android\jadx-headless-mcp-0.7.1-all.jar'
+New-Item -ItemType Directory -Force (Split-Path $jar) | Out-Null
+Invoke-WebRequest -Uri 'https://github.com/1013503897/jadx-headless-mcp/releases/download/v0.7.1/jadx-headless-mcp-0.7.1-all.jar' -OutFile $jar
+if ((Get-FileHash -Algorithm SHA256 $jar).Hash.ToLowerInvariant() -ne '6e5eacf500b64292bfb73c49797c1958f6ee44646e43e868039ae7feb573ff75') { throw 'JADX JAR SHA-256 does not match REA pinned metadata' }
+$env:REA_JADX_MCP_JAR = $jar
+```
+
+REA runs its packaged Java metadata bridge in source-file mode against the
+unmodified engine JAR. REA does not install Java, an Android SDK, JADX, a device
+service or an emulator. Set the path in the shell or MCP server environment:

 ```sh
 export REA_JADX_MCP_JAR=/absolute/path/jadx-headless-mcp-0.7.1-all.jar
@@ -38,6 +38,10 @@ graph TD
        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"]
        DOTNET["src/dotnet/<br/>Managed static metadata provider<br/>PE/CLI metadata inventory, members, IL"]
+        OBJDUMP["src/objdump/<br/>GNU objdump command profile"]
+        RIZIN["src/rizin/<br/>Static command profile and persistent debugger sessions"]
+        GDB["src/gdb/<br/>Persistent GDB/MI sessions<br/>owned process lifecycle and target state"]
+
        FIRMWARE["src/firmware/<br/>Binwalk / Unblob adapters<br/>owned processes, report normalization, private extraction"]
        ANDROID["src/android/<br/>Static APK adapter<br/>owned SDK stdio, pinned headless JADX, normalization"]
        REFERENCE["src/reference/<br/>Filesystem reference-source reader<br/>bounded traversal, validation, errors"]
@@ -50,7 +54,7 @@ graph TD
    end

    subgraph "Contracts Layer"
-        CONTRACTS["src/contracts/<br/>Caller-visible schemas for analysis, managed, browser, Electron, application, and session tools"]
+        CONTRACTS["src/contracts/<br/>Caller-visible schemas for analysis, reverse engineering, managed, browser, Electron, application, and session tools"]
        ERRORS["src/contracts/errorSchemas.ts<br/>Shared analysis error schemas<br/>CLI diagnostics, generated error catalog"]
    end

@@ -122,6 +126,14 @@ graph TD
    INSPECTOR --> DOMAIN
    APP --> DOTNET
    DOTNET --> DOMAIN
+    COMPOSITION --> OBJDUMP
+    OBJDUMP --> DOMAIN
+    COMPOSITION --> RIZIN
+    RIZIN --> DOMAIN
+    CLI --> GDB
+    SERVER --> GDB
+    GDB --> PROCESS
+    GDB --> DOMAIN
    HOPPER --> HOPPER_BRIDGE
    GHIDRA --> GHIDRA_BRIDGE
    HOPPER --> PROCESS
@@ -265,6 +265,21 @@ The user-level MCP paths follow the editor's VS Code layout. Trae's official
 documentation describes the raw JSON editor without specifying those paths;
 loading this registration in Trae remains unverified.

+## WorkBuddy
+
+For [WorkBuddy](https://www.workbuddy.cn/docs/workbuddy/From-Beginner-to-Expert-Guide/Function-Description/MCP-Guide), setup writes `mcpServers.rea` into the
+global `~/.workbuddy/mcp.json` (a standard `mcpServers` JSON file at the user
+home root on every platform). Entries carry `command`, `args`, and `env`;
+stdio is the default transport. WorkBuddy discovers personal skills from
+`~/.workbuddy/skills` (`%userprofile%\.workbuddy\skills` on Windows), so
+setup installs its skill copy there.
+
+The global MCP path is a plain JSON file; setup preserves every other server
+and top-level key when it upserts `rea`. WorkBuddy also reads a project-level
+`<project>/.workbuddy/mcp.json`, which setup does not touch: it registers the
+user-level file only. Loading this registration and real REA tool calls
+through WorkBuddy are verified.
+
 ## ZCode

 For [ZCode](https://zcode.z.ai/en/docs/mcp-services), setup registers `rea` in the nested `mcp.servers` table of the
@@ -0,0 +1,90 @@
+# PowerPC ELF analysis through Ghidra
+
+This lane extends native ELF analysis to 32-bit big-endian PowerPC; it does
+not execute PowerPC code. It is the first milestone of
+[#1741](https://github.com/morluto/rea/issues/1741). GameCube/Wii DOL
+executables, PPC64 and other PowerPC variants are separate follow-ups.
+
+## Admission boundary
+
+ELF `EM_PPC` (20) is retained as the provider-neutral `powerpc` family. The
+target also retains class, byte order, `e_type` and raw `e_flags`. Recognition
+does not mean every provider or PowerPC variant is supported. `EM_PPC64` (21)
+is not recognized as this family.
+
+The Ghidra lane admits ELF32 `ET_EXEC`, big endian. ELF64, little-endian,
+relocatable and shared objects are refused with a specific unsupported
+constraint. `EF_PPC_EMB` declares an embedded variant and is refused; the
+remaining `e_flags` bits only declare relocation conventions and do not change
+instruction decoding.
+
+## Bound language
+
+Ghidra's ELF loader selects `PowerPC:BE:32:e500:default` for a plain `EM_PPC`
+header with `e_flags` 0. e500 is the embedded Book-E variant whose SPE unit
+replaces the classic floating-point unit, so it misinterprets ordinary
+750-family code. Nothing in the header distinguishes the two.
+
+The lane therefore imports with `PowerPC:BE:32:default` and compiler spec
+`default`, commits that choice in the analysis profile as
+`language_selection: "provider-pinned-v1"`, and checks the language Ghidra
+reports after import. A different language set through
+`REA_GHIDRA_LANGUAGE_ID` (and optionally `REA_GHIDRA_COMPILER_SPEC_ID`)
+replaces the lane default and is admitted against the installed language
+catalog as usual; for example `PowerPC:BE:32:e500` for genuine e500 code.
+
+PowerPC header facts are committed in the analysis profile so interpretation
+changes invalidate profile-bound snapshots. The support-lane identity is
+versioned independently of file identity. Evidence, lifecycle outputs and saved
+snapshots preserve the `powerpc` family. Hopper and IDA adapters do not
+implicitly gain PowerPC support. Windows Ghidra P0 remains PE x86/x86-64 only.
+
+## Use
+
+With a caller-installed Ghidra and compatible JDK on a supported Linux/macOS host:
+
+```bash
+rea analyze /absolute/path/to/ppc.elf --provider ghidra --json
+rea function /absolute/path/to/ppc.elf function_name --provider ghidra --json
+```
+
+MCP uses the existing tools, including `open_binary` with an absolute path and
+`provider_id: "ghidra"`. No PowerPC-specific tool catalog or emulator is
+introduced.
+
+## Verification
+
+The source-owned freestanding fixture is `tests/conformance/c/ppc.c`. The
+optional cross-target lane requires a caller-supplied Clang with the PowerPC
+target, LLD, GNU `readelf`, Ghidra and its compatible JDK; ordinary host-native
+checks do not acquire these tools. `REA_PPC_CLANG` selects Clang and
+`REA_PPC_READELF` selects GNU readelf (for example
+`/opt/homebrew/opt/binutils/bin/readelf` on macOS). These are test-time
+prerequisites, not REA runtime providers.
+
+```bash
+GHIDRA_INSTALL_DIR=/absolute/path/to/ghidra npm run verify:ghidra:ppc
+```
+
+The lane compiles the fixture for a 750-family core (`-mcpu=750`) and never
+executes it. GNU readelf independently supplies header facts and symbol
+addresses, compared with the production REA target resolver before provider
+startup. Its version, bounded command output and selected tool path are
+retained in the verification report. The parser deliberately recognizes only
+this fixture's GNU output, rejecting missing, ambiguous or unexpected fields
+rather than substituting defaults.
+
+Existing public MCP operations check Ghidra's **observed** language and
+compiler specification and loaded-file hashes; an e500 import fails the lane.
+The source-owned `ppc-probe.S` supplies fixed `addi` and conditional-branch
+encodings: verification checks bytes, length, immediate and direct destination,
+independent of decompiler variable names. Global and string bytes are checked at
+independently reported symbol addresses.
+
+CLI/MCP discovery and decompilation, direct calls, target/profile/Evidence
+identity, snapshot round-trip and final process cleanup remain covered.
+Pseudocode presence is a liveness check, not proof of semantic equivalence.
+Instruction decoding and reader agreement do not establish runtime or complete
+floating-point semantics. Parser/profile tests are narrower checks, not
+substitutes for this real-provider run. Report the actual lanes run and leave
+unexecuted verification pending.
@@ -288,11 +288,12 @@ hosted connector; see [its configuration](client-configuration.md#other-client-s
 | Qoder              | `qoder`          |
 | Factory Droid      | `droid`          |
 | Trae               | `trae`           |
+| WorkBuddy          | `workbuddy`      |
 | ZCode              | `zcode`          |
 | DeepSeek Harness   | `dsh`            |

-`dsh` and `kimi` require a release containing them. The registry check above does
-not cover them.
+`dsh`, `kimi`, and `workbuddy` require a release containing them. The registry
+check above does not cover them.

 See [client configuration and troubleshooting](client-configuration.md) for
 configuration paths, profiles, trust policies, and schema-limit workarounds.
@@ -493,7 +494,13 @@ copying license secrets into logs, or killing unrelated Hopper processes.
 REA connects to an existing Ghidra installation on Linux x64/arm64, macOS x64/arm64,
 or experimental Windows x64 P0.
 It accepts Ghidra 12.1.x and the 64-bit full JDK declared by that installation's
-`application.java.min` and `application.java.max`. Current 12.1 releases require
+`application.java.min` and `application.java.max`. The JDK's reported platform must also
+map to the same native decompiler directory as the host: a JDK running under
+emulation (for example x86_64 under Rosetta on Apple silicon) or reporting a
+non-host `os.name` fails the doctor
+even when it is an otherwise accepted 64-bit full JDK. Select a matching JDK
+through `JAVA_HOME`; adding another decompiler in the JVM-selected directory
+does not satisfy this host-validated admission rule. Current 12.1 releases require
 JDK 21 or newer and set no maximum. The bridge is verified with Ghidra 12.1.4
 and JDK 21. Each installation must include the native decompiler for the host
 architecture in `Ghidra/Features/Decompiler/os/<platform>/` or the corresponding
@@ -556,6 +563,11 @@ registrations. See [Windows Ghidra P0](windows-ghidra-p0.md) for provider diagno
 Doctor validates the platform, architecture, Ghidra 12.1.x application version,
 `support/analyzeHeadless` or `support/analyzeHeadless.bat`, the installation's
 Java major range, 64-bit JDK bitness, and the presence of `javac`/`javac.exe`.
+It also checks that the reported JVM `os.arch` and `os.name` map to the same native decompiler
+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.
 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,
@@ -209,21 +209,34 @@ native/runtime questions retain their own authority and coverage.
 Names and semantic labels do not establish durable member identity. Comparison
 preserves observed differences, structural inferences, and unknowns separately.

-Cross-version matching first prefers exact CIL/signature identity. When that
-does not match, it pairs an exact declared type, method name, and raw signature
-before trying structural body shape. The exact-signature key uses names only
-as part of that full tuple; names alone never select a pair. Duplicate tuples
-remain ambiguous rather than being paired by token order. Fields follow the
-same exact tiers. The raw signature is exact whether or not REA decoded it, so
-undecoded signatures pair by that tuple too, but they never enter structural
-rounds. An unmatched member is therefore reported as `unknown`, not added or
-removed, when its own signature was not decoded or when the other side has an
-unpaired member, one-sided or ambiguous, with the same declared type and name
-whose signature was not decoded.
+Cross-version matching first pairs members of the same declared type and name:
+methods by exact CIL/signature identity, then by raw signature, then by
+structural body shape; fields by raw signature. Exact CIL/signature identity
+requires equal raw CIL digests and decoded fingerprints. Equal decoded
+fingerprints alone cannot establish an exact match: they omit switch targets
+and collapse distinct NaN encodings. Only then may a remap pair a
+renamed or moved member across declared types or names: a method by exact
+CIL/signature identity or structural body shape, a field by exact decoded
+signature. A remap never takes a member whose own declared type and name remain
+unpaired on the other side, so a member is never taken from its own changed
+counterpart. A member whose declaring type is unresolved could belong to any
+type, so it pairs by name only when no other member shares its exact
+CIL/signature key, never by structural shape alone, and any remap group it
+joins stays ambiguous. The exact-signature key uses names only as part of that
+full tuple; names alone never select a pair. Duplicate tuples remain ambiguous
+rather than being paired by token order. The raw signature is exact whether or
+not REA decoded it, so undecoded signatures pair by that tuple too, but they
+never enter structural rounds. An unmatched member is therefore reported as
+`unknown`, not added or removed, when its own signature was not decoded or when
+the other side has an unpaired member, one-sided or ambiguous, with the same
+declared type and name whose signature was not decoded.

 For a matched method, an unavailable or partial body makes body-shape facets
-unknown while preserving observed signature differences. Structural identity
-compares normalized signatures, the limited decoded-CIL tuple fingerprint,
+unknown while preserving observed signature differences. For complete bodies,
+a difference in either the raw CIL digest or the decoded fingerprint is reported
+in the `cil` dimension, including reordered switch targets with equal decoded
+fingerprints. Structural identity compares normalized signatures, the limited
+decoded-CIL tuple fingerprint,
 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.
@@ -360,8 +373,8 @@ 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
 exception semantics, and does not replace the raw CIL digest. Cross-build
-comparison uses the value only as one unique-only structural tier and reports
-ambiguity when that evidence is insufficient.
+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
 `[["nop","none",null],["ret","none",null]]` and has digest
@@ -58,7 +58,13 @@ side declares a required facet unavailable. An empty legacy facet without a
 coverage declaration is also unknown. A producer that actually observed an
 empty facet can declare it available so observed absence remains comparable.
 Call comparison additionally retains the Evidence relationship coverage
-checks. IDA declares the facets its selected adapter cannot provide; an empty
+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
+calls still prevent equality conclusions for those dimensions. Analysis profiles
+version these declarations so older snapshots are not reused as current results.
+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,
 identities, manifests and other fixed metadata before authentication. Imports
@@ -187,7 +193,8 @@ after target content was read.
 ## Retained application Evidence inputs

 `analyze_javascript_application`, `inspect_binary_layout`, `inspect_artifact`,
-`inspect_wasm_artifact`, and `inspect_macho` accept `detail`:
+`inspect_wasm_artifact`, `inspect_macho`, and `decode_interface_builder` accept
+`detail`:
 `complete` (default) returns the complete analysis Evidence; `summary` records
 that complete Evidence in the session and returns the derived
 `inspect_analysis_view` summary Evidence instead. Its
@@ -270,6 +277,17 @@ symbols (Electron Framework: 2.5 MB of imports and exports, 3.3 KB as a
 summary), so select `detail: "summary"` on the first call. Delivery detail
 never reaches the native tools or changes the source Evidence ID.

+Completed `decode_interface_builder` Evidence supports a summary and
+`view: {kind: "interface-builder", collection: "nodes", offset: 0, limit: 64}`.
+Collections are `documents`, `nodes`, `edges`, and `coverage`; for nodes and
+edges an optional `document` keeps rows whose evidence names that exact decoded
+`relative_path`, and an undecoded path is refused. The summary keeps every
+decoded document, the provider, node-kind, edge-relation and coverage-status
+counts, and whether the decoded graph is incomplete. Each page repeats `source_truncated`. A complete
+app graph often exceeds the default MCP response budget (XLD.app: 6.7 MB), so
+select `detail: "summary"` on the first call. Delivery detail never reaches the
+decoder or changes the source Evidence ID.
+
 `trace_application_feature`, `trace_javascript_semantics`,
 `compare_application_versions`, `compare_source_to_bundle`, and
 `compare_javascript_export_shapes` accept complete inline application Evidence
@@ -319,7 +319,16 @@ complete observation.
  digest are preserved; key interpretation follows the pinned plist decoder.
 - `decode_interface_builder` / `rea decode-interface-builder <app>` reads
  compiled storyboard and nib resources into objects, outlets, actions,
-  connections and controller/class names. `verify:interface-builder` compiles
+  connections and controller/class names. A full app graph is often several
+  megabytes (XLD.app: 6.7 MB, about 2M tokens, which also exceeds the default
+  MCP response budget), so MCP callers should pass `detail: "summary"`. The
+  summary keeps every decoded document and node/edge/coverage counts and
+  retains the complete Evidence; page rows with `inspect_analysis_view`
+  `view: {kind: "interface-builder", collection: "nodes", document:
+  "Contents/Resources/Base.lproj/MainMenu.nib", offset: 0, limit: 64}`.
+  `document` selects rows whose evidence names that exact
+  `documents[].relative_path` (nodes and edges only); `source_truncated` keeps
+  graph incompleteness visible on every page. `verify:interface-builder` compiles
  a source-owned AppKit XIB with `ibtool`. Storyboards require an installed
  iOS platform; unsupported archive forms remain explicit.
  XML plist archives accept UTF-8 and BOM-marked UTF-16 in either byte order.
@@ -1,6 +1,7 @@
 <div align="center">

 [English](../../README.md) · [简体中文](README_zh.md) · [繁體中文](README_zh-TW.md) · [日本語](README_ja.md) · [한국어](README_ko.md) · [Türkçe](README_tr.md) · [Русский](README_ru.md) · [Tiếng Việt](README_vi.md) · [ไทย](README_th.md) · [Deutsch](README_de.md) · [Español](README_es.md) · [Français](README_fr.md) · [Українська](README_uk.md) · [Polski](README_pl.md) · [Português (Brasil)](README_pt-BR.md) · **العربية** · [فارسی](README_fa.md) · [Bahasa Indonesia](README_id.md) · [Italiano](README_it.md)
+· [Ελληνικά](README_el.md)

 # REA: هندسة عكسية لأي شيء

@@ -19,10 +20,14 @@

 **[الموقع](https://rea.tools/ar/) · [الأدلة](https://rea.tools/ar/guides/) · [دراسات الحالة](https://rea.tools/ar/showcase/)**

-[البدء السريع](#البدء-السريع) · [كيف يعمل REA](#كيف-يعمل-rea) · [ما الذي يمكنك تحليله](#ما-الذي-يمكنك-تحليله) · [دراسات الحالة](#دراسات-الحالة) · [الأسئلة الشائعة](#الأسئلة-الشائعة) · [التوثيق](#التوثيق)
+[البدء السريع](#البدء-السريع) · [كيف يعمل REA](#كيف-يعمل-rea) · [ما الذي يمكنك تحليله](#ما-الذي-يمكنك-تحليله) · [دراسات الحالة](#دراسات-الحالة) · [الأسئلة الشائعة](#الأسئلة-الشائعة) · [أدوات GNU objdump وRizin](../reverse-engineering-tools.md) · [التوثيق](#التوثيق)

 <code>npx rea-agents setup</code>

+> [!IMPORTANT]
+> المصادر الرسمية لـ REA: [morluto/rea](https://github.com/morluto/rea), [rea.tools](https://rea.tools/), [rea-agents](https://www.npmjs.com/package/rea-agents) (npm).
+> استخدم روابط التثبيت والإصدارات المنشورة في هذه المصادر الرسمية. تشابه اسم المستودع أو نسخ ملف README لا يثبت أن التنزيل رسمي.
+
 <br />

 <img src="../assets/rea-hopper-analysis.png" alt="REA يشغّل جسر التحليل داخل Hopper لفحص ملف ثنائي للشيفرة الأصلية" width="1200" />
@@ -1,6 +1,7 @@
 <div align="center">

 [English](../../README.md) · [简体中文](README_zh.md) · [繁體中文](README_zh-TW.md) · [日本語](README_ja.md) · [한국어](README_ko.md) · [Türkçe](README_tr.md) · [Русский](README_ru.md) · [Tiếng Việt](README_vi.md) · [ไทย](README_th.md) · **Deutsch** · [Español](README_es.md) · [Français](README_fr.md) · [Українська](README_uk.md) · [Polski](README_pl.md) · [Português (Brasil)](README_pt-BR.md) · [العربية](README_ar.md) · [فارسی](README_fa.md) · [Bahasa Indonesia](README_id.md) · [Italiano](README_it.md)
+· [Ελληνικά](README_el.md)

 # REA: Alles reverse-engineeren

@@ -19,10 +20,14 @@

 **[Website](https://rea.tools/de/) · [Anleitungen](https://rea.tools/de/guides/) · [Fallstudien](https://rea.tools/de/showcase/)**

-[Schnellstart](#schnellstart) · [So funktioniert REA](#so-funktioniert-rea) · [Was du analysieren kannst](#was-du-analysieren-kannst) · [Fallstudien](#fallstudien) · [Häufige Fragen](#häufige-fragen) · [Dokumentation](#dokumentation)
+[Schnellstart](#schnellstart) · [So funktioniert REA](#so-funktioniert-rea) · [Was du analysieren kannst](#was-du-analysieren-kannst) · [Fallstudien](#fallstudien) · [Häufige Fragen](#häufige-fragen) · [GNU-objdump- und Rizin-Tools](../reverse-engineering-tools.md) · [Dokumentation](#dokumentation)

 <code>npx rea-agents setup</code>

+> [!IMPORTANT]
+> Offizielle Quellen für REA: [morluto/rea](https://github.com/morluto/rea), [rea.tools](https://rea.tools/), [rea-agents](https://www.npmjs.com/package/rea-agents) (npm).
+> Folge den Installations- und Release-Links aus diesen offiziellen Quellen. Ein ähnlicher Repository-Name oder eine kopierte README belegt nicht, dass ein Download offiziell ist.
+
 <br />

 <img src="../assets/rea-hopper-analysis.png" alt="REA startet seine Analysebrücke in Hopper und untersucht eine native Binärdatei" width="1200" />
@@ -0,0 +1,316 @@
+<div align="center">
+
+[English](../../README.md) · [简体中文](README_zh.md) · [繁體中文](README_zh-TW.md) · [日本語](README_ja.md) · [한국어](README_ko.md) · [Türkçe](README_tr.md) · [Русский](README_ru.md) · [Tiếng Việt](README_vi.md) · [ไทย](README_th.md) · [Deutsch](README_de.md) · [Español](README_es.md) · [Français](README_fr.md) · [Українська](README_uk.md) · [Polski](README_pl.md) · [Português (Brasil)](README_pt-BR.md) · [العربية](README_ar.md) · [فارسی](README_fa.md) · [Bahasa Indonesia](README_id.md) · [Italiano](README_it.md) · **Ελληνικά**
+
+# REA: Reverse Engineer Anything
+
+### Ένα MCP για αντίστροφη μηχανική δυαδικών αρχείων, εφαρμογών και της συμπεριφοράς τους κατά την εκτέλεση.
+
+**Βρήκες μια λειτουργία που σου αρέσει; Μάθε πώς λειτουργεί, μέχρι το επίπεδο του δυαδικού κώδικα.**
+
+[![Έκδοση npm](https://img.shields.io/npm/v/rea-agents?style=flat-square&color=cb3837)](https://www.npmjs.com/package/rea-agents)
+[![Κατάλογος εργαλείων MCP](https://img.shields.io/badge/MCP-tool_catalog-5c4ee5?style=flat-square)](../mcp-contracts.md#generated-catalog)
+[![Απαιτήσεις Node.js](https://img.shields.io/badge/Node.js-requirements-339933?style=flat-square&logo=nodedotjs&logoColor=white)](#τι-μπορείς-να-αναλύσεις)
+[![skills.sh](https://skills.sh/b/morluto/rea?style=flat-square)](https://skills.sh/morluto/rea/reverse-engineer-anything)
+[![Άδεια MIT](https://img.shields.io/badge/license-MIT-f4c430?style=flat-square)](../../LICENSE)
+[![Discord](https://img.shields.io/discord/1556595354999332884?logo=discord&logoColor=white&label=Discord&color=5865F2)](https://discord.gg/GkcryMnJDM)
+
+<a href="https://trendshift.io/repositories/82054?utm_source=repository-badge&amp;utm_medium=badge&amp;utm_campaign=badge-repository-82054" target="_blank" rel="noopener noreferrer"><img src="https://trendshift.io/api/badge/repositories/82054" alt="morluto%2Frea | Trendshift" width="250" height="55"/></a>
+
+**[Ιστότοπος](https://rea.tools/) · [Οδηγοί](https://rea.tools/guides/) · [Μελέτες περίπτωσης](https://rea.tools/showcase/)**
+
+[Γρήγορη εκκίνηση](#γρήγορη-εκκίνηση) · [Πώς λειτουργεί το REA](#πώς-λειτουργεί-το-rea) · [Τι μπορείς να αναλύσεις](#τι-μπορείς-να-αναλύσεις) · [Μελέτες περίπτωσης](#μελέτες-περίπτωσης) · [Συχνές ερωτήσεις](#συχνές-ερωτήσεις) · [Τεκμηρίωση](#τεκμηρίωση)
+
+<code>npx rea-agents setup</code>
+
+> [!IMPORTANT]
+> Κατέβασε το REA από το [morluto/rea](https://github.com/morluto/rea),
+> το [rea.tools](https://rea.tools/) ή το [πακέτο rea-agents στο npm](https://www.npmjs.com/package/rea-agents).
+> Ακολούθησε τους συνδέσμους εγκατάστασης και εκδόσεων από αυτές τις επίσημες πηγές.
+> Ένα παρόμοιο όνομα αποθετηρίου ή ένα αντιγραμμένο README δεν αποδεικνύει ότι
+> ένα αρχείο προς λήψη προέρχεται από επίσημη πηγή.
+
+<br />
+
+<img src="../assets/rea-hopper-analysis.png" alt="Το REA ξεκινά τη γέφυρα ανάλυσής του μέσα στο Hopper κατά την εξέταση ενός δυαδικού αρχείου με κώδικα μηχανής" width="1200" />
+
+<br /><br />
+
+<table aria-label="Κοινότητα REA">
+<tr>
+<td align="center" width="360">
+  <a href="https://discord.gg/GkcryMnJDM">
+    <img src="../assets/discord.svg" height="42" alt="Discord" /><br />
+    <strong>Έλα στην κοινότητα αντίστροφης μηχανικής</strong>
+  </a><br />
+  <sub>Discord · Ερωτήσεις και απαντήσεις · Παρουσίαση έργων</sub>
+</td>
+</tr>
+</table>
+
+<br />
+
+</div>
+
+---
+
+Είδες σε μια εφαρμογή μια λειτουργία που θα ήθελες στο δικό σου προϊόν; Ζήτησε από τον πράκτορά σου (agent) να τη διερευνήσει με το REA. Μπορεί να εξετάσει την εφαρμογή χωρίς τον πηγαίο κώδικά της, να εξηγήσει πώς λειτουργεί η συγκεκριμένη δυνατότητα, να παρουσιάσει τα τεκμήρια και να υλοποιήσει μια αντίστοιχη έκδοση για το έργο σου.
+
+Το REA συνδέει τον πράκτορά σου με εργαλεία εξέτασης δυαδικών αρχείων με κώδικα μηχανής, εφαρμογών JavaScript και Electron, assemblies .NET και ιστοτόπων. Μπορείς να χρησιμοποιήσεις τα ίδια εργαλεία και από το τερματικό. Η ανάλυση εκτελείται τοπικά και τα αποτελέσματα περιλαμβάνουν τα τεκμήρια και τους περιορισμούς στους οποίους βασίζεται κάθε συμπέρασμα.
+
+Η διαδικασία εγκατάστασης καταχωρίζει το REA στον agent σου και εγκαθιστά τις αντίστοιχες οδηγίες εργασίας. Για ανάλυση κώδικα μηχανής μπορείς να χρησιμοποιήσεις μια υπάρχουσα εγκατάσταση του Hopper, του Ghidra ή του IDA. Η διαδικασία εγκατάστασης μπορεί επίσης να εγκαταστήσει το Hopper, εφόσον το εγκρίνεις. Η στατική ανάλυση JavaScript δεν χρειάζεται μηχανή ανάλυσης κώδικα μηχανής.
+
+> **[Επισκέψου τον ιστότοπο του REA](https://rea.tools/)** για οδηγίες εγκατάστασης, εικονογραφημένους οδηγούς και πραγματικές μελέτες περίπτωσης.
+
+## Γρήγορη εκκίνηση
+
+### Ρύθμιση του agent σου
+
+Αφού εγκαταστήσεις το Node.js και το npm, εκτέλεσε:
+
+```bash
+npx rea-agents setup
+```
+
+Επίλεξε τους agents σου, έλεγξε τις προτεινόμενες αλλαγές και ενέκρινέ τες. Η διαδικασία προσθέτει τον MCP server του REA και τις αντίστοιχες οδηγίες εργασίας, κρατώντας αντίγραφα ασφαλείας των υφιστάμενων ρυθμίσεων. Στη συνέχεια, επανεκκίνησε τον agent σου.
+
+Υποστηρίζονται τα Claude Code, Codex, Cursor, Gemini CLI, Grok Build και [άλλοι agents](../installation.md#supported-agents). Για τη ρύθμιση των παρόχων ανάλυσης και τη μη αυτόματη καταχώριση MCP, δες τον οδηγό [εγκατάστασης και ρύθμισης](../installation.md).
+
+### Ζήτησε από τον agent σου
+
+```text
+Διερεύνησε πώς λειτουργεί η αναζήτηση στην εφαρμογή Notes, δείξε μου τα τεκμήρια και υλοποίησε μια παρόμοια λειτουργία για το έργο μου.
+```
+
+Αντικατάστησε το Notes με την εφαρμογή που θέλεις να εξετάσεις και προσδιόρισε τη λειτουργία που θέλεις να κατανοήσεις.
+
+### Χρήση από το τερματικό
+
+Εξέτασε τον αποσυμπιεσμένο φάκελο μιας εφαρμογής JavaScript/Electron ή ένα αρχείο ASAR:
+
+```bash
+npx -y rea-agents@latest analyze-javascript-application /absolute/path/to/app --json
+```
+
+Το αποτέλεσμα περιλαμβάνει modules, εισαγωγές, τα όρια μεταξύ των μερών της εφαρμογής Electron και τα αντίστοιχα τεκμήρια. Αντικατάστησε τη διαδρομή με τη δική σου, για παράδειγμα `"D:/apps/example"` στα Windows.
+
+Για να εγκαταστήσεις την εντολή `rea` για τακτική χρήση:
+
+```bash
+npm install --global rea-agents
+rea --help
+```
+
+Για ανάλυση κώδικα μηχανής, ρύθμισε πρώτα έναν πάροχο ανάλυσης. Ο [οδηγός CLI και Evidence](../cli.md) περιγράφει τις σχετικές εντολές, την επιλογή παρόχου, τα στιγμιότυπα και τη χρήση σε scripts.
+
+### Ενημέρωση του REA
+
+Το REA εξελίσσεται γρήγορα και οι νέες εκδόσεις περιλαμβάνουν συχνές διορθώσεις σφαλμάτων. Διατήρησε την εγκατάστασή σου ενημερωμένη.
+
+Για CLI εγκατεστημένο με `npm install --global` ή `bun add --global`:
+
+```bash
+rea update
+```
+
+Οι ενημερώσεις καθολικής εγκατάστασης με Bun απαιτούν Bun 1.2.17 ή νεότερο. Αν χρησιμοποιείς παλαιότερη έκδοση, αναβάθμισε το Bun χειροκίνητα πριν εκτελέσεις το `rea update`.
+
+Για να ανανεώσεις τις καταχωρίσεις των agents και το skill, εκτέλεσε την εντολή ρύθμισης που εμφανίζει η ενημέρωση.
+
+Αν χρησιμοποιείς `npx`, ενημέρωσε τη ρύθμιση του agent με:
+
+```bash
+npx rea-agents@latest setup
+```
+
+Έλεγξε τις αλλαγές και επανεκκίνησε τον agent σου. Για μεμονωμένες εντολές CLI, χρησιμοποίησε `npx rea-agents@latest` και στη συνέχεια την εντολή που θέλεις. Με Bun, χρησιμοποίησε `bunx rea-agents@latest` αντί για `npx rea-agents@latest`.
+
+## Πώς λειτουργεί το REA
+
+Ο agent σου καλεί το REA μέσω MCP για να εξετάσει τον στόχο και να ακολουθήσει τον σχετικό κώδικα. Το REA επιστρέφει ευρήματα μαζί με τα τεκμήριά τους. Ο agent τα χρησιμοποιεί για να θέσει πρόσθετες ερωτήσεις, να εξηγήσει τη συμπεριφορά ή να γράψει και να δοκιμάσει μια υλοποίηση. Οι εντολές CLI χρησιμοποιούν τις ίδιες ροές εργασίας.
+
+![Ροή διερεύνησης του REA: ο agent σου θέτει ένα ερώτημα για έναν τοπικό στόχο, το REA τον εξετάζει και ακολουθεί τον κώδικά του με εργαλεία ανάλυσης, και ο agent χρησιμοποιεί τον κώδικα, τις αναφορές και τα άγνωστα στοιχεία που επιστρέφονται για να εξηγήσει, να υλοποιήσει και να δοκιμάσει.](../../website/public/assets/figures/rea-investigation-flow.svg)
+
+[Άνοιγμα του διαγράμματος σε πλήρες μέγεθος](../../website/public/assets/figures/rea-investigation-flow.svg).
+
+<a id="current-status"></a>
+
+## Τι μπορείς να αναλύσεις
+
+Το REA απαιτεί Node.js 22.x (>=22.19), 24.x (>=24.11) ή 26+, καθώς και npm. Τα πρόσθετα εργαλεία και τα υποστηριζόμενα συστήματα εξαρτώνται από τον στόχο:
+
+| Στόχος                                 | Τι επιστρέφει το REA                                                                                                        | Απαιτήσεις και οδηγός                                                                                                               |
+| -------------------------------------- | --------------------------------------------------------------------------------------------------------------------------- | ----------------------------------------------------------------------------------------------------------------------------------- |
+| Δυαδικά αρχεία με κώδικα μηχανής       | Ψευδοκώδικα, κώδικα assembly, συμβολοσειρές, σύμβολα, κλήσεις και αναφορές                                                  | Hopper, Ghidra ή IDA· [ανάλυση κώδικα μηχανής](https://rea.tools/guides/native/)                                                    |
+| Δομή ELF χωρίς εκτέλεση                | Ενότητες, τμήματα, αρχικά σύμβολα/εγγραφές μετεγκατάστασης και ενδείξεις πιθανών μέτρων προστασίας από στατική ανάλυση      | pwntools που παρέχει ο χρήστης σε Linux x64· [διαγνωστικά δυαδικών αρχείων](../binary-diagnostics.md)                               |
+| Δυαδικά αρχεία Go                      | Ενσωματωμένη έκδοση μεταγλωττιστή, modules, αντικαταστάσεις και ρυθμίσεις μεταγλώττισης                                     | Ενσωματωμένη στατική εξέταση ELF/PE/thin Mach-O· δεν απαιτείται εγκατάσταση του Go                                                  |
+| Δυαδικά αρχεία Rust                    | Ενσωματωμένα ονόματα, εκδόσεις, πηγές και σχέσεις εξάρτησης των crates από το cargo-auditable                               | Ενσωματωμένη στατική εξέταση ELF/PE/thin Mach-O· [δυαδικά αρχεία Rust](../rust-binaries.md)                                         |
+| Αρχείο WASM                            | Έλεγχο εγκυρότητας, ενότητες, ακριβείς μορφές εισαγωγών/εξαγωγών σε WAT και αποκωδικοποιημένο WAT                           | WABT 1.0.42 που παρέχει ο χρήστης· `inspect_wasm_artifact`                                                                          |
+| Bytecode EVM                           | Selectors επιλογής συνάρτησης, μετατοπίσεις σε bytes, εκτιμώμενα ορίσματα και δυνατότητα μεταβολής της κατάστασης           | Τοπική είσοδος σε ακατέργαστα bytes ή δεκαεξαδική μορφή σε Linux x64· [οδηγός EVM χωρίς εκτέλεση](../evm-bytecode.md)               |
+| Βάσεις δεδομένων SQLite                | Πίνακες, στήλες, ευρετήρια, όψεις, triggers και επιλεγμένες γραμμές κανονικών πινάκων                                       | Τοπικό στιγμιότυπο βάσης δεδομένων/WAL· Node.js 24.x >=24.15 ή 26+ με εγγενή όρια SQLite                                            |
+| Καταγεγραμμένες καταρρεύσεις στο Linux | Ακατέργαστες εγγραφές notes, καταχωρητές/σήματα κάθε καταγεγραμμένου νήματος και προαιρετικές πιθανές αντιστοιχίσεις μνήμης | pwntools που παρέχει ο χρήστης σε Linux x64· προαιρετικά GDB/pwndbg· [καταγεγραμμένες καταρρεύσεις](../recorded-crashes.md)         |
+| JavaScript / Electron                  | Modules, εισαγωγές, source maps, διαδρομές, IPC και σχέσεις με πρόσθετα κώδικα μηχανής                                      | Node.js και npm· [ανάλυση εφαρμογών](https://rea.tools/guides/javascript/)                                                          |
+| Ιστότοποι                              | Δομή σελίδων, scripts, παρατηρήσεις δικτύου και ζητούμενα στιγμιότυπα οθόνης                                                | Πρόγραμμα περιήγησης της οικογένειας Chrome· [ανάλυση μέσω προγράμματος περιήγησης](https://rea.tools/guides/browser/)              |
+| Αποθηκευμένες καταγραφές δικτύου       | Αιτήματα, αποκρίσεις, διαθέσιμα δεδομένα φορτίου και θέσεις στην πηγή                                                       | HAR ή SAZ του Fiddler Everywhere· mitmdump σε Linux για την εγγενή μορφή mitmproxy· [οδηγός καταγραφών](../web-network-captures.md) |
+| Assemblies .NET                        | Μεταδεδομένα, εντολές CIL, δηλωμένες εξαρτήσεις από κώδικα μηχανής και συγκρίσεις builds                                    | Καμία πρόσθετη απαίτηση (μόνο στατική ανάλυση)· [οδηγός διαχειριζόμενου κώδικα](../managed-code-analysis.md)                        |
+| APK Android                            | Δηλώσεις manifest, κλάσεις, απομεταγλωττισμένες μέθοδοι και αναφορές                                                        | JADX χωρίς γραφικό περιβάλλον και πλήρες JDK σε Linux/macOS/Windows x64· [οδηγός Android](../android-analysis.md)                   |
+| Συσκευές Android                       | Συσκευές, πακέτα, διεργασίες, logs, dumps, στιγμιότυπα οθόνης και μεταφορές APK/αρχείων                                     | adb που παρέχει ο χρήστης (εξομοιωτής ή φυσική συσκευή)· [οδηγός ADB](../adb-device-analysis.md)                                    |
+| Έργα JEB                               | Μονάδες έργου από τη μηχανή ανάλυσης, ψηφιακά αποτυπώματα αρχείων και ψευδοκώδικα τύπων/μεθόδων                             | Client JEB που έχει ξεκινήσει ο χρήστης και παρέχει MCP, επιλεγμένος μέσω REA_JEB_MCP_URL· [οδηγός JEB](../jeb-analysis.md)         |
+| Πόροι Android                          | Αποκωδικοποιημένο manifest, πίνακες συμβολοσειρών, τοπικές ρυθμίσεις και στοιχεία έκδοσης                                   | Launcher του Apktool που παρέχει ο χρήστης· [οδηγός πόρων](../apktool-resource-analysis.md)                                         |
+| Εφαρμογές Flutter                      | Hash του στιγμιότυπου Dart, ψηφιακά αποτυπώματα βιβλιοθηκών ανά ABI και στοιχεία προέλευσης του build της μηχανής           | Καμία πρόσθετη απαίτηση (στατική ανάλυση APK)· [οδηγός Flutter](../flutter-analysis.md)                                             |
+| Firmware                               | Περιοχές, αποτελέσματα εξαγωγής και παραπομπές για ανάλυση κώδικα μηχανής                                                   | Binwalk / Unblob σε Linux· [οδηγός firmware](../firmware-analysis.md)                                                               |
+| Καταγεγραμμένα πακέτα                  | Επιλεγμένα πεδία, επαναλαμβανόμενες τιμές, φίλτρα και κανόνες decode-as                                                     | Linux: TShark 4.x + prlimit· [οδηγός καταγραφών](../packet-captures.md)                                                             |
+| Σύνοψη πακέτων                         | Τερματικά σημεία, DNS, TLS SNI, κατανομή πρωτοκόλλων και αιτήματα HTTP                                                      | tshark και capinfos που παρέχει ο χρήστης (επαληθευμένα σε Linux)· [οδηγός pcap](../pcap-analysis.md)                               |
+| Καταγεγραμμένα λογικά σήματα           | Σχολιασμούς αποκωδικοποιητή, εύρη δειγμάτων και διαγνωστικά                                                                 | Linux: sigrok-cli 0.7.2 + prlimit· [καταγραφές σημάτων](../signal-recordings.md)                                                    |
+| Πακέτα και πόροι                       | Καταλόγους αρχείων, ψηφιακά αποτυπώματα, plists, δομή bundles της Apple και εξαγόμενους πόρους                              | [Οδηγός αρχείων και JavaScript](../javascript-artifact-reconstruction.md), [εφαρμογές Apple](../apple-application-analysis.md)      |
+| Συμπεριφορά διεργασιών                 | Έξοδο τερματικού, αλληλεπιδράσεις, παρατηρήσεις τερματισμού και συστήματος αρχείων, καθώς και συγκρίσεις εκτελέσεων         | Linux/macOS με εγγενές PTY· [καταγραφή διεργασιών](../process-capture.md)                                                           |
+
+Για εξέταση δυαδικών αρχείων με κώδικα μηχανής χωρίς εκτέλεση, το REA υποστηρίζει
+επίσης GNU objdump και Rizin που παρέχει ο χρήστης. Δες τον
+[οδηγό εργαλείων αντίστροφης μηχανικής](../reverse-engineering-tools.md)
+για τη ρύθμιση των παρόχων, τις εντολές CLI και τους περιορισμούς των Evidence.
+Το REA δεν εγκαθιστά ούτε ενημερώνει κανέναν από τους δύο παρόχους.
+
+Η στατική εξέταση JavaScript και .NET διαβάζει τα αρχεία που παρέχονται χωρίς να εκτελεί την εφαρμογή. Η καταγραφή κατά την εκτέλεση εκτελεί τον επιλεγμένο στόχο ή αλληλεπιδρά μαζί του με τα δικαιώματα του χρήστη σου. Κάθε σχετικός οδηγός περιγράφει τις επιπτώσεις της.
+
+<a id="choosing-a-deep-analysis-provider"></a>
+
+Οι μορφές κώδικα μηχανής και τα υποστηριζόμενα συστήματα διαφέρουν ανά πάροχο ανάλυσης. Δες τη [ρύθμιση Hopper και Ghidra](../installation.md#hopper), τον [οδηγό IDA](../ida-provider.md) (που περιλαμβάνει προαιρετική γέφυρα MCP για κώδικα μηχανής) και την [πειραματική υποστήριξη Ghidra στα Windows](../windows-ghidra-p0.md). Το Ghidra υποστηρίζει επίσης [ανάλυση DOS 16 bit](../ghidra-dos.md). Για μεγάλα δυαδικά αρχεία, αύξησε το χρονικό όριο εκκίνησης του παρόχου με `REA_GHIDRA_STARTUP_TIMEOUT_MS` ή `REA_HOPPER_STARTUP_TIMEOUT_MS`. Για την επιλογή παρόχου, δες τον [οδηγό CLI](../cli.md#choose-a-provider). Έλεγξε τη [διαθεσιμότητα ανά έκδοση](../installation.md#released-package-and-main) για λειτουργίες που προστέθηκαν μετά την τελευταία έκδοση στο npm.
+
+## Μελέτες περίπτωσης
+
+[![Εικονογραφήσεις των μελετών περίπτωσης για τη στερεοφωνική τοποθέτηση ήχου στο DX-Ball, τη γέφυρα προχείρου του Notion και τον δακτύλιο βλημάτων του TH04](../assets/rea-showcases.png)](https://rea.tools/showcase/)
+
+### DX-Ball: ανακατασκευή υπολογισμού στερεοφωνικής τοποθέτησης ήχου
+
+Ακολούθησε μια κλήση ήχου μέχρι τη βοηθητική συνάρτηση που μετατρέπει τη θέση σε στερεοφωνική τοποθέτηση, εξέτασε τις εντολές και μετέτρεψε τον ελλιπή ψευδοκώδικα σε C. Η ανακατασκευή περνά 3.205 δοκιμές με τον αρχικό κώδικα x86 και αναπαράγει και τα 63 bytes της μεταγλωττισμένης συνάρτησης.
+
+[Διάβασε τη μελέτη περίπτωσης](https://rea.tools/showcase/dx-ball/) · [Αποθετήριο ανακατασκευής](https://github.com/N0zoM1z0/dx-ball)
+
+### Notion: ιχνηλάτηση της γέφυρας προχείρου στο Electron
+
+Βρες το API προχείρου του renderer, ακολούθησέ το μέσω του preload και του IPC μέχρι την κύρια διεργασία και εξέτασε τη μορφή εμπλουτισμένου περιεχομένου του προχείρου.
+
+[Διάβασε τη μελέτη περίπτωσης](https://rea.tools/showcase/notion/)
+
+### TH04: ανάκτηση υπολογισμού δακτυλίου βλημάτων σε DOS
+
+Εξέτασε τις εντολές 16 bit του αρχικού παιχνιδιού για PC-98, ανέκτησε τους υπολογισμούς σταθερών γωνιών και γωνιών στόχευσης και σύγκρινε τον ανακατασκευασμένο κώδικα C++ με την έξοδο του μεταγλωττιστή της εποχής.
+
+[Διάβασε τη μελέτη περίπτωσης](https://rea.tools/showcase/th04/) · [Αποθετήριο ανακατασκευής](https://github.com/N0zoM1z0/th04)
+
+Αν χρησιμοποίησες το REA για κάτι ενδιαφέρον, θα θέλαμε να το δούμε. Μοιράσου τη μελέτη σου σε ένα [issue](https://github.com/morluto/rea/issues) ή [pull request](https://github.com/morluto/rea/pulls), αναφέροντας τον στόχο, το ερώτημά σου, πώς βοήθησε το REA και τι ανακάλυψες.
+
+## Συχνές ερωτήσεις
+
+<details>
+<summary><strong>Ποιοι agents μπορούν να χρησιμοποιήσουν το REA;</strong></summary>
+
+Κάθε agent που υποστηρίζει τοπικούς MCP servers. Η διαδικασία εγκατάστασης ρυθμίζει τους [υποστηριζόμενους agents](../installation.md#supported-agents). Οι υπόλοιποι clients μπορούν να χρησιμοποιήσουν τη [μη αυτόματη καταχώριση MCP](../installation.md#mcp-registry).
+
+</details>
+
+<details>
+<summary><strong>Χρειάζομαι Hopper, Ghidra ή IDA;</strong></summary>
+
+Η εις βάθος ανάλυση κώδικα μηχανής χρησιμοποιεί ένα από αυτά. Η στατική εξέταση JavaScript και .NET λειτουργεί χωρίς μηχανή ανάλυσης κώδικα μηχανής. Η διαδικασία εγκατάστασης μπορεί να εγκαταστήσει το Hopper μετά την έγκρισή σου· το Ghidra και το IDA χρησιμοποιούν τις υπάρχουσες εγκαταστάσεις σου. Δες τη [ρύθμιση παρόχων ανάλυσης](../installation.md#hopper).
+
+</details>
+
+<details>
+<summary><strong>Πρέπει να ξεκινήσω πρώτα το Hopper;</strong></summary>
+
+Το REA ξεκινά το Hopper όταν το χρειάζεται κάποια λειτουργία. Στο macOS, κατά την πρώτη εκκίνηση μπορεί να εμφανιστεί παράθυρο που ζητά να επιλέξεις τη δοκιμαστική λειτουργία ή να ενεργοποιήσεις την άδειά σου. Ολοκλήρωσε αυτό το βήμα στην ίδια συνεδρία χρήστη πριν από την πρώτη λειτουργία CLI/MCP ή εργασία χωρίς επίβλεψη. Άνοιξε το Hopper χειροκίνητα αν χρειαστεί. Δες την [εκκίνηση και αντιμετώπιση προβλημάτων του Hopper](../installation.md#launcher-paths-and-troubleshooting).
+
+</details>
+
+<details>
+<summary><strong>Τι προσφέρει η εγκατάσταση του skill από το skills.sh;</strong></summary>
+
+Το skill παρέχει οδηγίες διερεύνησης για τον agent σου. Χρησιμοποίησε `npx rea-agents setup` για να καταχωρίσεις τον MCP server του REA και να εγκαταστήσεις τις αντίστοιχες οδηγίες. Στη συνέχεια, επανεκκίνησε τον agent σου. Δες την [εγκατάσταση μόνο του skill](../installation.md#skill-only-installation).
+
+</details>
+
+<details>
+<summary><strong>Τι κώδικα επιστρέφει το REA;</strong></summary>
+
+Η ανάλυση κώδικα μηχανής επιστρέφει ψευδοκώδικα και assembly. Η ανάλυση JavaScript/Electron ανακτά modules και τις σχέσεις τους. Ο agent σου χρησιμοποιεί αυτά τα ευρήματα για να γράψει και να δοκιμάσει μια υλοποίηση. Οι [μελέτες περίπτωσης](#μελέτες-περίπτωσης) παρουσιάζουν ολοκληρωμένα παραδείγματα.
+
+</details>
+
+<details>
+<summary><strong>Το REA ανεβάζει την εφαρμογή μου στο διαδίκτυο;</strong></summary>
+
+Το REA αναλύει τους στόχους τοπικά. Ο agent σου λαμβάνει τα αποτελέσματα των εργαλείων και ο πάροχος του μοντέλου του έχει τη δική του πολιτική δεδομένων.
+
+</details>
+
+<details>
+<summary><strong>Τι να κάνω αν συναντήσω κάποιο σφάλμα;</strong></summary>
+
+Κάνε πρώτα ενημέρωση· μια πρόσφατη έκδοση μπορεί να το έχει ήδη διορθώσει.
+
+Για CLI εγκατεστημένο με `npm install --global` ή `bun add --global`:
+
+```bash
+rea update
+```
+
+Για ρύθμιση agent μέσω `npx`:
+
+```bash
+npx rea-agents@latest setup
+```
+
+Αν χρησιμοποιείς agent, ολοκλήρωσε την [ανανέωση της ρύθμισης](#ενημέρωση-του-rea) και επανεκκίνησέ τον. Δοκίμασε ξανά την ίδια εργασία. Αν το πρόβλημα παραμένει, [άνοιξε ένα issue](https://github.com/morluto/rea/issues) με την έκδοση του REA, τον τύπο του στόχου, τα βήματα αναπαραγωγής και την έξοδο του σφάλματος.
+
+</details>
+
+## Τεκμηρίωση
+
+Ξεκίνα από τους [οδηγούς με ολοκληρωμένα παραδείγματα](https://rea.tools/guides/) στον ιστότοπο. Για τις ακριβείς επιλογές, τις προϋποθέσεις και τις προδιαγραφές των αποτελεσμάτων:
+
+- [Εγκατάσταση και ρύθμιση](../installation.md): καταχώριση agents, ρύθμιση παρόχων ανάλυσης, ενημερώσεις και απεγκατάσταση.
+- [Έλεγχος ετοιμότητας και αντιμετώπιση προβλημάτων](../installation.md#check-readiness-for-your-task): διάγνωση ενός agent ή μιας μηχανής ανάλυσης.
+- [CLI και Evidence](../cli.md): εντολές, επιλογή παρόχου, στιγμιότυπα, εισαγωγή/εξαγωγή και κωδικοί εξόδου.
+- [Προδιαγραφές MCP](../mcp-contracts.md) και [prompts για agents](../mcp-prompts.md): αποτελέσματα εργαλείων, συνεδρίες και καθοδηγούμενες διερευνήσεις.
+- [Κατάλογος εργαλείων](../mcp-contracts.md#generated-catalog): αυτόματα παραγόμενος κατάλογος εργαλείων, παρόχων και εντολών CLI.
+- [Οδικός χάρτης](../roadmap.md): προγραμματισμένες εργασίες και παρακολούθηση της ανάπτυξης δυνατοτήτων.
+
+Ανάφερε ευπάθειες ακολουθώντας το [SECURITY.md](../../SECURITY.md).
+
+## Ιστορικό αστεριών
+
+🎉 **180.000 αστέρια στο GitHub — σας ευχαριστούμε!**
+
+Ευχαριστούμε όλους όσοι χρησιμοποιούν το REA, αναφέρουν σφάλματα, ζητούν νέες λειτουργίες, δοκιμάζουν builds και συνεισφέρουν διορθώσεις.
+
+<a href="https://www.star-history.com/?repos=morluto%2Frea&amp;type=date">
+  <picture>
+    <source media="(prefers-color-scheme: dark)" srcset="https://api.star-history.com/chart?repos=morluto/rea&amp;type=date&amp;theme=dark&amp;legend=top-left" />
+    <source media="(prefers-color-scheme: light)" srcset="https://api.star-history.com/chart?repos=morluto/rea&amp;type=date" />
+    <img alt="Ιστορικό αστεριών του REA στο GitHub" src="https://api.star-history.com/chart?repos=morluto/rea&amp;type=date" />
+  </picture>
+</a>
+
+## Αποποίηση ευθύνης
+
+Δες το [DISCLAIMER.md](../../DISCLAIMER.md) για οδηγίες υπεύθυνης χρήσης.
+
+Η θέση του έργου σχετικά με την παράνομη ή μη εξουσιοδοτημένη χρήση ισχύει για όλες τις εκδόσεις του REA, τα κανάλια διανομής και τις πλατφόρμες. Κάθε αντίγραφο διατηρεί τις άδειες που το διέπουν.
+
+Το REA προορίζεται για νόμιμη έρευνα, εκπαίδευση, διαλειτουργικότητα, διατήρηση λογισμικού και ανάλυση λογισμικού που σου ανήκει ή για το οποίο έχεις άδεια εξέτασης.
+
+## Συνεισφορά
+
+Θα χαρούμε να βοηθήσεις στο REA! [Άνοιξε ένα issue](https://github.com/morluto/rea/issues) για να αναφέρεις ένα σφάλμα ή να προτείνεις μια λειτουργία, ή [στείλε ένα pull request](https://github.com/morluto/rea/pulls) για να βελτιώσεις τον κώδικα ή την τεκμηρίωση.
+
+Δες το [CONTRIBUTING.md](../../CONTRIBUTING.md) για τη ρύθμιση του περιβάλλοντος ανάπτυξης και τους ελέγχους, τον [οδηγό δοκιμών](../testing.md) για τις διαδικασίες επαλήθευσης και τον [χάρτη αρχιτεκτονικής](../architecture.mermaid) για τη δομή του έργου.
+
+## Άδεια χρήσης
+
+[MIT](../../LICENSE)
+
+Ο κώδικας που έχει δημιουργηθεί για το REA διατίθεται με την άδεια MIT. Το υλικό τρίτων διατηρεί τις δικές του άδειες και δηλώσεις πνευματικών δικαιωμάτων. Δες τις [δηλώσεις για υλικό τρίτων](../../THIRD_PARTY_NOTICES.md) για τον κατάλογο εξαρτήσεων, τα διατηρημένα κείμενα αδειών και τους όρους των εξωτερικών εργαλείων.
+
+**Πρέπει να αποκτήσεις νόμιμα τα εμπορικά εργαλεία και να έχεις δικαίωμα χρήσης τους**, συμπεριλαμβανομένων των IDA Pro / Hex-Rays decompilers, Hopper και JEB, σύμφωνα με τις αντίστοιχες άδειές τους. Το REA δεν παρέχει άδεια εργαλείου, κλειδί ή δικαίωμα ενεργοποίησης. Οι δοκιμαστικές εκδόσεις και τα demos εξακολουθούν να υπόκεινται στους όρους του προμηθευτή, συμπεριλαμβανομένων των περιορισμών χρήστη, λειτουργιών, αυτοματοποίησης και κοινοποίησης αποτελεσμάτων. Δεν επιτρέπονται σπασμένες εκδόσεις, κλεμμένα κλειδιά, μη εξουσιοδοτημένη κοινοποίηση κλειδιών ή παράκαμψη περιορισμών αδειοδότησης. Αυτές οι οδηγίες δεν τροποποιούν την άδεια MIT ή τις άδειες τρίτων.
@@ -1,6 +1,7 @@
 <div align="center">

 [English](../../README.md) · [简体中文](README_zh.md) · [繁體中文](README_zh-TW.md) · [日本語](README_ja.md) · [한국어](README_ko.md) · [Türkçe](README_tr.md) · [Русский](README_ru.md) · [Tiếng Việt](README_vi.md) · [ไทย](README_th.md) · [Deutsch](README_de.md) · **Español** · [Français](README_fr.md) · [Українська](README_uk.md) · [Polski](README_pl.md) · [Português (Brasil)](README_pt-BR.md) · [العربية](README_ar.md) · [فارسی](README_fa.md) · [Bahasa Indonesia](README_id.md) · [Italiano](README_it.md)
+· [Ελληνικά](README_el.md)

 # REA: Ingeniería inversa de cualquier cosa

@@ -19,10 +20,14 @@

 **[Sitio web](https://rea.tools/es/) · [Guías](https://rea.tools/es/guides/) · [Casos prácticos](https://rea.tools/es/showcase/)**

-[Inicio rápido](#inicio-rápido) · [Cómo funciona REA](#cómo-funciona-rea) · [Qué puedes analizar](#qué-puedes-analizar) · [Casos prácticos](#casos-prácticos) · [Preguntas frecuentes](#preguntas-frecuentes) · [Documentación](#documentación)
+[Inicio rápido](#inicio-rápido) · [Cómo funciona REA](#cómo-funciona-rea) · [Qué puedes analizar](#qué-puedes-analizar) · [Casos prácticos](#casos-prácticos) · [Preguntas frecuentes](#preguntas-frecuentes) · [Herramientas GNU objdump y Rizin](../reverse-engineering-tools.md) · [Documentación](#documentación)

 <code>npx rea-agents setup</code>

+> [!IMPORTANT]
+> Fuentes oficiales de REA: [morluto/rea](https://github.com/morluto/rea), [rea.tools](https://rea.tools/), [rea-agents](https://www.npmjs.com/package/rea-agents) (npm).
+> Usa los enlaces de instalación y de versiones publicados en estas fuentes oficiales. Un nombre de repositorio parecido o un README copiado no demuestra que una descarga sea oficial.
+
 <br />

 <img src="../assets/rea-hopper-analysis.png" alt="REA inicia su puente de análisis dentro de Hopper mientras inspecciona un binario nativo" width="1200" />
@@ -1,6 +1,7 @@
 <div align="center" dir="rtl">

 [English](../../README.md) · [简体中文](README_zh.md) · [繁體中文](README_zh-TW.md) · [日本語](README_ja.md) · [한국어](README_ko.md) · [Türkçe](README_tr.md) · [Русский](README_ru.md) · [Tiếng Việt](README_vi.md) · [ไทย](README_th.md) · [Deutsch](README_de.md) · [Español](README_es.md) · [Français](README_fr.md) · [Українська](README_uk.md) · [Polski](README_pl.md) · [Português (Brasil)](README_pt-BR.md) · [العربية](README_ar.md) · **فارسی** · [Bahasa Indonesia](README_id.md) · [Italiano](README_it.md)
+· [Ελληνικά](README_el.md)

 # REA: مهندسی معکوس هر چیزی

@@ -19,10 +20,14 @@

 **[وب‌سایت](https://rea.tools/fa/) · [راهنماها](https://rea.tools/fa/guides/) · [نمونه‌های عملی](https://rea.tools/fa/showcase/)**

-[شروع سریع](#شروع-سریع) · [REA چگونه کار می‌کند؟](#rea-چگونه-کار-می‌کند) · [چه چیزهایی را میتوان تحلیل کرد؟](#چه-چیزهایی-را-می‌توان-تحلیل-کرد) · [نمونه‌های عملی](#نمونه‌های-عملی) · [پرسش‌های متداول](#پرسش‌های-متداول) · [مستندات](#مستندات)
+[شروع سریع](#شروع-سریع) · [REA چگونه کار می‌کند؟](#rea-چگونه-کار-می‌کند) · [چه چیزهایی را میتوان تحلیل کرد؟](#چه-چیزهایی-را-می‌توان-تحلیل-کرد) · [نمونه‌های عملی](#نمونه‌های-عملی) · [پرسش‌های متداول](#پرسش‌های-متداول) · [ابزارهای GNU objdump و Rizin](../reverse-engineering-tools.md) · [مستندات](#مستندات)

 <code>npx rea-agents setup</code>

+> [!IMPORTANT]
+> منابع رسمی REA: [morluto/rea](https://github.com/morluto/rea), [rea.tools](https://rea.tools/), [rea-agents](https://www.npmjs.com/package/rea-agents) (npm).
+> از پیوندهای نصب و انتشار در این منابع رسمی استفاده کنید. نام مشابه مخزن یا README کپی‌شده ثابت نمی‌کند که فایل دانلودی رسمی است.
+
 <br />

 <img src="../assets/rea-hopper-analysis.png" alt="اجرای پل تحلیل REA در Hopper هنگام بررسی یک فایل باینری بومی" width="1200" />
@@ -1,6 +1,7 @@
 <div align="center">

 [English](../../README.md) · [简体中文](README_zh.md) · [繁體中文](README_zh-TW.md) · [日本語](README_ja.md) · [한국어](README_ko.md) · [Türkçe](README_tr.md) · [Русский](README_ru.md) · [Tiếng Việt](README_vi.md) · [ไทย](README_th.md) · [Deutsch](README_de.md) · [Español](README_es.md) · **Français** · [Українська](README_uk.md) · [Polski](README_pl.md) · [Português (Brasil)](README_pt-BR.md) · [العربية](README_ar.md) · [فارسی](README_fa.md) · [Bahasa Indonesia](README_id.md) · [Italiano](README_it.md)
+· [Ελληνικά](README_el.md)

 # REA : rétro-ingénierie de tout

@@ -19,10 +20,14 @@

 **[Site web](https://rea.tools/fr/) · [Guides](https://rea.tools/fr/guides/) · [Études de cas](https://rea.tools/fr/showcase/)**

-[Démarrage rapide](#démarrage-rapide) · [Comment fonctionne REA](#comment-fonctionne-rea) · [Ce que vous pouvez analyser](#ce-que-vous-pouvez-analyser) · [Études de cas](#études-de-cas) · [Questions fréquentes](#questions-fréquentes) · [Documentation](#documentation)
+[Démarrage rapide](#démarrage-rapide) · [Comment fonctionne REA](#comment-fonctionne-rea) · [Ce que vous pouvez analyser](#ce-que-vous-pouvez-analyser) · [Études de cas](#études-de-cas) · [Questions fréquentes](#questions-fréquentes) · [Outils GNU objdump et Rizin](../reverse-engineering-tools.md) · [Documentation](#documentation)

 <code>npx rea-agents setup</code>

+> [!IMPORTANT]
+> Sources officielles de REA : [morluto/rea](https://github.com/morluto/rea), [rea.tools](https://rea.tools/), [rea-agents](https://www.npmjs.com/package/rea-agents) (npm).
+> Utilisez les liens d’installation et de publication de ces sources officielles. Un nom de dépôt similaire ou un README copié ne prouve pas qu’un téléchargement est officiel.
+
 <br />

 <img src="../assets/rea-hopper-analysis.png" alt="REA lance son pont d'analyse dans Hopper pendant l'inspection d'un binaire natif" width="1200" />
@@ -1,6 +1,7 @@
 <div align="center">

 [English](../../README.md) · [简体中文](README_zh.md) · [繁體中文](README_zh-TW.md) · [日本語](README_ja.md) · [한국어](README_ko.md) · [Türkçe](README_tr.md) · [Русский](README_ru.md) · [Tiếng Việt](README_vi.md) · [ไทย](README_th.md) · [Deutsch](README_de.md) · [Español](README_es.md) · [Français](README_fr.md) · [Українська](README_uk.md) · [Polski](README_pl.md) · [Português (Brasil)](README_pt-BR.md) · [العربية](README_ar.md) · [فارسی](README_fa.md) · **Bahasa Indonesia** · [Italiano](README_it.md)
+· [Ελληνικά](README_el.md)

 # REA: Rekayasa Balik Apa Pun

@@ -19,10 +20,14 @@

 **[Situs Web](https://rea.tools/id/) · [Panduan](https://rea.tools/id/guides/) · [Contoh Penggunaan](https://rea.tools/id/showcase/)**

-[Mulai Cepat](#mulai-cepat) · [Cara Kerja REA](#cara-kerja-rea) · [Hal yang Dapat Anda Analisis](#hal-yang-dapat-anda-analisis) · [Contoh Penggunaan](#contoh-penggunaan) · [FAQ](#faq) · [Dokumentasi](#dokumentasi)
+[Mulai Cepat](#mulai-cepat) · [Cara Kerja REA](#cara-kerja-rea) · [Hal yang Dapat Anda Analisis](#hal-yang-dapat-anda-analisis) · [Contoh Penggunaan](#contoh-penggunaan) · [FAQ](#faq) · [Alat GNU objdump dan Rizin](../reverse-engineering-tools.md) · [Dokumentasi](#dokumentasi)

 <code>npx rea-agents setup</code>

+> [!IMPORTANT]
+> Sumber resmi REA: [morluto/rea](https://github.com/morluto/rea), [rea.tools](https://rea.tools/), [rea-agents](https://www.npmjs.com/package/rea-agents) (npm).
+> Gunakan tautan instalasi dan rilis dari sumber resmi ini. Nama repositori yang mirip atau README yang disalin tidak membuktikan bahwa unduhan tersebut resmi.
+
 <br />

 <img src="../assets/rea-hopper-analysis.png" alt="REA menjalankan jembatan analisisnya di dalam Hopper saat memeriksa biner native" width="1200" />
@@ -1,6 +1,7 @@
 <div align="center">

 [English](../../README.md) · [简体中文](README_zh.md) · [繁體中文](README_zh-TW.md) · [日本語](README_ja.md) · [한국어](README_ko.md) · [Türkçe](README_tr.md) · [Русский](README_ru.md) · [Tiếng Việt](README_vi.md) · [ไทย](README_th.md) · [Deutsch](README_de.md) · [Español](README_es.md) · [Français](README_fr.md) · [Українська](README_uk.md) · [Polski](README_pl.md) · [Português (Brasil)](README_pt-BR.md) · [العربية](README_ar.md) · [فارسی](README_fa.md) · [Bahasa Indonesia](README_id.md) · **Italiano**
+· [Ελληνικά](README_el.md)

 # REA: Reverse Engineer Anything

@@ -19,10 +20,14 @@

 **[Sito web](https://rea.tools/it/) · [Guide](https://rea.tools/it/guides/) · [Showcase](https://rea.tools/it/showcase/)**

-[Avvio rapido](#avvio-rapido) · [Come funziona REA](#come-funziona-rea) · [Cosa puoi analizzare](#cosa-puoi-analizzare) · [Showcase](#showcase) · [FAQ](#faq) · [Documentazione](#documentazione)
+[Avvio rapido](#avvio-rapido) · [Come funziona REA](#come-funziona-rea) · [Cosa puoi analizzare](#cosa-puoi-analizzare) · [Showcase](#showcase) · [FAQ](#faq) · [Strumenti GNU objdump e Rizin](../reverse-engineering-tools.md) · [Documentazione](#documentazione)

 <code>npx rea-agents setup</code>

+> [!IMPORTANT]
+> Fonti ufficiali di REA: [morluto/rea](https://github.com/morluto/rea), [rea.tools](https://rea.tools/), [rea-agents](https://www.npmjs.com/package/rea-agents) (npm).
+> Segui i link di installazione e delle versioni pubblicate da queste fonti ufficiali. Un nome di repository simile o un README copiato non dimostra che un download sia ufficiale.
+
 <br />

 <img src="../assets/rea-hopper-analysis.png" alt="REA che avvia il suo bridge di analisi dentro Hopper mentre ispeziona un binario nativo" width="1200" />
@@ -1,6 +1,7 @@
 <div align="center">

 [English](../../README.md) · [简体中文](README_zh.md) · [繁體中文](README_zh-TW.md) · **日本語** · [한국어](README_ko.md) · [Türkçe](README_tr.md) · [Русский](README_ru.md) · [Tiếng Việt](README_vi.md) · [ไทย](README_th.md) · [Deutsch](README_de.md) · [Español](README_es.md) · [Français](README_fr.md) · [Українська](README_uk.md) · [Polski](README_pl.md) · [Português (Brasil)](README_pt-BR.md) · [العربية](README_ar.md) · [فارسی](README_fa.md) · [Bahasa Indonesia](README_id.md) · [Italiano](README_it.md)
+· [Ελληνικά](README_el.md)

 # REA：あらゆるものをリバースエンジニアリング

@@ -19,10 +20,14 @@

 **[ウェブサイト](https://rea.tools/ja/) · [ガイド](https://rea.tools/ja/guides/) · [事例](https://rea.tools/ja/showcase/)**

-[クイックスタート](#クイックスタート) · [REA の仕組み](#rea-の仕組み) · [分析できる対象](#分析できる対象) · [事例](#事例) · [よくある質問](#よくある質問) · [ドキュメント](#ドキュメント)
+[クイックスタート](#クイックスタート) · [REA の仕組み](#rea-の仕組み) · [分析できる対象](#分析できる対象) · [事例](#事例) · [よくある質問](#よくある質問) · [GNU objdump・Rizin ツール](../reverse-engineering-tools.md) · [ドキュメント](#ドキュメント)

 <code>npx rea-agents setup</code>

+> [!IMPORTANT]
+> REA の公式配布元： [morluto/rea](https://github.com/morluto/rea), [rea.tools](https://rea.tools/), [rea-agents](https://www.npmjs.com/package/rea-agents) (npm).
+> インストールとリリースのリンクは、これらの公式配布元からたどってください。似たリポジトリ名やコピーされた README だけでは、公式の配布物とは確認できません。
+
 <br />

 <img src="../assets/rea-hopper-analysis.png" alt="REA が Hopper 内で分析ブリッジを起動し、ネイティブバイナリを調べる様子" width="1200" />
@@ -1,6 +1,7 @@
 <div align="center">

 [English](../../README.md) · [简体中文](README_zh.md) · [繁體中文](README_zh-TW.md) · [日本語](README_ja.md) · **한국어** · [Türkçe](README_tr.md) · [Русский](README_ru.md) · [Tiếng Việt](README_vi.md) · [ไทย](README_th.md) · [Deutsch](README_de.md) · [Español](README_es.md) · [Français](README_fr.md) · [Українська](README_uk.md) · [Polski](README_pl.md) · [Português (Brasil)](README_pt-BR.md) · [العربية](README_ar.md) · [فارسی](README_fa.md) · [Bahasa Indonesia](README_id.md) · [Italiano](README_it.md)
+· [Ελληνικά](README_el.md)

 # REA: 무엇이든 리버스 엔지니어링

@@ -23,6 +24,10 @@

 <code>npx rea-agents setup</code>

+> [!IMPORTANT]
+> REA 공식 배포처: [morluto/rea](https://github.com/morluto/rea), [rea.tools](https://rea.tools/), [rea-agents](https://www.npmjs.com/package/rea-agents) (npm).
+> 이 공식 배포처에서 제공하는 설치 및 릴리스 링크를 이용하세요. 비슷한 저장소 이름이나 복사된 README만으로는 공식 배포 파일인지 확인할 수 없습니다.
+
 <br />

 <img src="../assets/rea-hopper-analysis.png" alt="REA가 Hopper 안에서 분석 브리지를 실행해 네이티브 바이너리를 검사하는 모습" width="1200" />
@@ -51,7 +56,7 @@

 REA는 에이전트를 분석 도구에 연결해 네이티브 바이너리, JavaScript와 Electron 앱, .NET 어셈블리, 웹사이트를 검사할 수 있게 합니다. 터미널에서도 같은 도구를 사용할 수 있습니다. 분석은 로컬에서 실행되며, 결과에는 각 결론의 근거와 한계가 포함됩니다.

-설정 과정은 에이전트에 REA를 등록하고 해당 버전에 맞는 워크플로 지침을 설치합니다. 네이티브 분석에는 기존 Hopper 또는 Ghidra 설치를 사용할 수 있습니다. 승인하면 설정 과정에서 Hopper를 설치할 수도 있습니다. 정적 JavaScript 분석에는 두 엔진 모두 필요하지 않습니다.
+설정 과정은 에이전트에 REA를 등록하고 해당 버전에 맞는 워크플로 지침을 설치합니다. 네이티브 분석에는 기존 Hopper, Ghidra 또는 IDA 설치를 사용할 수 있습니다. 승인하면 설정 과정에서 Hopper를 설치할 수도 있습니다. 정적 JavaScript 분석에는 네이티브 분석 엔진이 필요하지 않습니다.

 > **[REA 웹사이트](https://rea.tools/ko/)**에서 설정 방법, 그림으로 설명한 가이드, 실제 사례를 확인하세요.

@@ -100,12 +105,14 @@ rea --help

 REA는 빠르게 바뀌며 새 릴리스에는 버그 수정이 자주 포함됩니다. 설치된 버전을 최신으로 유지하세요.

-npm으로 설치한 CLI에서는 다음 명령을 사용합니다.
+`npm install --global` 또는 `bun add --global`로 설치한 CLI에서는 다음 명령을 사용합니다.

 ```bash
 rea update
 ```

+Bun 전역 업데이트에는 Bun 1.2.17 이상이 필요합니다. 이전 버전의 Bun을 사용한다면 `rea update`를 실행하기 전에 Bun을 직접 업그레이드하세요.
+
 에이전트 등록과 스킬을 갱신하려면 업데이트 과정에서 출력된 설정 명령을 실행하세요.

 `npx`를 사용한다면 다음 명령으로 에이전트 설정을 갱신하세요.
@@ -114,7 +121,7 @@ rea update
 npx rea-agents@latest setup
 ```

-설정 변경 사항을 검토하고 에이전트를 다시 시작하세요. 일회성 CLI 작업에는 `npx rea-agents@latest` 뒤에 필요한 명령을 붙여 실행하세요.
+설정 변경 사항을 검토하고 에이전트를 다시 시작하세요. 일회성 CLI 작업에는 `npx rea-agents@latest` 뒤에 필요한 명령을 붙여 실행하세요. Bun을 사용한다면 위 명령의 `npx rea-agents@latest`를 `bunx rea-agents@latest`로 바꿔 실행하세요.

 ## REA의 작동 방식

@@ -134,14 +141,16 @@ REA에는 Node.js 22.x(>=22.19), 24.x(>=24.11) 또는 26+와 npm이 필요합니
 | --------------------- | ----------------------------------------------------------------------------- | --------------------------------------------------------------------------------------------------------------------------------- |
 | 네이티브 바이너리     | 의사 코드, 어셈블리, 문자열, 심볼, 호출, 참조                                 | Hopper, Ghidra 또는 IDA; [네이티브 분석](https://rea.tools/ko/guides/native/)                                                     |
 | 오프라인 ELF 레이아웃 | 섹션, 세그먼트, 원본 심볼/재배치 정보, 정적 분석으로 얻은 보안 완화 기법 후보 | Linux x64에서 호출자가 제공하는 pwntools; [바이너리 진단](../binary-diagnostics.md)                                               |
+| Go 바이너리           | 내장된 컴파일러 버전, 모듈, `replace` 지시어, 빌드 설정                       | REA 내장 ELF/PE/thin Mach-O 정적 검사; Go 설치 불필요                                                                             |
+| Rust 바이너리         | cargo-auditable이 내장한 크레이트 이름, 버전, 소스, 의존성 관계               | REA 내장 ELF/PE/thin Mach-O 정적 검사; [Rust 바이너리](../rust-binaries.md)                                                       |
+| WASM 아티팩트         | 유효성 검사, 섹션, import/export의 정확한 WAT 형식, 디코딩된 WAT              | 호출자 제공 WABT 1.0.42; `inspect_wasm_artifact`                                                                                  |
 | EVM 바이트코드        | 디스패치 선택자, 바이트 오프셋, 추론한 인자와 상태 변경 가능 여부             | Linux x64의 로컬 원시 바이트/16진수 입력 파일; [오프라인 EVM 가이드](../evm-bytecode.md)                                          |
-| Rust binaries         | Embedded cargo-auditable dependency list                                      | [Rust binaries](../rust-binaries.md)                                                                                              |
+| SQLite 데이터베이스   | 테이블, 열, 인덱스, 뷰, 트리거, 선택한 일반 테이블의 행                       | 로컬 데이터베이스/WAL 스냅샷; Node.js 24.15 이상의 24.x 또는 26 이상(네이티브 SQLite 한도 적용)                                   |
 | 기록된 Linux 크래시   | 원시 note 레코드, 기록된 모든 스레드의 레지스터/시그널, 선택적 매핑 후보      | Linux x64에서 호출자가 제공하는 pwntools; GDB/pwndbg는 선택 사항; [기록된 크래시](../recorded-crashes.md)                         |
 | JavaScript / Electron | 모듈, 임포트, 소스 맵, 라우트, IPC, 네이티브 애드온 관계                      | Node.js와 npm; [애플리케이션 분석](https://rea.tools/ko/guides/javascript/)                                                       |
 | 웹사이트              | 페이지 구조, 스크립트, 네트워크 관찰 결과, 요청한 스크린샷                    | Chrome 계열 브라우저; [브라우저 분석](https://rea.tools/ko/guides/browser/)                                                       |
-| 저장된 네트워크 캡처  | 요청, 응답, 접근 가능한 페이로드, 소스 위치                                   | HAR; mitmproxy 네이티브 캡처에는 Linux의 mitmdump; [캡처 가이드](../web-network-captures.md)                                      |
-| 패킷 캡처             | 엔드포인트, DNS, TLS SNI, 프로토콜 분포, HTTP 요청                            | 호출자 제공 tshark 및 capinfos; [pcap 가이드](../pcap-analysis.md)                                                                |
-| .NET 어셈블리         | 메타데이터, CIL 명령, 선언된 네이티브 의존성, 빌드 비교                       | 정적 검사; [관리 코드 가이드](../managed-code-analysis.md)                                                                        |
+| 저장된 네트워크 캡처  | 요청, 응답, 접근 가능한 페이로드, 소스 위치                                   | HAR 또는 Fiddler Everywhere SAZ; mitmproxy 네이티브 캡처에는 Linux의 mitmdump; [캡처 가이드](../web-network-captures.md)          |
+| .NET 어셈블리         | 메타데이터, CIL 명령, 선언된 네이티브 의존성, 빌드 비교                       | 별도 도구 불필요(정적 검사만); [관리 코드 가이드](../managed-code-analysis.md)                                                    |
 | Android APK           | 매니페스트 선언, 클래스, 디컴파일된 메서드, 참조                              | Linux/macOS/Windows x64의 헤드리스 JADX와 전체 JDK; [Android 가이드](../android-analysis.md)                                      |
 | Android 기기          | 기기, 패키지, 프로세스, 로그, 덤프, 화면 캡처, APK/파일 전송                  | 호출자 제공 adb(에뮬레이터 또는 실제 기기); [ADB 가이드](../adb-device-analysis.md)                                               |
 | JEB 프로젝트          | 엔진 기반 프로젝트 유닛, 아티팩트 해시, 타입/메서드 의사코드                  | 호출자가 시작하고 MCP를 제공하는 JEB 클라이언트(REA_JEB_MCP_URL로 선택); [JEB 가이드](../jeb-analysis.md)                         |
@@ -149,15 +158,18 @@ REA에는 Node.js 22.x(>=22.19), 24.x(>=24.11) 또는 26+와 npm이 필요합니
 | Flutter 앱            | Dart 스냅샷 해시, ABI별 라이브러리 해시, 엔진 빌드 출처                       | 없음 (정적 APK 파싱); [Flutter 가이드](../flutter-analysis.md)                                                                    |
 | 펌웨어                | 영역, 추출 결과, 네이티브 분석으로 전달할 정보                                | Linux의 Binwalk / Unblob; [펌웨어 가이드](../firmware-analysis.md)                                                                |
 | 기록된 패킷           | 선택 필드, 반복 값, 필터 및 디코딩 규칙                                       | Linux: TShark 4.x + prlimit; [캡처 가이드](../packet-captures.md)                                                                 |
+| 패킷 요약             | 엔드포인트, DNS, TLS SNI, 프로토콜 분포, HTTP 요청                            | 호출자 제공 tshark 및 capinfos(Linux에서 검증됨); [pcap 가이드](../pcap-analysis.md)                                              |
 | 기록된 논리 신호      | 디코더 주석, 샘플 범위, 진단 정보                                             | Linux: sigrok-cli 0.7.2 + prlimit; [신호 기록 가이드](../signal-recordings.md)                                                    |
 | 패키지 및 리소스      | 파일 목록, 다이제스트, plist, Apple 번들 구조, 추출한 리소스                  | [아티팩트 및 JavaScript 가이드](../javascript-artifact-reconstruction.md), [Apple 애플리케이션](../apple-application-analysis.md) |
 | 프로세스 동작         | 터미널 출력, 상호작용, 종료 및 파일 시스템 관찰 결과, 실행 비교               | 네이티브 PTY를 지원하는 Linux/macOS; [프로세스 캡처](../process-capture.md)                                                       |

 정적 JavaScript 및 .NET 검사는 제공된 파일을 읽으며 애플리케이션을 실행하지 않습니다. 런타임 캡처는 사용자의 권한으로 선택한 대상을 실행하거나 대상과 상호작용합니다. 각 런타임 가이드에서 그 영향을 설명합니다.

+오프라인 네이티브 바이너리 검사에는 호출자가 제공하는 GNU objdump와 Rizin도 사용할 수 있습니다. 제공자 설정, CLI 명령, Evidence의 한계는 [리버스 엔지니어링 도구 가이드](../reverse-engineering-tools.md)를 참고하세요. REA는 objdump와 Rizin을 설치하거나 업데이트하지 않습니다.
+
 <a id="choosing-a-deep-analysis-provider"></a>

-지원하는 네이티브 형식과 호스트는 제공자마다 다릅니다. [Hopper 및 Ghidra 설정](../installation.md#hopper), [IDA 가이드](../ida-provider.md), [실험적 Windows Ghidra 지원](../windows-ghidra-p0.md)을 참고하세요. Ghidra는 [16비트 DOS 분석](../ghidra-dos.md)도 지원합니다. 제공자 선택은 [CLI 가이드](../cli.md#choose-a-provider)를 참고하세요. 최신 npm 릴리스 이후 추가된 기능은 [릴리스에서의 사용 가능 여부](../installation.md#released-package-and-main)를 확인하세요.
+지원하는 네이티브 형식과 호스트는 제공자마다 다릅니다. [Hopper 및 Ghidra 설정](../installation.md#hopper), [IDA 가이드](../ida-provider.md)(선택 사항인 네이티브 MCP 브리지 포함), [실험적 Windows Ghidra 지원](../windows-ghidra-p0.md)을 참고하세요. Ghidra는 [16비트 DOS 분석](../ghidra-dos.md)도 지원합니다. 큰 바이너리를 분석할 때는 `REA_GHIDRA_STARTUP_TIMEOUT_MS` 또는 `REA_HOPPER_STARTUP_TIMEOUT_MS`로 제공자의 시작 제한 시간을 늘리세요. 제공자 선택은 [CLI 가이드](../cli.md#choose-a-provider)를 참고하세요. 최신 npm 릴리스 이후 추가된 기능은 [릴리스에서의 사용 가능 여부](../installation.md#released-package-and-main)를 확인하세요.

 ## 사례

@@ -204,7 +216,7 @@ REA로 흥미로운 대상을 분석했다면 알려 주세요. [issue](https://
 <details>
 <summary><strong>Hopper를 먼저 실행해야 하나요?</strong></summary>

-REA는 작업에 필요할 때 Hopper를 실행합니다. macOS에서는 처음 실행할 때 데모 모드를 선택하거나 라이선스를 활성화하라는 대화 상자가 나타날 수 있습니다. [Hopper 실행 및 문제 해결](../installation.md#launcher-paths-and-troubleshooting)을 참고하세요.
+REA는 작업에 필요할 때 Hopper를 실행합니다. macOS에서는 처음 실행할 때 데모 모드를 선택하거나 라이선스를 활성화하라는 대화 상자가 나타날 수 있습니다. 첫 CLI/MCP 작업이나 무인 작업을 실행하기 전에 같은 사용자 세션에서 이 대화 상자를 완료하세요. 필요하면 Hopper를 직접 여세요. [Hopper 실행 및 문제 해결](../installation.md#launcher-paths-and-troubleshooting)을 참고하세요.

 </details>

@@ -234,7 +246,7 @@ REA는 대상을 로컬에서 분석합니다. 에이전트는 도구 결과를

 먼저 업데이트하세요. 최근 릴리스에서 이미 수정되었을 수 있습니다.

-npm으로 설치한 CLI에서는 다음 명령을 사용합니다.
+`npm install --global` 또는 `bun add --global`로 설치한 CLI에서는 다음 명령을 사용합니다.

 ```bash
 rea update
@@ -281,11 +293,9 @@ REA를 사용하고, 버그를 보고하고, 기능을 요청하고, 빌드를

 책임 있는 사용에 관한 안내는 [DISCLAIMER.md](../../DISCLAIMER.md)를 참고하세요.

-REA는 법을 준수하는 연구, 교육, 상호운용성 확보, 소프트웨어 보존, 그리고 본인이 소유하거나 조사 권한을 가진 소프트웨어의 분석을 위한 도구입니다.
+불법적이거나 허가되지 않은 사용에 대한 프로젝트의 입장은 모든 REA 버전, 배포 채널, 플랫폼에 적용되며, 각 사본에는 해당 라이선스가 그대로 유지됩니다.

-The project's position on illegal or unauthorized use covers all REA
-versions, distribution channels, and platforms; each copy retains its
-applicable licenses.
+REA는 법을 준수하는 연구, 교육, 상호운용성 확보, 소프트웨어 보존, 그리고 본인이 소유하거나 조사 권한을 가진 소프트웨어의 분석을 위한 도구입니다.

 ## 기여

@@ -297,14 +307,6 @@ REA에 도움을 보태 주세요! [issue를 등록](https://github.com/morluto/

 [MIT](../../LICENSE)

-REA-authored code is licensed under MIT. Third-party material retains its own
-licenses and copyright notices; see [Third-party notices](../../THIRD_PARTY_NOTICES.md)
-for the dependency inventory, preserved license texts, and external-tool terms.
+REA가 작성한 코드는 MIT 라이선스를 따릅니다. 제3자 자료에는 각자의 라이선스와 저작권 고지가 그대로 유지됩니다. 의존성 목록, 보존된 라이선스 원문, 외부 도구 약관은 [제3자 고지](../../THIRD_PARTY_NOTICES.md)를 참고하세요.

-**You must legally obtain and be authorized to use commercial tools**, including
-IDA Pro / Hex-Rays decompilers, Hopper, and JEB, under their applicable licenses.
-REA grants no tool license, key, or activation entitlement. Trials and demos
-remain subject to vendor terms, including user, feature, automation, and output
-sharing restrictions. Cracked copies, stolen keys, unauthorized key sharing,
-and bypassing license restrictions are not authorized. This guidance does not
-amend MIT or third-party licenses.
+IDA Pro / Hex-Rays 디컴파일러, Hopper, JEB 등 상용 도구는 해당 라이선스에 따라 **합법적으로 취득하고 사용 권한을 갖춰야 합니다.** REA는 어떤 도구의 라이선스, 키, 활성화 권한도 부여하지 않습니다. 체험판과 데모는 사용자, 기능, 자동화, 출력 공유 제한을 포함한 공급업체 약관을 따릅니다. 크랙 버전, 도난 키, 무단 키 공유, 라이선스 제한 우회는 허용되지 않습니다. 이 안내는 MIT 라이선스나 제3자 라이선스를 변경하지 않습니다.
@@ -1,6 +1,7 @@
 <div align="center">

 [English](../../README.md) · [简体中文](README_zh.md) · [繁體中文](README_zh-TW.md) · [日本語](README_ja.md) · [한국어](README_ko.md) · [Türkçe](README_tr.md) · [Русский](README_ru.md) · [Tiếng Việt](README_vi.md) · [ไทย](README_th.md) · [Deutsch](README_de.md) · [Español](README_es.md) · [Français](README_fr.md) · [Українська](README_uk.md) · **Polski** · [Português (Brasil)](README_pt-BR.md) · [العربية](README_ar.md) · [فارسی](README_fa.md) · [Bahasa Indonesia](README_id.md) · [Italiano](README_it.md)
+· [Ελληνικά](README_el.md)

 # REA: Reverse Engineer Anything

@@ -19,10 +20,14 @@

 **[Strona internetowa](https://rea.tools/pl/) · [Przewodniki](https://rea.tools/pl/guides/) · [Przykłady zastosowań](https://rea.tools/pl/showcase/)**

-[Szybki start](#szybki-start) · [Jak działa REA](#jak-działa-rea) · [Co możesz analizować](#co-możesz-analizować) · [Przykłady zastosowań](#przykłady-zastosowań) · [FAQ](#faq) · [Dokumentacja](#dokumentacja)
+[Szybki start](#szybki-start) · [Jak działa REA](#jak-działa-rea) · [Co możesz analizować](#co-możesz-analizować) · [Przykłady zastosowań](#przykłady-zastosowań) · [FAQ](#faq) · [Narzędzia GNU objdump i Rizin](../reverse-engineering-tools.md) · [Dokumentacja](#dokumentacja)

 <code>npx rea-agents setup</code>

+> [!IMPORTANT]
+> Oficjalne źródła REA: [morluto/rea](https://github.com/morluto/rea), [rea.tools](https://rea.tools/), [rea-agents](https://www.npmjs.com/package/rea-agents) (npm).
+> Korzystaj z odnośników do instalacji i wydań w tych oficjalnych źródłach. Podobna nazwa repozytorium lub skopiowany README nie potwierdzają, że pobierany plik jest oficjalny.
+
 <br />

 <img src="../assets/rea-hopper-analysis.png" alt="REA uruchamia most analityczny w Hopperze podczas analizy natywnego pliku binarnego" width="1200" />
@@ -1,6 +1,7 @@
 <div align="center">

 [English](../../README.md) · [简体中文](README_zh.md) · [繁體中文](README_zh-TW.md) · [日本語](README_ja.md) · [한국어](README_ko.md) · [Türkçe](README_tr.md) · [Русский](README_ru.md) · [Tiếng Việt](README_vi.md) · [ไทย](README_th.md) · [Deutsch](README_de.md) · [Español](README_es.md) · [Français](README_fr.md) · [Українська](README_uk.md) · [Polski](README_pl.md) · **Português (Brasil)** · [العربية](README_ar.md) · [فارسی](README_fa.md) · [Bahasa Indonesia](README_id.md) · [Italiano](README_it.md)
+· [Ελληνικά](README_el.md)

 # REA: Faça engenharia reversa de qualquer coisa

@@ -19,10 +20,14 @@

 **[Site](https://rea.tools/pt-BR/) · [Guias](https://rea.tools/pt-BR/guides/) · [Casos de uso](https://rea.tools/pt-BR/showcase/)**

-[Início rápido](#início-rápido) · [Como o REA funciona](#como-o-rea-funciona) · [O que você pode analisar](#o-que-você-pode-analisar) · [Casos de uso](#casos-de-uso) · [Perguntas frequentes](#perguntas-frequentes) · [Documentação](#documentação)
+[Início rápido](#início-rápido) · [Como o REA funciona](#como-o-rea-funciona) · [O que você pode analisar](#o-que-você-pode-analisar) · [Casos de uso](#casos-de-uso) · [Perguntas frequentes](#perguntas-frequentes) · [Ferramentas GNU objdump e Rizin](../reverse-engineering-tools.md) · [Documentação](#documentação)

 <code>npx rea-agents setup</code>

+> [!IMPORTANT]
+> Fontes oficiais do REA: [morluto/rea](https://github.com/morluto/rea), [rea.tools](https://rea.tools/), [rea-agents](https://www.npmjs.com/package/rea-agents) (npm).
+> Use os links de instalação e de versões dessas fontes oficiais. Um nome de repositório parecido ou um README copiado não comprova que um download seja oficial.
+
 <br />

 <img src="../assets/rea-hopper-analysis.png" alt="REA inicia sua ponte de análise dentro do Hopper enquanto inspeciona um binário nativo" width="1200" />
@@ -1,6 +1,7 @@
 <div align="center">

 [English](../../README.md) · [简体中文](README_zh.md) · [繁體中文](README_zh-TW.md) · [日本語](README_ja.md) · [한국어](README_ko.md) · [Türkçe](README_tr.md) · **Русский** · [Tiếng Việt](README_vi.md) · [ไทย](README_th.md) · [Deutsch](README_de.md) · [Español](README_es.md) · [Français](README_fr.md) · [Українська](README_uk.md) · [Polski](README_pl.md) · [Português (Brasil)](README_pt-BR.md) · [العربية](README_ar.md) · [فارسی](README_fa.md) · [Bahasa Indonesia](README_id.md) · [Italiano](README_it.md)
+· [Ελληνικά](README_el.md)

 # REA: Реверс-инжиниринг чего угодно

@@ -19,10 +20,14 @@

 **[Сайт](https://rea.tools/ru/) · [Руководства](https://rea.tools/ru/guides/) · [Примеры](https://rea.tools/ru/showcase/)**

-[Быстрый старт](#быстрый-старт) · [Как работает REA](#как-работает-rea) · [Что можно анализировать](#что-можно-анализировать) · [Примеры](#примеры) · [Частые вопросы](#частые-вопросы) · [Документация](#документация)
+[Быстрый старт](#быстрый-старт) · [Как работает REA](#как-работает-rea) · [Что можно анализировать](#что-можно-анализировать) · [Примеры](#примеры) · [Частые вопросы](#частые-вопросы) · [Инструменты GNU objdump и Rizin](../reverse-engineering-tools.md) · [Документация](#документация)

 <code>npx rea-agents setup</code>

+> [!IMPORTANT]
+> Официальные источники REA: [morluto/rea](https://github.com/morluto/rea), [rea.tools](https://rea.tools/), [rea-agents](https://www.npmjs.com/package/rea-agents) (npm).
+> Используйте ссылки на установку и выпуски из этих официальных источников. Похожее название репозитория или скопированный README не подтверждают, что скачиваемый файл является официальным.
+
 <br />

 <img src="../assets/rea-hopper-analysis.png" alt="REA запускает мост анализа внутри Hopper при исследовании нативного бинарного файла" width="1200" />
@@ -1,6 +1,7 @@
 <div align="center">

 [English](../../README.md) · [简体中文](README_zh.md) · [繁體中文](README_zh-TW.md) · [日本語](README_ja.md) · [한국어](README_ko.md) · [Türkçe](README_tr.md) · [Русский](README_ru.md) · [Tiếng Việt](README_vi.md) · **ไทย** · [Deutsch](README_de.md) · [Español](README_es.md) · [Français](README_fr.md) · [Українська](README_uk.md) · [Polski](README_pl.md) · [Português (Brasil)](README_pt-BR.md) · [العربية](README_ar.md) · [فارسی](README_fa.md) · [Bahasa Indonesia](README_id.md) · [Italiano](README_it.md)
+· [Ελληνικά](README_el.md)

 # REA: Reverse Engineer Anything

@@ -19,10 +20,14 @@

 **[เว็บไซต์](https://rea.tools/th/) · [คู่มือ](https://rea.tools/th/guides/) · [ตัวอย่างการใช้งาน](https://rea.tools/th/showcase/)**

-[เริ่มต้นอย่างรวดเร็ว](#เริ่มต้นอย่างรวดเร็ว) · [REA ทำงานอย่างไร](#rea-ทำงานอย่างไร) · [สิ่งที่วิเคราะห์ได้](#สิ่งที่วิเคราะห์ได้) · [ตัวอย่างการใช้งาน](#ตัวอย่างการใช้งาน) · [คำถามที่พบบ่อย](#คำถามที่พบบ่อย) · [เอกสาร](#เอกสาร)
+[เริ่มต้นอย่างรวดเร็ว](#เริ่มต้นอย่างรวดเร็ว) · [REA ทำงานอย่างไร](#rea-ทำงานอย่างไร) · [สิ่งที่วิเคราะห์ได้](#สิ่งที่วิเคราะห์ได้) · [ตัวอย่างการใช้งาน](#ตัวอย่างการใช้งาน) · [คำถามที่พบบ่อย](#คำถามที่พบบ่อย) · [เครื่องมือ GNU objdump และ Rizin](../reverse-engineering-tools.md) · [เอกสาร](#เอกสาร)

 <code>npx rea-agents setup</code>

+> [!IMPORTANT]
+> แหล่งทางการของ REA: [morluto/rea](https://github.com/morluto/rea), [rea.tools](https://rea.tools/), [rea-agents](https://www.npmjs.com/package/rea-agents) (npm).
+> โปรดใช้ลิงก์ติดตั้งและรุ่นเผยแพร่จากแหล่งทางการเหล่านี้ ชื่อที่เก็บโค้ดที่คล้ายกันหรือ README ที่คัดลอกมาไม่ได้ยืนยันว่าไฟล์ดาวน์โหลดเป็นของทางการ
+
 <br />

 <img src="../assets/rea-hopper-analysis.png" alt="REA เริ่มบริดจ์การวิเคราะห์ภายใน Hopper ขณะตรวจสอบไบนารีเนทีฟ" width="1200" />
@@ -1,6 +1,7 @@
 <div align="center">

 [English](../../README.md) · [简体中文](README_zh.md) · [繁體中文](README_zh-TW.md) · [日本語](README_ja.md) · [한국어](README_ko.md) · **Türkçe** · [Русский](README_ru.md) · [Tiếng Việt](README_vi.md) · [ไทย](README_th.md) · [Deutsch](README_de.md) · [Español](README_es.md) · [Français](README_fr.md) · [Українська](README_uk.md) · [Polski](README_pl.md) · [Português (Brasil)](README_pt-BR.md) · [العربية](README_ar.md) · [فارسی](README_fa.md) · [Bahasa Indonesia](README_id.md) · [Italiano](README_it.md)
+· [Ελληνικά](README_el.md)

 # REA: Her şeyi tersine mühendislikle inceleyin

@@ -19,10 +20,14 @@

 **[Web sitesi](https://rea.tools/tr/) · [Rehberler](https://rea.tools/tr/guides/) · [Örnek çalışmalar](https://rea.tools/tr/showcase/)**

-[Hızlı başlangıç](#hızlı-başlangıç) · [REA nasıl çalışır](#rea-nasıl-çalışır) · [Neleri analiz edebilirsiniz](#neleri-analiz-edebilirsiniz) · [Örnek çalışmalar](#örnek-çalışmalar) · [Sık sorulan sorular](#sık-sorulan-sorular) · [Belgeler](#belgeler)
+[Hızlı başlangıç](#hızlı-başlangıç) · [REA nasıl çalışır](#rea-nasıl-çalışır) · [Neleri analiz edebilirsiniz](#neleri-analiz-edebilirsiniz) · [Örnek çalışmalar](#örnek-çalışmalar) · [Sık sorulan sorular](#sık-sorulan-sorular) · [GNU objdump ve Rizin araçları](../reverse-engineering-tools.md) · [Belgeler](#belgeler)

 <code>npx rea-agents setup</code>

+> [!IMPORTANT]
+> REA’nın resmî kaynakları: [morluto/rea](https://github.com/morluto/rea), [rea.tools](https://rea.tools/), [rea-agents](https://www.npmjs.com/package/rea-agents) (npm).
+> Kurulum ve sürüm bağlantılarını bu resmî kaynaklardan takip edin. Benzer bir depo adı veya kopyalanmış bir README, indirilen dosyanın resmî olduğunu kanıtlamaz.
+
 <br />

 <img src="../assets/rea-hopper-analysis.png" alt="REA, yerel bir ikili dosyayı incelerken Hopper içinde analiz köprüsünü başlatıyor" width="1200" />
@@ -137,6 +142,7 @@ REA; Node.js 22.x (>=22.19), 24.x (>=24.11) veya 26+ ve npm gerektirir. Ek araç
 | Yerel makine kodu içeren ikili dosyalar | Sözde kod, assembly kodu, dizeler, semboller, çağrılar ve referanslar                                       | Hopper, Ghidra veya IDA; [yerel kod analizi](https://rea.tools/tr/guides/native/)                                                  |
 | Çevrimdışı ELF yapısı                   | Bölümler, segmentler, özgün semboller/yeniden konumlandırmalar ve statik analizden olası güvenlik önlemleri | Linux x64 üzerinde çağıranın sağladığı pwntools; [ikili dosya tanılama](../binary-diagnostics.md)                                  |
 | Go ikili dosyaları                      | Gömülü compiler sürümü, modüller, replacement'lar ve build ayarları                                         | Yerleşik statik ELF/PE/thin Mach-O incelemesi; Go kurulumu gerekmez                                                                |
+| Rust ikili dosyaları                    | Gömülü cargo-auditable crate adları, sürümleri, kaynakları ve bağımlılık ilişkileri                         | Yerleşik statik ELF/PE/thin Mach-O incelemesi; [Rust ikili dosyaları](../rust-binaries.md)                                         |
 | WASM artefaktı                          | Doğrulama, bölümler, tam import/export WAT biçimleri ve decode edilmiş WAT                                  | Çağıranın sağladığı WABT 1.0.42; `inspect_wasm_artifact`                                                                           |
 | EVM bayt kodu                           | Yönlendirme seçicileri, bayt ofsetleri, çıkarımla belirlenen argümanlar ve durum değiştirilebilirliği       | Linux x64 üzerinde yerel ham bayt/onaltılık girdi; [çevrimdışı EVM rehberi](../evm-bytecode.md)                                    |
 | SQLite veritabanları                    | Tablolar, sütunlar, index'ler, view'lar, trigger'lar ve seçilen sıradan tablo satırları                     | Yerel database/WAL snapshot'ı; native SQLite sınırlarıyla Node.js 24.x >=24.15 veya 26+                                            |
@@ -1,6 +1,7 @@
 <div align="center">

 [English](../../README.md) · [简体中文](README_zh.md) · [繁體中文](README_zh-TW.md) · [日本語](README_ja.md) · [한국어](README_ko.md) · [Türkçe](README_tr.md) · [Русский](README_ru.md) · [Tiếng Việt](README_vi.md) · [ไทย](README_th.md) · [Deutsch](README_de.md) · [Español](README_es.md) · [Français](README_fr.md) · **Українська** · [Polski](README_pl.md) · [Português (Brasil)](README_pt-BR.md) · [العربية](README_ar.md) · [فارسی](README_fa.md) · [Bahasa Indonesia](README_id.md) · [Italiano](README_it.md)
+· [Ελληνικά](README_el.md)

 # REA: Реверс-інжиніринг будь-чого

@@ -19,10 +20,14 @@

 **[Сайт](https://rea.tools/uk/) · [Посібники](https://rea.tools/uk/guides/) · [Приклади](https://rea.tools/uk/showcase/)**

-[Швидкий старт](#швидкий-старт) · [Як працює REA](#як-працює-rea) · [Що можна аналізувати](#що-можна-аналізувати) · [Приклади](#приклади) · [Поширені запитання](#поширені-запитання) · [Документація](#документація)
+[Швидкий старт](#швидкий-старт) · [Як працює REA](#як-працює-rea) · [Що можна аналізувати](#що-можна-аналізувати) · [Приклади](#приклади) · [Поширені запитання](#поширені-запитання) · [Інструменти GNU objdump і Rizin](../reverse-engineering-tools.md) · [Документація](#документація)

 <code>npx rea-agents setup</code>

+> [!IMPORTANT]
+> Офіційні джерела REA: [morluto/rea](https://github.com/morluto/rea), [rea.tools](https://rea.tools/), [rea-agents](https://www.npmjs.com/package/rea-agents) (npm).
+> Користуйтеся посиланнями на встановлення та випуски з цих офіційних джерел. Схожа назва репозиторію або скопійований README не підтверджують, що завантажуваний файл є офіційним.
+
 <br />

 <img src="../assets/rea-hopper-analysis.png" alt="REA запускає міст аналізу всередині Hopper під час дослідження нативного бінарного файла" width="1200" />
@@ -1,6 +1,7 @@
 <div align="center">

 [English](../../README.md) · [简体中文](README_zh.md) · [繁體中文](README_zh-TW.md) · [日本語](README_ja.md) · [한국어](README_ko.md) · [Türkçe](README_tr.md) · [Русский](README_ru.md) · **Tiếng Việt** · [ไทย](README_th.md) · [Deutsch](README_de.md) · [Español](README_es.md) · [Français](README_fr.md) · [Українська](README_uk.md) · [Polski](README_pl.md) · [Português (Brasil)](README_pt-BR.md) · [العربية](README_ar.md) · [فارسی](README_fa.md) · [Bahasa Indonesia](README_id.md) · [Italiano](README_it.md)
+· [Ελληνικά](README_el.md)

 # REA: Kỹ thuật dịch ngược cho mọi thứ

@@ -19,10 +20,14 @@

 **[Trang web](https://rea.tools/vi/) · [Hướng dẫn](https://rea.tools/vi/guides/) · [Các nghiên cứu điển hình](https://rea.tools/vi/showcase/)**

-[Bắt đầu nhanh](#bắt-đầu-nhanh) · [Cách REA hoạt động](#cách-rea-hoạt-động) · [Những gì bạn có thể phân tích](#những-gì-bạn-có-thể-phân-tích) · [Các nghiên cứu điển hình](#các-nghiên-cứu-điển-hình) · [Câu hỏi thường gặp](#câu-hỏi-thường-gặp) · [Tài liệu](#tài-liệu)
+[Bắt đầu nhanh](#bắt-đầu-nhanh) · [Cách REA hoạt động](#cách-rea-hoạt-động) · [Những gì bạn có thể phân tích](#những-gì-bạn-có-thể-phân-tích) · [Các nghiên cứu điển hình](#các-nghiên-cứu-điển-hình) · [Câu hỏi thường gặp](#câu-hỏi-thường-gặp) · [Công cụ GNU objdump và Rizin](../reverse-engineering-tools.md) · [Tài liệu](#tài-liệu)

 <code>npx rea-agents setup</code>

+> [!IMPORTANT]
+> Các nguồn chính thức của REA: [morluto/rea](https://github.com/morluto/rea), [rea.tools](https://rea.tools/), [rea-agents](https://www.npmjs.com/package/rea-agents) (npm).
+> Hãy dùng các liên kết cài đặt và bản phát hành từ những nguồn chính thức này. Tên kho mã tương tự hoặc README được sao chép không chứng minh rằng tệp tải xuống là bản chính thức.
+
 <br />

 <img src="../assets/rea-hopper-analysis.png" alt="REA khởi chạy cầu nối phân tích bên trong Hopper khi kiểm tra tệp nhị phân mã máy" width="1200" />
@@ -1,6 +1,7 @@
 <div align="center">

 [English](../../README.md) · [简体中文](README_zh.md) · **繁體中文** · [日本語](README_ja.md) · [한국어](README_ko.md) · [Türkçe](README_tr.md) · [Русский](README_ru.md) · [Tiếng Việt](README_vi.md) · [ไทย](README_th.md) · [Deutsch](README_de.md) · [Español](README_es.md) · [Français](README_fr.md) · [Українська](README_uk.md) · [Polski](README_pl.md) · [Português (Brasil)](README_pt-BR.md) · [العربية](README_ar.md) · [فارسی](README_fa.md) · [Bahasa Indonesia](README_id.md) · [Italiano](README_it.md)
+· [Ελληνικά](README_el.md)

 # REA：逆向一切

@@ -19,10 +20,14 @@

 **[網站](https://rea.tools/zh-TW/) · [指南](https://rea.tools/zh-TW/guides/) · [案例](https://rea.tools/zh-TW/showcase/)**

-[快速開始](#快速開始) · [REA 的工作原理](#rea-的工作原理) · [可以分析什麼](#可以分析什麼) · [案例](#案例) · [常見問題](#常見問題) · [文件](#文件)
+[快速開始](#快速開始) · [REA 的工作原理](#rea-的工作原理) · [可以分析什麼](#可以分析什麼) · [案例](#案例) · [常見問題](#常見問題) · [GNU objdump 與 Rizin 工具](../reverse-engineering-tools.md) · [文件](#文件)

 <code>npx rea-agents setup</code>

+> [!IMPORTANT]
+> REA 官方來源： [morluto/rea](https://github.com/morluto/rea), [rea.tools](https://rea.tools/), [rea-agents](https://www.npmjs.com/package/rea-agents) (npm).
+> 請透過這些官方來源提供的安裝與發布連結取得 REA。相似的儲存庫名稱或複製的 README 不能證明下載內容來自官方。
+
 <br />

 <img src="../assets/rea-hopper-analysis.png" alt="REA 在 Hopper 中啟動分析橋，檢查原生二進位檔案" width="1200" />
@@ -1,6 +1,7 @@
 <div align="center">

 [English](../../README.md) · **简体中文** · [繁體中文](README_zh-TW.md) · [日本語](README_ja.md) · [한국어](README_ko.md) · [Türkçe](README_tr.md) · [Русский](README_ru.md) · [Tiếng Việt](README_vi.md) · [ไทย](README_th.md) · [Deutsch](README_de.md) · [Español](README_es.md) · [Français](README_fr.md) · [Українська](README_uk.md) · [Polski](README_pl.md) · [Português (Brasil)](README_pt-BR.md) · [العربية](README_ar.md) · [فارسی](README_fa.md) · [Bahasa Indonesia](README_id.md) · [Italiano](README_it.md)
+· [Ελληνικά](README_el.md)

 # REA：逆向一切

@@ -19,10 +20,14 @@

 **[网站](https://rea.tools/zh-CN/) · [指南](https://rea.tools/zh-CN/guides/) · [案例](https://rea.tools/zh-CN/showcase/)**

-[快速开始](#快速开始) · [REA 的工作原理](#rea-的工作原理) · [可以分析什么](#可以分析什么) · [案例](#案例) · [常见问题](#常见问题) · [文档](#文档)
+[快速开始](#快速开始) · [REA 的工作原理](#rea-的工作原理) · [可以分析什么](#可以分析什么) · [案例](#案例) · [常见问题](#常见问题) · [GNU objdump 和 Rizin 工具](../reverse-engineering-tools.md) · [文档](#文档)

 <code>npx rea-agents setup</code>

+> [!IMPORTANT]
+> REA 官方来源： [morluto/rea](https://github.com/morluto/rea), [rea.tools](https://rea.tools/), [rea-agents](https://www.npmjs.com/package/rea-agents) (npm).
+> 请通过这些官方来源提供的安装和发布链接获取 REA。相似的仓库名称或复制的 README 不能证明下载内容来自官方。
+
 <br />

 <img src="../assets/rea-hopper-analysis.png" alt="REA 在 Hopper 中启动分析桥，检查原生二进制文件" width="1200" />
@@ -0,0 +1,81 @@
+# Reverse-engineering tool integrations
+
+REA can call caller-supplied GNU objdump, GDB, and Rizin installations. REA does not install or update these tools. Static objdump/Rizin workflows use the same services from CLI and MCP. GDB's persistent MCP sessions and ephemeral CLI command are available only on Linux. Persistent debugger sessions exist only for the lifetime of the REA MCP server process.
+
+In a shared MCP server, unrestricted Rizin and Cutter commands, GDB console
+commands, Rizin debugger start/close, and GDB debugger close retire any active
+binary analysis session after the operation. This prevents a fresh analysis
+provider from being rebound automatically while command effects may still be
+writing. After the command and any background writers have stopped, call
+`open_binary` explicitly before continuing analysis. A successful command
+response alone does not prove that background writers have stopped. An active
+Ghidra analysis session must be closed explicitly before an external mutation
+can run.
+
+## GNU objdump
+
+`inspect_with_objdump` and `rea inspect-with-objdump` capture stdout and stderr as inline Evidence. Artifact inspection Evidence records SHA-256 digests before and after the provider call and whether those observed digests differ; this cannot prove the exact bytes consumed if the artifact changes during execution. Choose one artifact operation: `file_headers`, `section_headers`, `symbols`, `relocations`, `dwarf`, `disassemble`, or `disassemble_all`. `disassemble` maps to `-d`; `disassemble_all` maps to `-D` and has different symbol-boundary semantics. The pathless `inspect_with_objdump {"operation":"architectures"}` MCP call or `rea list-objdump-architectures` CLI command maps to `-i` and produces subjectless provider-capability Evidence describing BFD targets compiled into the installed binary, not a guarantee that a target is supported.
+
+For DWARF, REA does not follow local debug links by default. Set `follow_debug_links` to `true` to pass `-WK`; external debug-file identities are not bound to Evidence when following links is enabled. The CLI option is `--follow-debug-links true`. REA clears `DEBUGINFOD_URLS` in the objdump child environment to disable debuginfod network access on GNU binutils versions that do not support `-WE`. The complete raw output is authoritative; REA does not normalize it into a version-independent disassembly schema.
+
+```sh
+rea inspect-with-objdump ./program disassemble
+rea inspect-with-objdump ./program dwarf --follow-debug-links false
+```
+
+Captured stdout and stderr are retained up to a 16 MiB combined limit; `output_truncated` marks output beyond that limit.
+
+## GDB
+
+Persistent GDB sessions, including the MCP workflow and one-shot `rea debug-with-gdb` CLI, are available only on Linux. REA verifies a live inferior's output descriptors through `/proc` before detaching it; it does not claim equivalent descriptor verification on macOS. Windows is also unsupported because owned Job Object cleanup can terminate debugger inferiors. On supported Linux hosts, `start_gdb_session`, `gdb_console`, `gdb_session_status`, and `close_gdb_session` manage a persistent GDB/MI session in MCP. The one-shot CLI requires a caller-selected durable TTY: `rea debug-with-gdb --inferior-tty /dev/pts/7 'file ./program'`. The TTY must remain available after GDB exits so a detached inferior can continue writing. MCP starts GDB with `--nx --quiet --interpreter=mi3`, retries initialization with MI2 if MI3 initialization returns an error and cleanup succeeds, waits for its prompt, and disables auto-loading. Pass optional `inferior_tty` to select the terminal that future inferior runs use; REA configures it before the session becomes ready and checks the exact value with GDB/MI. `gdb_session_status` reports the selected path, whether REA still considers the configuration verified, and whether invalid UTF-8 has been observed in MI bytes (`mi_invalid_utf8`) or stderr (`stderr_invalid_utf8`). Commands that can change the inferior TTY or redirect a run revoke that verification permanently for the session, so close retains GDB ownership if a live inferior remains. Closing a session with one live inferior and a verified TTY automatically detaches the target, which can resume execution; MCP clients receive destructive-operation annotations for this behavior. The console remains unrestricted and can run shell commands, change target state, access files, and use network in the REA process user's security context. Raw MI records and console/target/log streams are returned with Evidence; command output is retained up to 16 MiB and explicitly marked when truncated.
+
+Before a normal close, REA checks for live GDB thread groups. If one remains and a durable `inferior_tty` is verified, REA asks GDB to detach it and confirms the detach before stopping the debugger. Without that sink, select and apply a disposition to each inferior through `gdb_console` (for example, detach or kill as appropriate), then retry close. REA never infers target ownership from a PID or path. When the MCP server shuts down, REA terminates only its directly owned GDB process after confirming no inferior or a sink-backed detach; otherwise it retains the process and its output streams for retry.
+
+```text
+start_gdb_session {"inferior_tty":"/dev/pts/7"}
+gdb_console {"session_id":"…","command":"file ./program"}
+gdb_console {"session_id":"…","command":"info files"}
+gdb_session_status {"session_id":"…"}
+close_gdb_session {"session_id":"…"}
+```
+
+## Rizin
+
+`start_rizin_debug_session` and `rea debug-with-rizin` require an explicit `inferior_tty` on Linux. Rizin launches the target with stdout and stderr routed to that caller-selected TTY and `dbg.exitkills=false`; the target can therefore continue running and writing to the TTY after REA closes the debugger. This is a visible target output effect, and the caller is responsible for selecting a suitable terminal that remains available. Before closing, REA serializes after the active command, parses the structured `olj` descriptor list, detaches only a unique `ptrace://` descriptor, confirms its absence, reads back `dbg.exitkills=false`, and verifies `/proc/<pid>/fd/1` and `/proc/<pid>/fd/2` still identify the selected TTY by device and inode. Any malformed or ambiguous readback keeps the Rizin session owned for retry. Persistent Rizin debugger sessions are unavailable outside Linux because this verification depends on `/proc`.
+
+Persistent command Evidence retains the caller-selected and resolved startup paths in `parameters.launch_provenance`. These paths describe launch provenance, not the current target after unrestricted commands. `rizin_debug_session_status.output_invalid_utf8` reports whether lossy frame decoding has been observed during the session. Command Evidence sets `output_invalid_utf8` only when loss was observed in frames collected for that command, including retained partial output; earlier startup or command frames do not mark later valid output as lossy.
+
+```sh
+rea debug-with-rizin --inferior-tty /dev/pts/7 ./program dr
+```
+
+`execute_rizin_command` and `rea execute-rizin-command` run one selected command against a local artifact. The one-shot profile uses `-N -q -c`, suppressing user settings and scripts while leaving plugins available. Commands and plugins run with the REA process user's security context and may access files, the network, or perform other actions. REA preserves raw output and does not claim that command-specific JSON has one universal schema.
+
+```sh
+rea execute-rizin-command ./program iI
+```
+
+Captured stdout and stderr are retained up to a 16 MiB combined limit; `output_truncated` marks output beyond that limit.
+
+## Cutter bridge
+
+Cutter commands use the local REA Python bridge and authenticated loopback messages.
+The client and plugin must support the current v2 handshake; older peers are
+rejected before dispatch because they cannot preserve cross-caller recovery proof.
+The client checks the final encoded request against the bridge's 16 MiB frame
+limit before dispatch; an oversized command does not retain a cleanup owner.
+Discovery admits at most 256 descriptor candidates and reads and probes at most
+eight concurrently. A larger directory reports `bridge_directory_capacity_exceeded`.
+
+Terminal receipts are recovered before another command, including from a new CLI
+invocation. A signed status response also confirms completion of older bridge
+connections after its UI dispatch, preserving recovery proof for a caller whose
+receipt another caller has acknowledged. Neither receipt recovery nor a status
+probe retries the original command. Failed acknowledgements remain pending and
+block new commands until they can be retried successfully.
+
+## Configuration and coverage
+
+The optional `REA_OBJDUMP_COMMAND`, `REA_GDB_COMMAND`, and `REA_RIZIN_COMMAND` values select executable names or paths; otherwise REA searches `PATH`. Values are passed directly to process spawning, not expanded as shell command lines. Provider versions and capabilities remain unknown until observed from the installed tool and verified in the corresponding real-provider lane.
+
+Provider capture limits are separate from the MCP response budget. A result that fits the provider limit may still exceed the default 10 MiB MCP response budget and return a transport `resource_constraint`. Configure `REA_MCP_MAX_RESPONSE_BYTES` to match a larger client receive buffer when that client supports one; otherwise use retained Evidence/export or complete CLI JSON output. See [MCP response limits](./mcp-contracts.md#tool-results).
@@ -262,8 +262,33 @@ CI owns coverage; aggregate and domain/contract thresholds are maintained in
 `vitest.config.ts`. They are glob-specific and never updated automatically.
 Coverage does not replace named boundary or real-provider scenarios.

-CI runs six native Vitest shards without retries. Each shard emits a blob
-report; the merge job produces JUnit and JSON diagnostics, a module timing
+The `reverse-tools` commands run the focused tests and report coverage and
+mutation results without an additional percentage requirement. Test failures,
+invalid mutation configuration, and runner failures still fail the check.
+Mutation targets cover reviewed production behavior, including signal-terminated
+commands, Cutter descriptor identity, debugger cleanup ownership, GDB command
+exit and cancellation, and Rizin startup, timeout Evidence, and process exit.
+A second exhaustive pass runs the focused Cutter and Rizin timeout semantics tests against every compilable
+selected mutant to avoid relying on per-test mutation coverage associations;
+Stryker reports compile-error mutants separately because those cannot execute
+tests. Both passes resolve named syntax owners, scoped declarations, guards,
+callbacks, and reviewed expressions from the current TypeScript source; no
+hand-maintained line ranges select mutations. Missing, ambiguous, or malformed
+source targets stop configuration before mutation generation. The selections
+retain their review scope rather than expanding to whole source files. The
+focused suite also exercises affected MCP routes and the Python Cutter bridge.
+Contract, integration, and real-provider lanes complement these focused tests;
+simulated providers do not establish real-provider behavior.
+
+Fast CI starts the source-only projects after classification, independently of
+build and static checks, and runs five built-dependent shards after the build.
+The six test jobs still select every portable test. Full qualification keeps six
+complete Vitest shards with coverage. Neither tier retries tests. Canonical
+project ownership lives in `scripts/lib/test-projects.mjs`, shared by Vitest,
+CI group execution, and development-test preparation. A test importing compiled
+verifier code belongs to the built group even if its directory is conformance.
+Source and built blobs have separate labels to prevent merge collisions.
+Each test job emits a blob report; the merge job produces JUnit and JSON diagnostics, a module timing
 report, plus aggregate coverage during full qualification. It retains available
 reports even when a shard fails, uploads available diagnostics, and writes the
 slowest files to the workflow summary. Partial reports are marked diagnostic
@@ -298,6 +323,30 @@ portable test shard. These limits leave headroom for cold setup and full coverag
 while preventing a hung job from holding the active main run for hours. They do
 not bound runner queue waiting. Timeouts still fail the required gate.

+The separate `CI timings` workflow also diagnoses missing, startup-failed,
+queued/running, approval-required, canceled, and failed PR CI through bounded,
+read-only API scans. Incomplete pagination or changing identity stays unknown.
+It never closes/reopens PRs, reruns workflows, posts comments, or changes gates.
+The timing manifest retains each run's event and head branch, including explicit
+unknowns, so PR runs cannot consume the eligible-main comparison budget.
+Its static failure report compares supported lint/duplication diagnostics with
+failed main runs using exact source, ancestry, task hashes, and toolchain proof.
+A match describes those diagnostics only; it does not establish that every
+failure in a job came from main. Duplication reports retain the actual checker
+exit result, so passing duplication checks preserve lint failures and caller
+thresholds remain owned by the producer. Missing checker outcomes leave
+those duplication diagnostics unclassified. Missing historical receipts remain unknown.
+The reporter executes only trusted default-branch code and parses whitelisted
+JSON artifact members without executing artifact content. Reports and limits
+are retained in `ci-health-diagnostics`; its ten-minute job limit covers the
+existing four-minute timing collection plus the bounded diagnostic stages.
+
+Shard-count reduction remains deferred: the available successful timing sample
+covers an older inventory, and file phase weights are not job wall time. The
+one-source/five-built split retains six test jobs; it makes no hosted speedup
+claim. Current complete timings and completed candidate hosted runs are needed
+before enabling a smaller adaptive matrix.
+
 The separate `CI timings` workflow batches recent completed CI and Release runs
 hourly or on manual dispatch. It uploads `ci-job-timings-batch` with raw API pages and
 a normalized point-in-time report for each collected workflow attempt, plus a
@@ -346,6 +395,17 @@ that explicit lane, not the routine iteration requirement.

 ## Real-toolchain verification lanes

+Linux Rizin debugger ownership changes require a separate real-provider gate on
+Rizin 0.7.4. Run
+`npm run test:focused -- tests/boundary/providers/RizinInferiorTty.test.ts`.
+This test verifies the caller-selected TTY against the actual inferior's file
+descriptors, confirms `dbg.exitkills=false`, detaches through the `olj`/`o-`
+protocol, and checks that the target survives debugger close and writes to the
+selected TTY. The suite is included in the reverse-tools Vitest and mutation
+typecheck selections; its real-provider assertions skip when Rizin 0.7.4 is
+unavailable. Run the focused command above on a host with that provider for the
+explicit real-provider verification gate.
+
 Each real-toolchain command must require only the host tools needed to prove
 its stated claim. Use a host-native fixture for host/provider acceptance, and
 place optional cross-target formats or platform-specific runners in separate
@@ -887,8 +947,13 @@ successful ancestor push run of this CI workflow, rather than the preceding
 push: changes from failed or cancelled intermediate runs remain in scope.
 Missing, untrusted, or non-ancestor history selects every lane. The lookup is
 bounded to the last 100 successful push runs; lookup failures broaden selection.
-Every code change still runs all six portable shards, without coverage
-instrumentation in this fast tier. This deliberately avoids relying on an
+Every code change still selects all six portable shards, without coverage
+instrumentation in this fast tier. Ordinary PR matrices cancel remaining
+shards after the first failure; passing PRs execute the complete suite. Main,
+`ci:full`, scheduled, manual, and release qualification retain every shard's
+results after a failure. Available failed-shard reports remain diagnostic only:
+failed, cancelled, skipped, or missing selected shards cannot pass the gate or
+publish successful timing history. This deliberately avoids relying on an
 import graph for generated files, filesystem inputs, subprocesses, or bridges.

 Scheduled qualification is admitted hourly at minute 23. A forced full run at
@@ -916,10 +981,11 @@ classification job records its selection and broad-fallback reasons in the run
 summary. Representative Git-diff cases live in
 `tests/boundary/filesystem/ciChangeScope.test.ts`.

-`CI required` always runs and checks classification and every selected lane,
+`CI required` checks classification and every selected lane,
 rejecting failed, cancelled, or unexpectedly skipped work. It rejects missing or
 unknown tiers and requires every canonical scope to be explicitly `true` or
 `false`; incomplete classification cannot silently omit verification lanes.
+The aggregate remains cancellable so superseded runs release their slots.
 Branch protection
 should require this aggregate, `Static checks`, `Portable tests`, and
 `Conventional PR title`. Title validation runs separately on PR opening,
@@ -968,7 +1034,7 @@ filename cases.

 POSIX package verification resolves the runner's configured npm cache before
 creating disposable client homes and passes that location to its child installs.
-This reuses the download cache restored by `setup-node` while keeping client
+This reuses the download cache restored by the shared npm-cache action while keeping client
 configuration isolated. Installs retain their existing registry and integrity
 checks.

@@ -998,7 +1064,24 @@ thresholds remain enforced in full qualification.
 Missing or unusable history falls back to Vitest's default assignment. Only
 timing data is reused; live provider results are never cached. Turbo outputs are
 restored by task hash, with trusted default-branch runs as cache writers and PRs
-as restore-only consumers. npm caches hold downloads, not `node_modules`.
+as restore-only consumers. The static job also restores its Turbo snapshot;
+each task revalidates its own input hash before reusing outputs or diagnostics.
+The Node pin and npm install settings are global hash inputs. Knip includes its ignore rules and VitePress
+entry, so changes to those inputs cannot reuse an older successful check.
+
+The shared npm-cache action caches downloads, not `node_modules`, and every
+consumer still runs `npm ci`. Keys cover the runner OS, actual Node architecture,
+root lockfiles, and `.npmrc`. Only push, schedule, and manual events on the
+repository's default branch may save shared npm caches; PRs restore only.
+Static snapshots use the same write authority and are saved only after the
+static job's validation succeeds. Main warms the new cache namespace after
+merge; a cold cache or changed lockfile falls back to the normal install.
+This prevents each PR from saving another branch-scoped copy of the same
+download cache. It does not remove runner assignment waits.
+
+Portable owner-based test selection remains observational. Its reports compare
+proposed omissions with the full run; incomplete or skipped observations do not
+establish that fewer tests would preserve regression detection.

 The eight automatic browser, binary-layout, EVM, JavaScript-recovery,
 recorded-crash, and web provider lanes download `rea-build-cache` from their own
@@ -224,6 +224,22 @@ are 0 and 50. The view groups request lines by method, WHATWG URL origin and
 path template, sorted by origin, template and method. Query strings, fragments,
 credentials, headers and bodies are not copied into groups.

+Select `include_query_parameter_names: true` in the view to include distinct
+observed query names per group. The default and explicit `false` omit the
+`query_parameter_names` field. Selected groups include an array of
+`{name, record_ordinals}` entries, sorted by Unicode code point; each name's
+ordinals follow the parent's selected record order and occur once per record.
+For `/products?category=books&page=1` and
+`/products?category=games&sort=price`, names are `category`, `page` and `sort`.
+Query values and fragments are not copied into this inventory.
+
+Names use WHATWG `URLSearchParams` form-query decoding: literal `+` becomes
+space, `%2B` becomes `+`, percent escapes decode as UTF-8, and empty names are
+retained. Empty query pairs are ignored. Original URLs stay in parent Evidence;
+name decoding does not normalize the recorded paths used for grouping.
+The inventory describes observed names, without inferring required parameters,
+value types or a complete API schema. Sensitive-parent refusals still apply.
+
 | Rule       | Exact segment format                                 | Placeholder  |
 | ---------- | ---------------------------------------------------- | ------------ |
 | `numeric`  | ASCII digits only                                    | `{numeric}`  |
@@ -269,13 +285,16 @@ and validation scratch by walking the parent without cloning it. Inline requests
 borrow the Evidence value through CLI/MCP envelope validation; the application
 admits the entire portable envelope before authenticated parsing or hashing.
 The envelope clone, hash scratch and capture clone stay charged during grouping.
-Temporary URL/path/MIME work and retained state must
+Temporary URL/path/MIME/query work and retained state must
 fit together within a 96 MiB projection representation budget. A whole group or
 merge is admitted before allocating its retained sets, map entries or array
 values. Fixed group storage, distinct text, exclusions and limitations count as
 well as matched segments. Per-record reservations are released after each
 record; repeated values do not accumulate another retained charge. Only the
 requested page is materialized as output groups.
+Opt-in query projection reserves parser tuples and the per-record name set
+before decoding, then admits distinct retained names and source ordinals before
+changing a group. Repeated names within a request retain that ordinal once.

 The representation budget uses conservative storage and worst-case JSON-escaping
 allowances; it is not an enforced V8 heap or process-memory limit. It is separate
@@ -9,6 +9,7 @@
    "src/application/RealHopperAssertions.ts",
    "scripts/verify/fridaBindingsCompatibility.ts",
    "scripts/verify/termux/browser.mjs",
+    "scripts/verify-real-cutter.mjs",
    "scripts/verify-real-ghidra-noreturn.mjs",
    "tests/processOwnershipRunner.ts"
  ],
@@ -17,7 +18,12 @@
    "entry": ["docs/.vitepress/config.ts"]
  },
  "ignoreBinaries": ["mkfifo"],
-  "ignoreDependencies": ["electron", "esbuild", "webpack"],
+  "ignoreDependencies": [
+    "@stryker-mutator/command-runner",
+    "electron",
+    "esbuild",
+    "webpack"
+  ],
  "ignoreExportsUsedInFile": true,
  "rules": {
    "files": "error",
@@ -1,12 +1,12 @@
 REA third-party npm runtime copyright, license and NOTICE texts
 Collected from integrity-verified locked npm archives. Release-specific supplements and upstream notice gaps are identified in their entries.
 External engines are separately obtained; see THIRD_PARTY_NOTICES.md.
-Runtime dependency fingerprint (SHA-256): 63d5b55e027a0cfb3e77989ca337aca905f1652d10cfbfc4d4dd876d3ecd4f3a
+Runtime dependency fingerprint (SHA-256): b0d0151958456c6e72c5720af80d783893be88a7b4c9a409db8505296fe10bd9

-===== @babel/helper-string-parser@8.0.0 =====
+===== @babel/helper-string-parser@8.0.6 =====
 Declared license: MIT
-Archive: https://registry.npmjs.org/@babel/helper-string-parser/-/helper-string-parser-8.0.0.tgz
-Integrity: sha512-6mJgmFFFIIO82vvoLt9XtRC7/TkzXfts1t/SpRX4IHSzMgqoPYCWesVu1udUPUWioAE/2fcG6WuI8zrkE1gwrg==
+Archive: https://registry.npmjs.org/@babel/helper-string-parser/-/helper-string-parser-8.0.6.tgz
+Integrity: sha512-+6HYEBnsDRZHJtUTAEfMQd0zMDqdFHS7C+5vAsEZqBUr3vEMWOWlHJ0Jz1gS8jJIlCvLsuA6ivmeoqzH936YcA==
 Scope: npm runtime dependency, including optional platform packages.

 --- package/LICENSE ---
@@ -34,11 +34,10 @@ LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN ACTION
 OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN CONNECTION
 WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE SOFTWARE.

-
-===== @babel/helper-validator-identifier@8.0.4 =====
+===== @babel/helper-validator-identifier@8.0.6 =====
 Declared license: MIT
-Archive: https://registry.npmjs.org/@babel/helper-validator-identifier/-/helper-validator-identifier-8.0.4.tgz
-Integrity: sha512-4wFaiLd0bVo4cIoTXI3zKI038NIWE/cr3jvBjejOVYVxV/m8Ltav1USiGzG1fmS5J2RhgEOgXNNK46cRPnRsrg==
+Archive: https://registry.npmjs.org/@babel/helper-validator-identifier/-/helper-validator-identifier-8.0.6.tgz
+Integrity: sha512-HYx0YmuchRxRSpnjvcNffnPS8DCauKNmnumIuHG3jnQAVewRA7kfJGwF13ayEpSZs7s4R/oBH4eppuXcAQd5JQ==
 Scope: npm runtime dependency, including optional platform packages.

 --- package/LICENSE ---
@@ -66,7 +65,6 @@ LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN ACTION
 OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN CONNECTION
 WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE SOFTWARE.

-
 ===== @babel/parser@8.0.4 =====
 Declared license: MIT
 Archive: https://registry.npmjs.org/@babel/parser/-/parser-8.0.4.tgz
@@ -8443,5 +8441,3 @@ HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY,
 WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
 FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
 OTHER DEALINGS IN THE SOFTWARE.
-
-
@@ -154,7 +154,9 @@ node scripts/verify-windows-native-inspection.mjs
 This artifact injects one access-denied disposition result while using a real
 NTFS directory. The verifier checks retained ownership, recovery on a later
 query, cached successful admission, and cleanup on environment teardown. It
-does not reproduce or identify the host-dependent cause reported in #1701.
+renames the retained probe and recreates its old path, checking that cleanup
+removes the owned inode while preserving the replacement directory.
+It does not reproduce or identify the host-dependent cause reported in #1701.
 The fixture is stored under `.cache/` and is excluded from the shipped addon.

 The native lane checks DACL readback independently, source write/replacement
@@ -123,7 +123,8 @@ napi_value wrap(napi_env env, std::unique_ptr<Resource> resource);
 Resource& resource(napi_env env, napi_value value, Kind kind);

 std::unique_ptr<File> openFile(const std::wstring& path, DWORD access = GENERIC_READ,
-                             bool directory = false);
+                             bool directory = false,
+                             DWORD finalSharing = FILE_SHARE_READ);
 napi_value identity(napi_env env, HANDLE handle, const std::wstring& path);
 std::unique_ptr<Runtime> createRuntime(const std::wstring& parent, const std::wstring& prefix);
 void closeRuntime(Runtime& runtime);
@@ -61,7 +61,8 @@ static Handle openComponent(const std::wstring& path, DWORD access, bool directo
  return handle;
 }

-std::unique_ptr<File> openFile(const std::wstring& requested, DWORD access, bool directory) {
+std::unique_ptr<File> openFile(const std::wstring& requested, DWORD access, bool directory,
+                               DWORD finalSharing) {
  auto result = std::make_unique<File>(localPath(requested));
  result->requestedPath = requested;
  const auto& path = result->path;
@@ -77,7 +78,8 @@ std::unique_ptr<File> openFile(const std::wstring& requested, DWORD access, bool
  while (offset < path.size()) {
    auto end = path.find(L'\\', offset);
    const bool final = end == std::wstring::npos;
-    auto child = openComponent(path.substr(0, end), final ? access : 0, !final || directory);
+    auto child = openComponent(path.substr(0, end), final ? access : 0, !final || directory,
+                               final ? finalSharing : FILE_SHARE_READ);
    // Once the next component is pinned against deletion, its parent cannot
    // become empty and NTFS cannot convert that parent into a reparse point.
    // Relax only ancestor write sharing so unrelated child renames can proceed.
@@ -238,6 +240,100 @@ static void verifyPrivate(HANDLE handle, const std::wstring& path, bool root) {
  require(user && system, "Private DACL must allow only current user and SYSTEM", path, ERROR_ACCESS_DENIED);
 }

+static std::unique_ptr<File> openPrivateBridgeDirectory(const std::wstring& path) {
+  // The retained runtime root handle carries DELETE authority for owned
+  // cleanup. Windows checks sharing in both directions, so this read-only
+  // inspection handle must share DELETE even though it requests no such right.
+  auto directory = openFile(path, FILE_LIST_DIRECTORY, true,
+                            FILE_SHARE_READ | FILE_SHARE_DELETE);
+  verifyPrivate(directory->get(), path, true);
+  return directory;
+}
+
+static Handle openPrivateBridgeDescriptor(HANDLE parent, const std::wstring& name,
+                                          const std::wstring& path) {
+  require(!name.empty() && name.find_first_of(L"\\/:\"<>|?*") == std::wstring::npos &&
+          name != L"." && name != L".." && name.size() <= 255,
+          "Private descriptor requires one filename component", path, ERROR_INVALID_PARAMETER);
+  UNICODE_STRING coordinate{};
+  coordinate.Buffer = const_cast<wchar_t*>(name.data());
+  coordinate.Length = static_cast<USHORT>(name.size() * sizeof(wchar_t));
+  coordinate.MaximumLength = coordinate.Length;
+  OBJECT_ATTRIBUTES attributes;
+  InitializeObjectAttributes(&attributes, &coordinate, OBJ_CASE_INSENSITIVE, parent, nullptr);
+  IO_STATUS_BLOCK observation{};
+  HANDLE raw = INVALID_HANDLE_VALUE;
+  // Immutable runtime files retain their original writer handle so later
+  // opens cannot acquire write or delete access. This read-only handle must
+  // share those granted rights for Windows' symmetric share check to pass.
+  const auto status = NtCreateFile(&raw, GENERIC_READ | READ_CONTROL | FILE_READ_ATTRIBUTES | SYNCHRONIZE,
+                                   &attributes, &observation, nullptr, FILE_ATTRIBUTE_NORMAL,
+                                   FILE_SHARE_READ | FILE_SHARE_WRITE | FILE_SHARE_DELETE, FILE_OPEN,
+                                   FILE_SYNCHRONOUS_IO_NONALERT | FILE_OPEN_REPARSE_POINT |
+                                       FILE_NON_DIRECTORY_FILE,
+                                   nullptr, 0);
+  if (status < 0)
+    throw Failure("Open Cutter bridge descriptor relative to its verified directory failed", path,
+                  RtlNtStatusToDosError(status));
+  Handle descriptor(raw);
+  require(descriptor.valid() && observation.Information == FILE_OPENED,
+          "Cutter bridge descriptor was not opened as an existing file", path, ERROR_ACCESS_DENIED);
+  FILE_STANDARD_INFO standard{};
+  require(GetFileInformationByHandleEx(descriptor.get(), FileStandardInfo, &standard, sizeof(standard)),
+          "Read Cutter bridge descriptor metadata failed", path);
+  require(!standard.Directory && standard.EndOfFile.QuadPart >= 0 &&
+              standard.EndOfFile.QuadPart <= 64 * 1024,
+          "Cutter bridge descriptor exceeds its byte limit or is not a file", path, ERROR_FILE_TOO_LARGE);
+  FILE_ATTRIBUTE_TAG_INFO fileAttributes{};
+  require(GetFileInformationByHandleEx(descriptor.get(), FileAttributeTagInfo, &fileAttributes, sizeof(fileAttributes)),
+          "Read Cutter bridge descriptor attributes failed", path);
+  require((fileAttributes.FileAttributes & FILE_ATTRIBUTE_REPARSE_POINT) == 0,
+          "Cutter bridge descriptor cannot be a reparse point", path, ERROR_REPARSE_TAG_INVALID);
+  verifyPrivate(descriptor.get(), path, false);
+  return descriptor;
+}
+
+static napi_value verifyCutterBridgeDirectory(napi_env env, const std::vector<napi_value>& args) {
+  require(args.size() == 1, "Cutter bridge directory verification requires one path", L"",
+          ERROR_INVALID_PARAMETER);
+  const auto path = wide(env, args[0]);
+  const auto directory = openPrivateBridgeDirectory(path);
+  auto result = object(env);
+  set(env, result, "privateDacl", boolean(env, true));
+  set(env, result, "identity", identity(env, directory->get(), path));
+  return result;
+}
+
+static napi_value readCutterBridgeDescriptor(napi_env env, const std::vector<napi_value>& args) {
+  require(args.size() == 3, "Cutter bridge descriptor read requires directory, name, and byte limit", L"",
+          ERROR_INVALID_PARAMETER);
+  const auto path = wide(env, args[0]);
+  const auto name = wide(env, args[1]);
+  const double maximum = numeric(env, args[2]);
+  require(maximum >= 1 && maximum <= 64 * 1024 && maximum == static_cast<size_t>(maximum),
+          "Cutter bridge descriptor limit is outside the supported range", path, ERROR_INVALID_PARAMETER);
+  auto directory = openPrivateBridgeDirectory(path);
+  auto descriptor = openPrivateBridgeDescriptor(directory->get(), name, path + L"\\" + name);
+  FILE_STANDARD_INFO standard{};
+  require(GetFileInformationByHandleEx(descriptor.get(), FileStandardInfo, &standard, sizeof(standard)),
+          "Read Cutter bridge descriptor size failed", path);
+  require(standard.EndOfFile.QuadPart <= maximum,
+          "Cutter bridge descriptor exceeds the requested byte limit", path, ERROR_FILE_TOO_LARGE);
+  std::vector<unsigned char> bytes(static_cast<size_t>(standard.EndOfFile.QuadPart));
+  size_t offset = 0;
+  while (offset < bytes.size()) {
+    DWORD count = 0;
+    const DWORD requested = static_cast<DWORD>(std::min<size_t>(bytes.size() - offset, MAXDWORD));
+    require(ReadFile(descriptor.get(), bytes.data() + offset, requested, &count, nullptr) && count > 0,
+            "Read Cutter bridge descriptor through its verified handle failed", path);
+    offset += count;
+  }
+  napi_value result;
+  const void* source = bytes.empty() ? static_cast<const void*>("") : bytes.data();
+  check(napi_create_buffer_copy(env, bytes.size(), source, nullptr, &result));
+  return result;
+}
+
 std::unique_ptr<Runtime> createRuntime(const std::wstring& parentPath, const std::wstring& prefix) {
  std::unique_ptr<File> parent;
  try { parent = openFile(parentPath, 0, true); }
@@ -559,6 +655,8 @@ static napi_value snapshot(napi_env env, napi_value rootValue, Runtime& root,
 }

 napi_value filesystemCall(napi_env env, const std::wstring& operation, const std::vector<napi_value>& args) {
+  if (operation == L"cutter_bridge_verify_directory") return verifyCutterBridgeDirectory(env, args);
+  if (operation == L"cutter_bridge_read_descriptor") return readCutterBridgeDescriptor(env, args);
  const auto expected = operation == L"open" || operation == L"close" || operation == L"runtime_close" || operation == L"runtime_snapshot_cancel" ? 1
                      : operation == L"runtime_create" || operation == L"runtime_mkdir" || operation == L"runtime_open" ? 2 : 3;
  require(args.size() == static_cast<size_t>(expected), "Wrong native filesystem argument count", L"", ERROR_INVALID_PARAMETER);
@@ -464,6 +464,7 @@ napi_value inspect(napi_env env, std::unique_ptr<Runtime>& probe) {
  set(env, result, "architecture", string(env, std::string("x64")));
  set(env, result, "filesystem", filesystem);
  set(env, result, "privateDacl", boolean(env, true));
+  set(env, result, "privateBridgeDescriptorAccess", boolean(env, true));
  set(env, result, "atomicJobAssignment", boolean(env, true));
  set(env, result, "killOnOwnerClose", boolean(env, true));
  set(env, result, "protocolStdin", boolean(env, true));
@@ -21,10 +21,14 @@
    "mcp-server",
    "model-context-protocol",
    "static-analysis",
-    "disassembler",
+    "rea",
    "decompiler",
    "mcp",
-    "hopper"
+    "hopper",
+    "electron",
+    "javascript-analysis",
+    "ghidra",
+    "ida"
  ],
  "publishConfig": {
    "access": "public"
@@ -45,6 +49,7 @@
    "dist",
    "install.sh",
    "bridge/hopper_bridge.py",
+    "bridge/cutter/rea_bridge.py",
    "bridge/native/NativeUIChildren.swift",
    "bridge/native/ReaNativeUI.swift",
    "bridge/native/rea_lldb_tracer.py",
@@ -113,7 +118,9 @@
    "precheck:pr": "npm run deps:check",
    "check:pr": "node scripts/run-exclusive.mjs artifacts turbo run typecheck lint:built format:check knip metadata:check test:run && npm run test:python && npm run docs:check && npm run jscpd && npm run scan:todos",
    "precheck:ci": "npm run deps:check",
-    "check:ci": "node scripts/run-exclusive.mjs artifacts turbo run typecheck lint:built format:check knip metadata:check jscpd scan:todos && npm run verify:architecture-guards",
+    "check:ci": "node scripts/run-exclusive.mjs artifacts npm run check:ci:tasks && npm run verify:architecture-guards",
+    "check:ci:tasks": "turbo run typecheck lint:built format:check knip metadata:check jscpd scan:todos",
+    "ci:static-context": "node scripts/ci/static-task-context.mjs",
    "build:cached": "node scripts/check-dependency-install.mjs && node scripts/run-exclusive.mjs artifacts turbo run artifacts:skills",
    "build:termux": "node scripts/run-exclusive.mjs artifacts npm run build:termux:run",
    "build:termux:run": "npm run compile:termux && npm run artifacts:skills",
@@ -133,7 +140,8 @@
    "test:run": "vitest run",
    "test:python": "python3 -m unittest bridge/native/test_rea_lldb_tracer.py -v && python3 -m unittest discover -s experiments/analysis-efficiency -p 'test_*.py' -v",
    "test:local": "node scripts/run-development-tests.mjs local",
-    "test:fast": "vitest run --project domain --project services --project adapters --project composition --project conformance --project evaluation",
+    "test:fast": "node scripts/ci/test-group.mjs source",
+    "test:built": "node scripts/ci/test-group.mjs built",
    "pretest:boundary": "npm run test:prepare",
    "test:boundary": "vitest run --project boundary --project cli-boundary --project process-boundary --project mcp-boundary --project process-global",
    "pretest:mcp": "node scripts/check-dependency-install.mjs && node scripts/run-exclusive.mjs artifacts turbo run artifacts:skills artifacts:mcp-catalog",
@@ -145,6 +153,10 @@
    "test:integration": "vitest run --project boundary --project cli-boundary --project process-boundary --project mcp-boundary --project process-global --project acceptance",
    "pretest:coverage": "npm run test:prepare",
    "test:coverage": "vitest run --coverage.enabled --coverage.provider=v8",
+    "test:coverage:reverse-tools": "vitest run --config vitest.reverse-tools.config.ts --coverage.enabled",
+    "pretest:coverage:reverse-tools": "npm run test:prepare",
+    "test:mutation:reverse-tools": "node scripts/prepare-reverse-tools-mutation.mjs && stryker run stryker.config.mjs && stryker run stryker.semantics.config.mjs",
+    "pretest:mutation:reverse-tools": "npm run test:prepare",
    "test:ci": "npm run test:coverage -- --reporter=default --reporter=junit --outputFile.junit=test-results.xml",
    "pretest:ci:shard": "npm run test:prepare",
    "test:ci:shard": "npm run test:ci:shard:run",
@@ -174,6 +186,7 @@
    "verify:ghidra": "npm run build:cached && node scripts/verify-real-ghidra.mjs",
    "verify:ghidra:seeds": "npm run build:cached && node scripts/verify/ghidra/seeds.mjs",
    "verify:ghidra:mips": "npm run build:cached && node scripts/verify-real-ghidra-mips.mjs",
+    "verify:ghidra:ppc": "npm run build:cached && node scripts/verify-real-ghidra-ppc.mjs",
    "verify:ghidra:riscv": "npm run build:cached && node scripts/verify-real-ghidra-riscv.mjs",
    "verify:ghidra:6502": "npm run build:cached && node scripts/verify-real-ghidra-6502.mjs",
    "verify:ghidra:mips-shared": "npm run build:cached && node scripts/verify-real-ghidra-mips-shared.mjs",
@@ -253,6 +266,7 @@
    "build:windows-native": "node scripts/build-windows-native.mjs",
    "verify:windows-native:artifact": "node scripts/verify-windows-native-artifact.mjs",
    "prepublishOnly": "npm run verify:windows-native:artifact",
+    "verify:cutter": "npm run build:cached && node scripts/verify-real-cutter.mjs",
    "verify:windows-native": "node scripts/verify-windows-native.mjs",
    "verify:ghidra:windows:package": "node scripts/verify-packaged-ghidra-windows.mjs",
    "build:ghidra:nativeaot": "node scripts/build-ghidra-nativeaot.mjs",
@@ -344,6 +358,9 @@
  },
  "devDependencies": {
    "@fast-check/vitest": "0.5.0",
+    "@stryker-mutator/core": "10.0.0",
+    "@stryker-mutator/typescript-checker": "10.0.0",
+    "@stryker-mutator/vitest-runner": "10.0.0",
    "@types/cross-spawn": "6.0.6",
    "@types/node": "26.6.4",
    "@types/semver": "7.7.1",
@@ -360,6 +377,7 @@
    "turbo": "2.11.7",
    "typescript": "7.0.2",
    "typescript-termux": "npm:typescript@5.9.3",
+    "typescript6": "npm:typescript@6.0.3",
    "vitepress": "1.6.4",
    "vitest": "5.0.3"
  },
@@ -0,0 +1,67 @@
+import type { NormalizedWorkflowRun } from "./workflow-run.mjs";
+import type { StaticTaskContext } from "./static-task-context.mjs";
+export type { StaticTaskContext } from "./static-task-context.mjs";
+
+export interface StaticDiagnostic {
+  readonly kind: "lint" | "jscpd";
+  readonly path?: string;
+  readonly paths?: readonly string[];
+  readonly code?: string;
+  readonly message?: string;
+  readonly severity?: string;
+}
+
+export function classifyFailure(input: {
+  repository: string;
+  defaultBranch: string;
+  run: Record<string, unknown> & NormalizedWorkflowRun;
+  job: Record<string, unknown>;
+  diagnostics: readonly StaticDiagnostic[];
+  context: StaticTaskContext;
+  mainRun: Record<string, unknown> & NormalizedWorkflowRun;
+  mainJob: Record<string, unknown>;
+  mainDiagnostics: readonly StaticDiagnostic[];
+  mainContext: StaticTaskContext;
+  executionSourcesVerified: boolean;
+  ancestry: { status: string; mergeBaseSha: string | null };
+  blobs: Readonly<Record<string, boolean>>;
+}): {
+  classification: "matching-main-diagnostics" | "unknown";
+  reason: string;
+};
+
+export function diagnosticPaths(item: StaticDiagnostic): readonly string[];
+export function isFailedConclusion(
+  value: unknown,
+): value is "failure" | "timed_out";
+export function isEligibleMainRun(
+  run: Record<string, unknown>,
+  defaultBranch: string,
+  repository: string,
+): boolean;
+export function pullRequestBaseSha(
+  run: Record<string, unknown>,
+  repository: string,
+): string | null;
+export function safeSha(value: unknown): string | null;
+export function validRepoPath(value: unknown): value is string;
+export function selectStaticCheckArtifact(
+  page: unknown,
+  run: { id?: string | number; head_sha?: string },
+  runAttempt: number,
+): unknown | null;
+
+export function readStaticCheckArtifact(options: {
+  repository: string;
+  artifactId: number | string;
+  token: string;
+  fetchImpl?: typeof fetch;
+}): Promise<{
+  context: StaticTaskContext;
+  diagnostics: readonly StaticDiagnostic[];
+}>;
+
+export function parseStaticArchive(archivePath: string): Promise<{
+  context: StaticTaskContext;
+  diagnostics: readonly StaticDiagnostic[];
+}>;
@@ -0,0 +1,442 @@
+import { execFile } from "node:child_process";
+import { mkdtemp, rm, writeFile } from "node:fs/promises";
+import { tmpdir } from "node:os";
+import { isAbsolute, join } from "node:path";
+import { promisify } from "node:util";
+
+import { parseStaticTaskContext } from "./static-task-context.mjs";
+
+const execute = promisify(execFile);
+const MAX_ARCHIVE_BYTES = 32 * 1024 * 1024;
+const MAX_MEMBER_BYTES = 4 * 1024 * 1024;
+const MAX_DIAGNOSTICS = 10_000;
+const API_ROOT = "https://api.github.com/repos";
+const SHA = /^[0-9a-f]{40}$/iu;
+const TASK_HASH = /^[a-f0-9]{16}$/u;
+
+/** Compare parsed static diagnostics only after source, task, host, and ancestry proof. */
+export function classifyFailure({
+  repository,
+  defaultBranch,
+  run,
+  job,
+  diagnostics,
+  context,
+  mainRun,
+  mainJob,
+  mainDiagnostics,
+  mainContext,
+  executionSourcesVerified,
+  ancestry,
+  blobs,
+}) {
+  const failedGate = firstFailure([
+    () => [
+      !isCompletedFailure(run) ||
+        job?.status !== "completed" ||
+        !isFailedConclusion(job?.conclusion),
+      "run_or_job_not_a_completed_failure",
+    ],
+    () => [
+      !isEligiblePullRun(run, repository) ||
+        !isEligibleMainRun(mainRun, defaultBranch, repository),
+      "run_event_or_repository_identity_unverified",
+    ],
+    () => [
+      job?.name !== mainJob?.name || job?.name !== "Static checks",
+      "failed_job_not_the_same_static_gate",
+    ],
+    () => [
+      !Array.isArray(diagnostics) ||
+        diagnostics.length === 0 ||
+        !Array.isArray(mainDiagnostics),
+      "structured_diagnostics_missing_or_invalid",
+    ],
+    () => [
+      !validContext(context) || !validContext(mainContext),
+      "static_task_context_missing_or_invalid",
+    ],
+    () => [
+      executionSourcesVerified !== true,
+      "checkout_sources_not_verified_against_api_commits",
+    ],
+    () => [
+      !sameExecutionContext(context, mainContext),
+      "task_or_toolchain_identity_differs",
+    ],
+    () => [
+      ancestry?.status !== "ahead" && ancestry?.status !== "identical",
+      "main_head_not_proven_ancestor_of_tested_base",
+    ],
+    () => [
+      ancestry?.mergeBaseSha !== mainRun?.head_sha,
+      "main_head_is_not_the_verified_merge_base",
+    ],
+    () => [
+      !sameTaskHashes(diagnostics, context, mainContext),
+      "owning_task_hash_differs",
+    ],
+  ]);
+  if (failedGate) return outcome("unknown", failedGate.reason);
+  const keys = diagnostics.map(diagnosticKey);
+  const mainKeys = new Set(mainDiagnostics.map(diagnosticKey).filter(Boolean));
+  if (keys.some((key) => key === null))
+    return outcome("unknown", "diagnostic_shape_not_supported");
+  if (!keys.every((key) => mainKeys.has(key)))
+    return outcome(
+      "unknown",
+      "main_run_does_not_contain_the_same_structured_diagnostic",
+    );
+  const blobStatus = blobIdentityStatus(diagnostics, blobs);
+  if (blobStatus !== "identical")
+    return outcome(
+      "unknown",
+      blobStatus === "different"
+        ? "diagnostic_source_content_not_identical_at_main_and_tested_base"
+        : "diagnostic_source_blob_identity_unavailable",
+    );
+  return outcome(
+    "matching-main-diagnostics",
+    "supported_diagnostics_match_failed_task_and_source_on_verified_main_ancestor",
+  );
+}
+
+export function diagnosticPaths(item) {
+  return item.kind === "jscpd" ? item.paths : [item.path];
+}
+export function isFailedConclusion(value) {
+  return value === "failure" || value === "timed_out";
+}
+export function isEligibleMainRun(run, defaultBranch, repository) {
+  return (
+    isEligibleStaticRun(run) &&
+    run.repository.full_name === repository &&
+    run.head_branch === defaultBranch &&
+    ["push", "schedule", "workflow_dispatch"].includes(run.event)
+  );
+}
+export function pullRequestBaseSha(run, repository) {
+  if (
+    run?.event !== "pull_request" ||
+    run.repository?.full_name !== repository ||
+    !Array.isArray(run.pull_requests) ||
+    run.pull_requests.length !== 1
+  )
+    return null;
+  const base = run.pull_requests[0].base;
+  return base?.repo?.full_name === repository && SHA.test(base?.sha ?? "")
+    ? base.sha
+    : null;
+}
+
+export function safeSha(value) {
+  return typeof value === "string" && SHA.test(value)
+    ? value.toLowerCase()
+    : null;
+}
+export function selectStaticCheckArtifact(page, run, runAttempt) {
+  if (runAttempt !== 1) return null;
+  if (
+    !Array.isArray(page?.artifacts) ||
+    !Number.isSafeInteger(page.total_count) ||
+    page.total_count !== page.artifacts.length ||
+    page.artifacts.length >= 100
+  )
+    return null;
+  const matches = page.artifacts.filter(
+    (artifact) => artifact.name === "static-check-reports",
+  );
+  if (matches.length !== 1) return null;
+  const [artifact] = matches;
+  if (
+    artifact.expired !== false ||
+    String(artifact.workflow_run?.id) !== String(run.id) ||
+    artifact.workflow_run?.head_sha !== run.head_sha
+  )
+    return null;
+  return artifact;
+}
+export function validRepoPath(value) {
+  return (
+    typeof value === "string" &&
+    value.length > 0 &&
+    !isAbsolute(value) &&
+    !value.includes("\\") &&
+    value
+      .split("/")
+      .every((part) => part !== "" && part !== "." && part !== "..") &&
+    !value.startsWith("/")
+  );
+}
+
+/** Download one named static report archive and parse only whitelisted JSON members. */
+export async function readStaticCheckArtifact({
+  repository,
+  artifactId,
+  token,
+  fetchImpl = fetch,
+}) {
+  if (!/^[A-Za-z0-9_.-]+\/[A-Za-z0-9_.-]+$/u.test(repository ?? ""))
+    throw new Error("artifact repository identity is invalid");
+  if (!token) throw new Error("artifact download requires a GitHub token");
+  const url = `${API_ROOT}/${repository}/actions/artifacts/${positiveId(artifactId)}/zip`;
+  let response = await fetchImpl(url, {
+    headers: {
+      Accept: "application/vnd.github+json",
+      Authorization: `Bearer ${token}`,
+      "X-GitHub-Api-Version": "2022-11-28",
+    },
+    signal: AbortSignal.timeout(10_000),
+    redirect: "manual",
+  });
+  if ([301, 302, 303, 307, 308].includes(response.status)) {
+    const location = response.headers.get("location");
+    if (!location || new URL(location).protocol !== "https:")
+      throw new Error("static_artifact_redirect_invalid");
+    response = await fetchImpl(location, {
+      signal: AbortSignal.timeout(10_000),
+      redirect: "error",
+    });
+  }
+  if (!response.ok) throw new Error(`github_http_${response.status}`);
+  const bytes = await readLimitedBytes(response, MAX_ARCHIVE_BYTES);
+  const directory = await mkdtemp(join(tmpdir(), "rea-ci-static-artifact-"));
+  const archive = join(directory, "artifact.zip");
+  try {
+    await writeFile(archive, bytes, { flag: "wx" });
+    return await parseStaticArchive(archive);
+  } finally {
+    await rm(directory, { recursive: true, force: true });
+  }
+}
+
+/** Accept only the three named producer outputs; no archive paths are written to disk. */
+export async function parseStaticArchive(archivePath) {
+  const { stdout: listing } = await execute("unzip", ["-Z1", archivePath], {
+    encoding: "utf8",
+    maxBuffer: 128 * 1024,
+    timeout: 5_000,
+  });
+  const allowed = new Set([
+    "oxlint.json",
+    "jscpd/jscpd-report.json",
+    "static-task-context.json",
+  ]);
+  const members = new Map();
+  for (const name of listing.split(/\r?\n/u).filter(Boolean)) {
+    const prefix = ".cache/static-checks/";
+    const member = name.startsWith(prefix) ? name.slice(prefix.length) : name;
+    if (!allowed.has(member)) continue;
+    if (members.has(member))
+      throw new Error("static artifact contains duplicate report members");
+    const { stdout } = await execute("unzip", ["-p", archivePath, name], {
+      encoding: "buffer",
+      maxBuffer: MAX_MEMBER_BYTES,
+      timeout: 5_000,
+    });
+    members.set(member, JSON.parse(stdout.toString("utf8")));
+  }
+  const context = parseStaticTaskContext(
+    members.get("static-task-context.json"),
+  );
+  const diagnostics = normalizeDiagnostics(
+    members.get("oxlint.json"),
+    members.get("jscpd/jscpd-report.json"),
+  );
+  if (diagnostics.length > MAX_DIAGNOSTICS)
+    throw new Error("static diagnostic count exceeds the reporting limit");
+  return { context, diagnostics };
+}
+
+function normalizeDiagnostics(lint, duplicate) {
+  return [...normalizeLint(lint), ...normalizeDuplicates(duplicate)];
+}
+
+function normalizeLint(report) {
+  if (report === undefined) return [];
+  if (!Array.isArray(report?.diagnostics))
+    throw new Error("oxlint diagnostics are malformed");
+  return report.diagnostics
+    .filter((item) => item.severity === "error")
+    .map((item) => {
+      if (
+        typeof item.filename !== "string" ||
+        !repoPath(item.filename) ||
+        typeof item.code !== "string" ||
+        typeof item.message !== "string"
+      )
+        throw new Error("oxlint error diagnostic is malformed");
+      return {
+        kind: "lint",
+        path: item.filename,
+        code: item.code,
+        message: item.message,
+        severity: "error",
+      };
+    });
+}
+
+function normalizeDuplicates(report) {
+  if (report === undefined) return [];
+  if (
+    !Array.isArray(report?.duplicates) ||
+    !Number.isFinite(report.statistics?.total?.percentage) ||
+    report.statistics.total.percentage < 0
+  )
+    throw new Error("jscpd report is malformed");
+  // The producer owns the checker outcome; do not copy or infer its threshold.
+  const exitCode = report.reaCheck?.exitCode;
+  if (exitCode === undefined || exitCode === 0) return [];
+  if (!Number.isSafeInteger(exitCode) || exitCode < 0)
+    throw new Error("jscpd checker outcome is malformed");
+  if (exitCode !== 1) return [];
+  return report.duplicates.map((item) => {
+    const paths = [item.firstFile?.name, item.secondFile?.name].sort();
+    if (paths.some((path) => !repoPath(path)))
+      throw new Error("jscpd duplicate paths are malformed");
+    return { kind: "jscpd", paths };
+  });
+}
+
+async function readLimitedBytes(response, limit) {
+  const reader = response.body?.getReader();
+  if (!reader) throw new Error("static artifact response has no body");
+  const chunks = [];
+  let size = 0;
+  try {
+    for (;;) {
+      const { done, value } = await reader.read();
+      if (done) break;
+      size += value.byteLength;
+      if (size > limit) {
+        await reader.cancel();
+        throw new Error("static artifact exceeds the download limit");
+      }
+      chunks.push(value);
+    }
+  } finally {
+    reader.releaseLock();
+  }
+  const result = new Uint8Array(size);
+  let offset = 0;
+  for (const chunk of chunks) {
+    result.set(chunk, offset);
+    offset += chunk.byteLength;
+  }
+  return result;
+}
+
+function repoPath(value) {
+  return (
+    typeof value === "string" &&
+    value.length > 0 &&
+    !value.startsWith("/") &&
+    !value.includes("\\") &&
+    value
+      .split("/")
+      .every((part) => part !== "" && part !== "." && part !== "..")
+  );
+}
+
+function positiveId(value) {
+  const id = Number(value);
+  if (!Number.isSafeInteger(id) || id < 1)
+    throw new Error("artifact ID must be a positive integer");
+  return String(id);
+}
+
+function diagnosticKey(item) {
+  if (
+    item?.kind === "lint" &&
+    validRepoPath(item.path) &&
+    typeof item.code === "string" &&
+    typeof item.message === "string" &&
+    item.severity === "error"
+  )
+    return JSON.stringify([
+      item.kind,
+      item.path,
+      item.code,
+      item.message,
+      item.severity,
+    ]);
+  if (
+    item?.kind === "jscpd" &&
+    Array.isArray(item.paths) &&
+    item.paths.length === 2 &&
+    item.paths.every(validRepoPath)
+  )
+    return JSON.stringify([item.kind, ...[...item.paths].sort()]);
+  return null;
+}
+function blobIdentityStatus(diagnostics, blobs) {
+  const paths = [...new Set(diagnostics.flatMap(diagnosticPaths))];
+  if (paths.some((path) => blobs?.[path] === false)) return "different";
+  return paths.length > 0 && paths.every((path) => blobs?.[path] === true)
+    ? "identical"
+    : "unavailable";
+}
+function sameTaskHashes(diagnostics, left, right) {
+  return diagnostics.every((diagnostic) => {
+    const task = diagnostic.kind === "jscpd" ? "jscpd" : "lint:built";
+    return (
+      TASK_HASH.test(left.taskHashes[task] ?? "") &&
+      left.taskHashes[task] === right.taskHashes[task]
+    );
+  });
+}
+function validContext(value) {
+  return (
+    value?.version === 1 &&
+    SHA.test(value.sourceSha ?? "") &&
+    value.sourceClean === true &&
+    typeof value.nodeVersion === "string" &&
+    typeof value.npmVersion === "string" &&
+    typeof value.turboVersion === "string" &&
+    typeof value.platform === "string" &&
+    typeof value.architecture === "string" &&
+    value.taskHashes &&
+    typeof value.taskHashes === "object" &&
+    !Array.isArray(value.taskHashes)
+  );
+}
+function sameExecutionContext(left, right) {
+  return [
+    "nodeVersion",
+    "npmVersion",
+    "turboVersion",
+    "platform",
+    "architecture",
+  ].every((key) => left[key] === right[key]);
+}
+function isEligiblePullRun(run, repository) {
+  return (
+    isEligibleStaticRun(run) && pullRequestBaseSha(run, repository) !== null
+  );
+}
+function isEligibleStaticRun(run) {
+  return (
+    isRecord(run) &&
+    run.status === "completed" &&
+    isFailedConclusion(run.conclusion) &&
+    SHA.test(run.head_sha ?? "") &&
+    typeof run.repository?.full_name === "string" &&
+    run.workflowPath?.file === ".github/workflows/ci.yml"
+  );
+}
+function isCompletedFailure(run) {
+  return run?.status === "completed" && isFailedConclusion(run.conclusion);
+}
+function isRecord(value) {
+  return typeof value === "object" && value !== null && !Array.isArray(value);
+}
+function firstFailure(checks) {
+  for (const check of checks) {
+    const [failed, reason] = check();
+    if (failed) return { reason };
+  }
+  return null;
+}
+function outcome(classification, reason) {
+  return { classification, reason };
+}
@@ -0,0 +1,159 @@
+import type { NormalizedWorkflowRun } from "./workflow-run.mjs";
+import type { StaticDiagnostic } from "./artifacts.mjs";
+export type { StaticDiagnostic } from "./artifacts.mjs";
+import type { StaticTaskContext } from "./static-task-context.mjs";
+export type { StaticTaskContext } from "./static-task-context.mjs";
+export interface CiRun {
+  readonly id?: string | number;
+  readonly run_attempt?: number;
+  readonly status?: string;
+  readonly conclusion?: string | null;
+  readonly event?: string;
+  readonly path?: string;
+  readonly head_sha?: string;
+  readonly head_branch?: string;
+  readonly repository?: { readonly full_name?: string };
+  readonly pull_requests?: readonly {
+    readonly base?: {
+      readonly sha?: string;
+      readonly repo?: { readonly full_name?: string };
+    };
+  }[];
+}
+
+export interface CiJob {
+  readonly name?: string;
+  readonly status?: string;
+  readonly conclusion?: string | null;
+  readonly run_id?: number;
+  readonly run_attempt?: number;
+}
+
+export interface StaticEvidence {
+  readonly context: StaticTaskContext;
+  readonly diagnostics: readonly StaticDiagnostic[];
+}
+
+export interface CiFailureReport {
+  readonly version: 1;
+  readonly observedAt: string;
+  readonly repository: string;
+  readonly defaultBranch: string;
+  readonly informationalOnly: true;
+  readonly gateEffect: string;
+  readonly boundedInput: {
+    readonly candidateRuns: number;
+    readonly limit: number;
+    readonly mainCandidateLimit: number;
+    readonly mainCandidateRuns: number;
+    readonly mainRunLookupLimit: number;
+    readonly manifestTruncated: boolean;
+  };
+  readonly counts: Readonly<
+    Record<"matching-main-diagnostics" | "unknown", number>
+  >;
+  readonly failures: readonly {
+    readonly runId: number | string;
+    readonly runAttempt: number;
+    readonly classification: "matching-main-diagnostics" | "unknown";
+    readonly reason: string;
+    readonly candidateFailures?: readonly {
+      readonly runId: number | string;
+      readonly runAttempt: number;
+      readonly reason: string;
+    }[];
+    readonly runHead?: string | null;
+    readonly event?: string | null;
+    readonly job?: string;
+    readonly conclusion?: string | null;
+  }[];
+}
+
+export interface FailureReportApi {
+  getRun(repository: string, id: string | number): Promise<CiRun>;
+  getJobs(
+    repository: string,
+    id: string | number,
+    attempt: number,
+  ): Promise<readonly CiJob[]>;
+  getStaticEvidence(
+    repository: string,
+    run: CiRun,
+    job: CiJob,
+  ): Promise<StaticEvidence | null>;
+  getAncestry(
+    repository: string,
+    ancestor: string,
+    descendant: string,
+  ): Promise<{ status: string; mergeBaseSha: string | null }>;
+  verifyExecutionSource(
+    repository: string,
+    run: CiRun,
+    context: StaticTaskContext,
+  ): Promise<boolean>;
+  getBlobIdentity(
+    repository: string,
+    ancestor: string,
+    descendant: string,
+    paths: readonly string[],
+  ): Promise<Readonly<Record<string, boolean>>>;
+}
+
+export function classifyFailure(input: {
+  repository: string;
+  defaultBranch: string;
+  run: CiRun & NormalizedWorkflowRun;
+  job: CiJob;
+  diagnostics: readonly StaticDiagnostic[];
+  context: StaticTaskContext;
+  mainRun: CiRun & NormalizedWorkflowRun;
+  mainJob: CiJob;
+  mainDiagnostics: readonly StaticDiagnostic[];
+  mainContext: StaticTaskContext;
+  executionSourcesVerified: boolean;
+  ancestry: { status: string; mergeBaseSha: string | null };
+  blobs: Readonly<Record<string, boolean>>;
+}): {
+  classification: "matching-main-diagnostics" | "unknown";
+  reason: string;
+};
+
+export function reportCiFailures(options: {
+  repository: string;
+  defaultBranch: string;
+  batchManifest: {
+    readonly repository: string;
+    readonly runs: readonly {
+      readonly id: number | string;
+      readonly runAttempt: number;
+      readonly workflow: string;
+      readonly event?: string | null;
+      readonly headBranch?: string | null;
+      readonly conclusion: string | null;
+      readonly status: string;
+    }[];
+    readonly truncation?: readonly unknown[];
+    readonly coverage?: { readonly exhaustive?: boolean };
+    readonly selection?: { readonly exhaustive?: boolean };
+  };
+  api: FailureReportApi;
+  outputDirectory?: string;
+  summaryPath?: string;
+  observedAt?: string;
+}): Promise<CiFailureReport>;
+
+export function createGitHubApi(options?: {
+  token?: string;
+  fetchImpl?: typeof fetch;
+  apiRequest?: (
+    method: string,
+    endpoint: string,
+    budget: { requests: number; startedAt: number; limitMs: number },
+  ) => Promise<unknown>;
+}): FailureReportApi;
+
+export function writeFailureReport(
+  report: CiFailureReport,
+  outputDirectory?: string,
+  summaryPath?: string,
+): Promise<{ jsonPath: string; markdownPath: string }>;
@@ -0,0 +1,533 @@
+import { appendFile, mkdir, readFile, writeFile } from "node:fs/promises";
+import { join, resolve } from "node:path";
+import { fileURLToPath } from "node:url";
+
+import {
+  classifyFailure,
+  diagnosticPaths,
+  isEligibleMainRun,
+  isFailedConclusion,
+  pullRequestBaseSha,
+  readStaticCheckArtifact,
+  safeSha,
+  selectStaticCheckArtifact,
+  validRepoPath,
+} from "./artifacts.mjs";
+import { createGhApiRequest } from "./github-api.mjs";
+import { normalizeWorkflowRun } from "./workflow-run.mjs";
+
+const OUTPUT_DIRECTORY = join(".cache", "ci-failures");
+const BATCH_DIRECTORY = join(".cache", "ci-job-timings");
+const MAX_JSON_BYTES = 8 * 1024 * 1024;
+const MAX_FAILED_RUNS = 20;
+const MAX_MAIN_CANDIDATES = 5;
+const MAX_MAIN_RUN_LOOKUPS = 100;
+const MAX_PATHS_PER_COMPARISON = 50;
+const MAX_API_REQUESTS = 250;
+
+export { classifyFailure } from "./artifacts.mjs";
+
+/** Uses only failed CI candidates present in the supplied bounded timing manifest. */
+export async function reportCiFailures({
+  repository,
+  defaultBranch,
+  batchManifest,
+  api,
+  outputDirectory = OUTPUT_DIRECTORY,
+  summaryPath,
+  observedAt = new Date().toISOString(),
+}) {
+  validateRepository(repository);
+  if (typeof defaultBranch !== "string" || defaultBranch.length === 0)
+    throw new Error("defaultBranch is required");
+  if (
+    !batchManifest ||
+    batchManifest.repository !== repository ||
+    !Array.isArray(batchManifest.runs)
+  )
+    throw new Error("timing manifest repository or runs are invalid");
+
+  const failedRuns = batchManifest.runs.filter(
+    (entry) =>
+      entry.workflow === "CI" &&
+      entry.status === "collected" &&
+      isFailedConclusion(entry.conclusion),
+  );
+  const prRuns = failedRuns.filter((entry) => entry.event === "pull_request");
+  const candidates = prRuns.slice(0, MAX_FAILED_RUNS);
+  const mainEntries = failedRuns.filter(
+    (entry) =>
+      entry.headBranch === defaultBranch &&
+      ["push", "schedule", "workflow_dispatch"].includes(entry.event),
+  );
+  const lookupRun = verifiedRunLookup(repository, api);
+  const mainCandidates = [];
+  for (const entry of mainEntries.slice(0, MAX_MAIN_RUN_LOOKUPS)) {
+    try {
+      const run = await lookupRun(entry);
+      if (isEligibleMainRun(run, defaultBranch, repository))
+        mainCandidates.push({
+          run,
+          getEvidence: mainEvidenceLookup(repository, api, entry, run),
+        });
+      if (mainCandidates.length === MAX_MAIN_CANDIDATES) break;
+    } catch {
+      // Unverified metadata cannot consume the eligible-main candidate budget.
+    }
+  }
+  const entries = [];
+  for (const entry of candidates) {
+    try {
+      entries.push(
+        ...(await analyzeFailedRun({
+          repository,
+          defaultBranch,
+          entry,
+          run: await lookupRun(entry),
+          mainCandidates,
+          api,
+        })),
+      );
+    } catch (error) {
+      entries.push({
+        runId: entry.id,
+        runAttempt: entry.runAttempt,
+        classification: "unknown",
+        reason: errorCode(error),
+      });
+    }
+  }
+  const report = {
+    version: 1,
+    observedAt,
+    repository,
+    defaultBranch,
+    informationalOnly: true,
+    gateEffect: "none; this report never changes CI conclusions",
+    boundedInput: {
+      candidateRuns: candidates.length,
+      limit: MAX_FAILED_RUNS,
+      mainCandidateLimit: MAX_MAIN_CANDIDATES,
+      mainCandidateRuns: mainCandidates.length,
+      mainRunLookupLimit: MAX_MAIN_RUN_LOOKUPS,
+      manifestTruncated:
+        prRuns.length > MAX_FAILED_RUNS ||
+        batchManifest.truncation?.length > 0 ||
+        batchManifest.selection?.exhaustive === false ||
+        batchManifest.coverage?.exhaustive === false,
+    },
+    counts: Object.fromEntries(
+      ["matching-main-diagnostics", "unknown"].map((category) => [
+        category,
+        entries.filter((item) => item.classification === category).length,
+      ]),
+    ),
+    failures: entries,
+  };
+  await writeFailureReport(report, outputDirectory, summaryPath);
+  return report;
+}
+
+async function analyzeFailedRun({
+  repository,
+  defaultBranch,
+  entry,
+  run,
+  mainCandidates,
+  api,
+}) {
+  const jobs = await api.getJobs(repository, entry.id, entry.runAttempt);
+  if (jobs.some((job) => !hasJobIdentity(job, entry)))
+    throw new Error("workflow_job_identity_mismatch");
+  const failedJobs = jobs.filter(
+    (job) => job.status === "completed" && isFailedConclusion(job.conclusion),
+  );
+  if (failedJobs.length === 0)
+    return [
+      {
+        runId: entry.id,
+        runAttempt: entry.runAttempt,
+        classification: "unknown",
+        reason: "failed_jobs_not_present_in_api_snapshot",
+      },
+    ];
+  const failures = [];
+  for (const job of failedJobs) {
+    let classification;
+    try {
+      const evidence = await api.getStaticEvidence(repository, run, job);
+      classification = await attributeToMain({
+        repository,
+        defaultBranch,
+        run,
+        job,
+        evidence,
+        mainCandidates,
+        api,
+      });
+    } catch (error) {
+      classification = outcome("unknown", errorCode(error));
+    }
+    failures.push({
+      runId: entry.id,
+      runAttempt: entry.runAttempt,
+      runHead: safeSha(run.head_sha),
+      event: run.event ?? null,
+      job: job.name,
+      conclusion: job.conclusion,
+      ...classification,
+    });
+  }
+  return failures;
+}
+
+async function attributeToMain({
+  repository,
+  defaultBranch,
+  run,
+  job,
+  evidence,
+  mainCandidates,
+  api,
+}) {
+  const baseSha = pullRequestBaseSha(run, repository);
+  if (baseSha === null || evidence === null)
+    return outcome("unknown", "no_matching_verified_main_failure");
+  let result = outcome("unknown", "no_matching_verified_main_failure");
+  const candidateFailures = [];
+  for (const candidate of mainCandidates) {
+    try {
+      const main = await compareCandidate({
+        repository,
+        defaultBranch,
+        run,
+        job,
+        evidence,
+        baseSha,
+        candidate,
+        api,
+      });
+      if (main !== null) result = main;
+      if (main?.classification === "matching-main-diagnostics") break;
+    } catch (error) {
+      candidateFailures.push({
+        runId: candidate.run.id,
+        runAttempt: candidate.run.run_attempt,
+        reason: errorCode(error),
+      });
+    }
+  }
+  return {
+    ...result,
+    ...(candidateFailures.length > 0 ? { candidateFailures } : {}),
+  };
+}
+
+async function compareCandidate({
+  repository,
+  defaultBranch,
+  run,
+  job,
+  evidence,
+  baseSha,
+  candidate,
+  api,
+}) {
+  const mainRun = candidate.run;
+  const main = await candidate.getEvidence(job.name);
+  if (!main?.evidence) return null;
+  const { job: mainJob, evidence: mainEvidence } = main;
+  const ancestry = await api.getAncestry(repository, mainRun.head_sha, baseSha);
+  const paths = [
+    ...new Set(evidence.diagnostics.flatMap(diagnosticPaths)),
+  ].slice(0, MAX_PATHS_PER_COMPARISON);
+  const blobs = await api.getBlobIdentity(
+    repository,
+    mainRun.head_sha,
+    baseSha,
+    paths,
+  );
+  const executionSourcesVerified =
+    (await api.verifyExecutionSource(repository, run, evidence.context)) &&
+    (await api.verifyExecutionSource(
+      repository,
+      mainRun,
+      mainEvidence.context,
+    ));
+  return classifyFailure({
+    repository,
+    defaultBranch,
+    run,
+    job,
+    diagnostics: evidence.diagnostics,
+    context: evidence.context,
+    mainRun,
+    mainJob,
+    mainDiagnostics: mainEvidence.diagnostics,
+    mainContext: mainEvidence.context,
+    executionSourcesVerified,
+    ancestry,
+    blobs,
+  });
+}
+
+// Each verified main run/attempt owns its lazy jobs and parsed evidence for this
+// report only. Cache nulls and rejected reads too; retries cannot spend the budget
+// repeatedly on an immutable observation that was unavailable or invalid.
+function mainEvidenceLookup(repository, api, entry, run) {
+  let jobs;
+  const evidence = new Map();
+  return (name) => {
+    jobs ??= Promise.resolve().then(async () => {
+      const items = await api.getJobs(repository, entry.id, entry.runAttempt);
+      return items.some((item) => !hasJobIdentity(item, entry)) ? [] : items;
+    });
+    if (!evidence.has(name))
+      evidence.set(
+        name,
+        jobs.then(async (items) => {
+          const job = items.find(
+            (item) =>
+              item.name === name &&
+              item.status === "completed" &&
+              isFailedConclusion(item.conclusion),
+          );
+          if (!job) return null;
+          return {
+            job,
+            evidence: await api.getStaticEvidence(repository, run, job),
+          };
+        }),
+      );
+    return evidence.get(name);
+  };
+}
+
+// Selection and attribution share one immutable API observation per run/attempt.
+function verifiedRunLookup(repository, api) {
+  const runs = new Map();
+  return (entry) => {
+    const key = `${positiveId(entry.id)}:${positiveId(entry.runAttempt)}`;
+    if (!runs.has(key))
+      runs.set(
+        key,
+        Promise.resolve().then(async () => {
+          const run = normalizeWorkflowRun(
+            await api.getRun(repository, entry.id),
+          );
+          if (!hasRunIdentity(run, entry))
+            throw new Error("workflow_run_identity_mismatch");
+          return run;
+        }),
+      );
+    return runs.get(key);
+  };
+}
+
+function hasRunIdentity(run, entry) {
+  return (
+    String(run?.id) === String(entry.id) &&
+    Number(run.run_attempt) === Number(entry.runAttempt) &&
+    run.workflowPath?.file === ".github/workflows/ci.yml"
+  );
+}
+function hasJobIdentity(job, entry) {
+  return (
+    String(job.run_id) === String(entry.id) &&
+    Number(job.run_attempt) === Number(entry.runAttempt)
+  );
+}
+
+export function createGitHubApi({
+  token = process.env.GH_TOKEN,
+  fetchImpl = fetch,
+  apiRequest = createGhApiRequest(),
+} = {}) {
+  if (!token) throw new Error("GH_TOKEN is required");
+  const budget = { requests: 0, startedAt: Date.now(), limitMs: 60_000 };
+  const request = async (endpoint) => {
+    if (budget.requests >= MAX_API_REQUESTS)
+      throw new Error("github_api_request_budget_exhausted");
+    budget.requests += 1;
+    return await apiRequest("GET", endpoint, budget);
+  };
+  return {
+    async getRun(repository, id) {
+      return await request(
+        `repos/${repository}/actions/runs/${positiveId(id)}`,
+      );
+    },
+    async getJobs(repository, id, attempt) {
+      const page = await request(
+        `repos/${repository}/actions/runs/${positiveId(id)}/attempts/${positiveId(attempt)}/jobs?per_page=100&page=1`,
+      );
+      if (!Array.isArray(page.jobs) || page.jobs.length >= 100)
+        throw new Error("job_list_incomplete_or_malformed");
+      return page.jobs;
+    },
+    async getStaticEvidence(repository, run, job) {
+      if (job.name !== "Static checks") return null;
+      if (run.run_attempt !== 1) return null;
+      const page = await request(
+        `repos/${repository}/actions/runs/${positiveId(run.id)}/artifacts?per_page=100&page=1`,
+      );
+      const artifact = selectStaticCheckArtifact(page, run, run.run_attempt);
+      if (!artifact) return null;
+      return await readStaticCheckArtifact({
+        repository,
+        artifactId: artifact.id,
+        token,
+        fetchImpl,
+      });
+    },
+    async getAncestry(repository, ancestor, descendant) {
+      const comparison = await request(
+        `repos/${repository}/compare/${safeSha(ancestor)}...${safeSha(descendant)}`,
+      );
+      return {
+        status: comparison.status,
+        mergeBaseSha: safeSha(comparison.merge_base_commit?.sha),
+      };
+    },
+    async verifyExecutionSource(repository, run, context) {
+      const sourceSha = safeSha(context.sourceSha);
+      if (!sourceSha) return false;
+      const pr =
+        run.event === "pull_request"
+          ? pullRequestBaseSha(run, repository)
+          : null;
+      if (
+        run.event === "pull_request" &&
+        (pr === null || context.prHeadSha !== run.head_sha)
+      )
+        return false;
+      const commit = await request(`repos/${repository}/commits/${sourceSha}`);
+      if (safeSha(commit.sha) !== sourceSha) return false;
+      if (sourceSha === run.head_sha) return true;
+      if (
+        pr === null ||
+        !Array.isArray(commit.parents) ||
+        commit.parents.length !== 2
+      )
+        return false;
+      return (
+        safeSha(commit.parents[0]?.sha) === pr &&
+        safeSha(commit.parents[1]?.sha) === run.head_sha
+      );
+    },
+    async getBlobIdentity(repository, ancestor, descendant, paths) {
+      const result = {};
+      for (const path of paths) {
+        if (!validRepoPath(path)) continue;
+        const [left, right] = await Promise.all([
+          getContentBlobSha(repository, ancestor, path, request),
+          getContentBlobSha(repository, descendant, path, request),
+        ]);
+        result[path] = left === right;
+      }
+      return result;
+    },
+  };
+}
+
+async function getContentBlobSha(repository, ref, path, request) {
+  const encodedPath = path.split("/").map(encodeURIComponent).join("/");
+  const value = await request(
+    `repos/${repository}/contents/${encodedPath}?ref=${encodeURIComponent(ref)}`,
+  );
+  const sha = safeSha(value?.sha);
+  if (value?.type !== "file" || sha === null)
+    throw new Error("github_content_blob_identity_unavailable");
+  return sha;
+}
+
+function outcome(classification, reason) {
+  return { classification, reason };
+}
+function validateRepository(value) {
+  if (
+    typeof value !== "string" ||
+    !/^[A-Za-z0-9_.-]+\/[A-Za-z0-9_.-]+$/u.test(value)
+  )
+    throw new Error("repository must be owner/name");
+}
+function positiveId(value) {
+  if (!Number.isSafeInteger(Number(value)) || Number(value) < 1)
+    throw new Error(
+      "GitHub run and attempt IDs must be positive safe integers",
+    );
+  return String(Number(value));
+}
+function errorCode(error) {
+  return error instanceof Error ? error.message.slice(0, 160) : "unknown_error";
+}
+
+export async function writeFailureReport(
+  report,
+  outputDirectory = OUTPUT_DIRECTORY,
+  summaryPath,
+) {
+  await mkdir(outputDirectory, { recursive: true });
+  const jsonPath = join(outputDirectory, "report.json");
+  const markdownPath = join(outputDirectory, "report.md");
+  const matchedDiagnostics = report.failures.filter(
+    (item) => item.classification === "matching-main-diagnostics",
+  );
+  const lines = [
+    "# CI failure attribution",
+    "",
+    "Informational only. This report never changes a CI conclusion or required gate.",
+    "",
+    `Observed ${report.observedAt}; ${String(matchedDiagnostics.length)} of ${String(report.failures.length)} failure reports contain supported diagnostics matching a failed static task on an ancestor main revision. This does not attribute other failures in the same job. Unknown means evidence was incomplete or did not match.`,
+    "",
+  ];
+  for (const item of report.failures)
+    lines.push(
+      `- Run ${item.runId} attempt ${item.runAttempt}${item.job ? `, ${item.job}` : ""}: **${item.classification}** — ${item.reason}.`,
+    );
+  if (report.failures.length === 0)
+    lines.push(
+      "- No completed failed CI attempts were present in the supplied timing snapshot.",
+    );
+  await writeFile(jsonPath, `${JSON.stringify(report, null, 2)}\n`, "utf8");
+  await writeFile(markdownPath, `${lines.join("\n")}\n`, "utf8");
+  if (summaryPath)
+    await appendFile(summaryPath, `${lines.join("\n")}\n`, "utf8");
+  return { jsonPath, markdownPath };
+}
+
+async function main() {
+  const repository = process.env.CI_REPOSITORY;
+  const defaultBranch = process.env.CI_DEFAULT_BRANCH;
+  const batchPath =
+    process.env.CI_TIMING_MANIFEST ??
+    join(BATCH_DIRECTORY, "batch-manifest.json");
+  const batchManifest = JSON.parse(await readBoundedFile(batchPath));
+  const api = createGitHubApi();
+  await reportCiFailures({
+    repository,
+    defaultBranch,
+    batchManifest,
+    api,
+    outputDirectory: process.env.CI_FAILURES_DIRECTORY ?? OUTPUT_DIRECTORY,
+    summaryPath: process.env.GITHUB_STEP_SUMMARY,
+  });
+}
+async function readBoundedFile(path) {
+  const bytes = await readFile(path);
+  if (bytes.byteLength > MAX_JSON_BYTES)
+    throw new Error("timing manifest exceeds size limit");
+  return bytes.toString("utf8");
+}
+
+if (
+  process.argv[1] &&
+  resolve(process.argv[1]) === fileURLToPath(import.meta.url)
+) {
+  try {
+    await main();
+  } catch (error) {
+    console.error(error instanceof Error ? error.message : String(error));
+    process.exitCode = 1;
+  }
+}
@@ -0,0 +1,16 @@
+export interface GitHubApiBudget {
+  requests: number;
+  startedAt: number;
+  limitMs: number;
+}
+
+export type GitHubApiRequest = (
+  method: string,
+  endpoint: string,
+  budget: GitHubApiBudget,
+) => Promise<unknown>;
+
+export function createGhApiRequest(options?: {
+  env?: NodeJS.ProcessEnv;
+  now?: () => number;
+}): GitHubApiRequest;
@@ -0,0 +1,33 @@
+import { execFile } from "node:child_process";
+import { promisify } from "node:util";
+
+const execFileAsync = promisify(execFile);
+const MAX_BUFFER = 8 * 1024 * 1024;
+const REQUEST_TIMEOUT_MS = 10_000;
+
+/** Create a bounded GitHub CLI JSON transport with GET-only authority. */
+export function createGhApiRequest({ env = process.env, now = Date.now } = {}) {
+  return async (method, endpoint, budget) => {
+    if (method !== "GET")
+      throw new Error(
+        "GitHub API transport supports read-only GET requests only",
+      );
+    const remaining = budget.limitMs - (now() - budget.startedAt);
+    if (remaining <= 0) throw new Error("lookup_time_budget_exhausted");
+    const { stdout } = await execFileAsync(
+      "gh",
+      ["api", "--method", "GET", endpoint],
+      {
+        env,
+        encoding: "utf8",
+        maxBuffer: MAX_BUFFER,
+        timeout: Math.min(REQUEST_TIMEOUT_MS, remaining),
+      },
+    );
+    try {
+      return JSON.parse(stdout);
+    } catch (cause) {
+      throw new Error("github_api_malformed_json", { cause });
+    }
+  };
+}
@@ -103,6 +103,9 @@ export async function collectJobTimingsBatch({
          id: run.id,
          runAttempt: run.run_attempt,
          workflow: workflow.name,
+          event: typeof run.event === "string" ? run.event : null,
+          headBranch:
+            typeof run.head_branch === "string" ? run.head_branch : null,
          conclusion: run.conclusion ?? null,
          createdAt: run.created_at ?? null,
        });
@@ -62,6 +62,10 @@ function classifyPackage(base, head) {
    ]),
  ].filter((key) => beforeScripts?.[key] !== afterScripts?.[key]);
  for (const script of changed) {
+    if (/^test:(coverage|mutation):reverse-tools$/.test(script)) {
+      selected.reverse_tools = true;
+      continue;
+    }
    if (script === "test:sqlite" || script === "test:sqlite:run") {
      selected.sqlite = true;
      continue;
@@ -162,6 +166,7 @@ if (tier === "reused") {
      .split("\0")
      .filter(Boolean);
    for (const file of files) {
+      if (matchesAny(file, SCOPES.reverse_tools)) selected.reverse_tools = true;
      if (
        /^(README(?:_[^/]+)?\.md|AGENTS\.md|CONTRIBUTING\.md|CHANGELOG\.md|SECURITY\.md|native\/windows\/README\.md|\.github\/pull_request_template\.md)$/.test(
          file,
@@ -0,0 +1,95 @@
+export interface PrCiHead {
+  repositoryId: number | null;
+  repository?: string | null;
+  ref: string | null;
+  sha: string | null;
+}
+
+export interface PrCiPullRequest {
+  number: number | null;
+  draft?: boolean;
+  identityValid: boolean;
+  mergeable?: boolean;
+  head: PrCiHead;
+  [key: string]: unknown;
+}
+
+export interface PrCiWorkflowRun {
+  id?: number;
+  run_attempt?: number;
+  event?: string;
+  head_branch?: string;
+  head_repository?: { id?: number | null } | null;
+  head_sha?: string;
+  status?: string;
+  conclusion?: string | null;
+  created_at?: string;
+  html_url?: string;
+  pull_requests?: Array<{
+    number?: number;
+    head?: { sha?: string; repo?: { id?: number | null } } | null;
+  }>;
+  [key: string]: unknown;
+}
+
+export interface PrCiClassification {
+  category: string;
+  reason: string;
+  runs: Array<Record<string, unknown>>;
+}
+
+export interface PrCiHealthReport {
+  schemaVersion: 1;
+  observedAt: string;
+  repository: string;
+  repositoryId: number | null;
+  defaultBranch: string;
+  lookbackDays: number;
+  scan: {
+    status: "complete" | "partial";
+    cutoff: string;
+    pullRequestPages: number;
+    workflowRunPages: number;
+    pullRequestInventoryComplete: boolean;
+    workflowRunInventoryComplete: boolean;
+    pullRequestDetailsFetched: number;
+    pullRequestDetailsLimit: number;
+    apiRequests: number;
+    apiRequestLimit: number;
+    errors: string[];
+  };
+  counts: Record<string, number>;
+  pullRequests: Array<PrCiPullRequest & { ci: PrCiClassification }>;
+}
+
+export function collectPrCiHealth(options: {
+  repository: string;
+  repositoryId?: number | string | null;
+  defaultBranch: string;
+  lookbackDays?: number;
+  now?: number;
+  apiRequest?: GitHubApiRequest;
+}): Promise<PrCiHealthReport>;
+
+export function classifyPrCi(input: {
+  pull: PrCiPullRequest;
+  runs: Array<{ run: PrCiWorkflowRun; basis: string }>;
+  runsComplete: boolean;
+  pullRequestInventoryComplete: boolean;
+  ambiguousRuns: Array<{ run: PrCiWorkflowRun; basis: string }>;
+}): PrCiClassification;
+
+export function matchRunToPullRequest(input: {
+  run: PrCiWorkflowRun;
+  pull: PrCiPullRequest;
+  identityCounts: Map<string, number>;
+  inventoryComplete?: boolean;
+}): { kind: "none" } | { kind: "ambiguous" } | { kind: "match"; basis: string };
+
+export function writePrCiHealthReport(
+  report: PrCiHealthReport,
+  outputDirectory?: string,
+): Promise<{ jsonPath: string; markdownPath: string }>;
+
+export function renderMarkdownSummary(report: PrCiHealthReport): string;
+import type { GitHubApiRequest } from "./github-api.mjs";
@@ -0,0 +1,718 @@
+import { appendFile, mkdir, writeFile } from "node:fs/promises";
+import { join, resolve } from "node:path";
+import { fileURLToPath } from "node:url";
+
+import { createGhApiRequest } from "./github-api.mjs";
+
+const OUTPUT_DIRECTORY = join(".cache", "ci-pr-health");
+const PAGE_SIZE = 100;
+const MAX_PULL_REQUEST_PAGES = 3;
+const MAX_WORKFLOW_RUN_PAGES = 5;
+const MAX_PULL_REQUEST_DETAILS = 40;
+const MAX_API_REQUESTS =
+  MAX_PULL_REQUEST_PAGES + MAX_WORKFLOW_RUN_PAGES + MAX_PULL_REQUEST_DETAILS;
+const LOOKBACK_DAYS = 14;
+const MAX_LOOKBACK_DAYS = 30;
+const MAX_TOTAL_TIME_MS = 60_000;
+const SHA_PATTERN = /^[0-9a-f]{40}$/iu;
+const COMPLETED_CATEGORIES = new Map([
+  ["failure", "completed_failure"],
+  ["timed_out", "completed_failure"],
+  ["cancelled", "completed_canceled"],
+  ["success", "completed_success"],
+]);
+
+/** Read workflow/PR state and conservatively distinguish absent CI from unknown. */
+export async function collectPrCiHealth({
+  repository,
+  repositoryId,
+  defaultBranch,
+  lookbackDays = LOOKBACK_DAYS,
+  now = Date.now(),
+  apiRequest = createGhApiRequest(),
+}) {
+  const identity = validateInputs({
+    repository,
+    repositoryId,
+    defaultBranch,
+    lookbackDays,
+    now,
+  });
+  const observedAt = new Date(now).toISOString();
+  const cutoff = now - identity.lookbackDays * 24 * 60 * 60 * 1_000;
+  const snapshot = await collectApiSnapshot({ apiRequest, identity, cutoff });
+  const classified = classifyPullRequests(snapshot);
+  const { pulls, runs, budget } = snapshot;
+  const errors = [pulls.error, runs.error, ...snapshot.detailsErrors].filter(
+    Boolean,
+  );
+  return {
+    schemaVersion: 1,
+    observedAt,
+    repository: identity.repository,
+    repositoryId: identity.repositoryId,
+    defaultBranch: identity.defaultBranch,
+    lookbackDays: identity.lookbackDays,
+    scan: {
+      status:
+        errors.length === 0 && pulls.complete && runs.complete
+          ? "complete"
+          : "partial",
+      cutoff: new Date(cutoff).toISOString(),
+      pullRequestPages: pulls.pages,
+      workflowRunPages: runs.pages,
+      pullRequestInventoryComplete: pulls.complete,
+      workflowRunInventoryComplete: runs.complete,
+      pullRequestDetailsFetched: snapshot.detailsFetched,
+      pullRequestDetailsLimit: MAX_PULL_REQUEST_DETAILS,
+      apiRequests: budget.requests,
+      apiRequestLimit: MAX_API_REQUESTS,
+      errors,
+    },
+    counts: countCategories(classified),
+    pullRequests: classified,
+  };
+}
+
+async function collectApiSnapshot({ apiRequest, identity, cutoff }) {
+  const budget = {
+    requests: 0,
+    startedAt: Date.now(),
+    limitMs: MAX_TOTAL_TIME_MS,
+  };
+  const requestGet = async (endpoint) => {
+    if (budget.requests >= MAX_API_REQUESTS)
+      throw new Error("lookup_request_budget_exhausted");
+    if (Date.now() - budget.startedAt >= budget.limitMs)
+      throw new Error("lookup_time_budget_exhausted");
+    budget.requests += 1;
+    return await apiRequest("GET", endpoint, budget);
+  };
+  const pulls = await listOpenPullRequests({
+    requestGet,
+    identity,
+  });
+  const candidates = pulls.items.filter(
+    (pull) =>
+      !pull.draft &&
+      Number.isFinite(Date.parse(pull.createdAt ?? "")) &&
+      Date.parse(pull.createdAt ?? "") >= cutoff &&
+      pull.base.ref === identity.defaultBranch &&
+      (identity.repositoryId === null ||
+        pull.base.repositoryId === identity.repositoryId),
+  );
+  const runs = await listRecentWorkflowRuns({
+    requestGet,
+    identity,
+    cutoff,
+  });
+
+  const detailResult = await resolveMergeability({
+    candidates,
+    requestGet,
+    identity,
+  });
+  return {
+    budget,
+    pulls,
+    runs,
+    candidates,
+    detailsFetched: detailResult.fetched,
+    detailsErrors: detailResult.errors,
+  };
+}
+
+function classifyPullRequests(snapshot) {
+  const { candidates, pulls, runs } = snapshot;
+  const identityCounts = countHeadIdentities(pulls.items);
+  return candidates.map((pull) => {
+    const relevant = [];
+    const ambiguousRuns = [];
+    for (const run of runs.items) {
+      const match = matchRunToPullRequest({
+        run,
+        pull,
+        identityCounts,
+        inventoryComplete: pulls.complete,
+      });
+      if (match.kind === "match") relevant.push({ run, basis: match.basis });
+      if (match.kind === "ambiguous")
+        ambiguousRuns.push({ run, basis: "ambiguous_head_identity" });
+    }
+    return {
+      ...pull,
+      ci: classifyPrCi({
+        pull,
+        runs: relevant,
+        runsComplete: runs.complete,
+        pullRequestInventoryComplete: pulls.complete,
+        ambiguousRuns,
+      }),
+    };
+  });
+}
+
+async function resolveMergeability({ candidates, requestGet, identity }) {
+  let detailsFetched = 0;
+  const errors = new Set();
+  for (const pull of candidates) {
+    if (pull.mergeable !== undefined) continue;
+    if (detailsFetched >= MAX_PULL_REQUEST_DETAILS) {
+      errors.add("pull_request_mergeability_lookup_budget_exhausted");
+      continue;
+    }
+    const result = await getPullRequestDetails({ requestGet, identity, pull });
+    detailsFetched += 1;
+    if (result.error !== null) {
+      errors.add(result.error);
+      continue;
+    }
+    const { detail } = result;
+    const identityChanged =
+      detail.headSha !== pull.head.sha ||
+      detail.headRef !== pull.head.ref ||
+      detail.headRepositoryId !== pull.head.repositoryId ||
+      detail.baseRef !== pull.base.ref ||
+      detail.baseRepositoryId !== pull.base.repositoryId;
+    if (identityChanged) {
+      pull.identityValid = false;
+      pull.identityChangedDuringScan = true;
+      errors.add("pull_request_identity_changed_during_scan");
+    } else {
+      pull.mergeable = detail.mergeable;
+      pull.detailFetched = true;
+      if (detail.mergeable === undefined)
+        errors.add("pull_request_mergeability_unknown");
+    }
+  }
+  return { fetched: detailsFetched, errors: [...errors] };
+}
+
+function countCategories(pulls) {
+  const categories = [
+    "missing",
+    "startup_failure",
+    "queued_pending_executing",
+    "approval_required",
+    "completed_success",
+    "completed_failure",
+    "completed_canceled",
+    "completed_other",
+    "merge_conflict",
+    "unknown",
+    "deferred",
+  ];
+  return Object.fromEntries(
+    categories.map((category) => [
+      category,
+      pulls.filter(({ ci }) => ci.category === category).length,
+    ]),
+  );
+}
+
+/** Pure status classifier; missing requires complete, identity-safe inventories. */
+export function classifyPrCi({
+  pull,
+  runs,
+  runsComplete,
+  pullRequestInventoryComplete,
+  ambiguousRuns,
+}) {
+  if (!pull.identityValid)
+    return result("unknown", "pull_request_head_identity_invalid", []);
+  const evidence = [...runs, ...ambiguousRuns];
+  if (pull.mergeable === false)
+    return result("merge_conflict", "github_reports_merge_conflict", evidence);
+  // The newest collected run is not proven current until pagination completes:
+  // an older-created run outside this snapshot may have been retried recently.
+  if (!runsComplete)
+    return result(
+      evidence.length > 0 ? "unknown" : "deferred",
+      evidence.length > 0
+        ? "workflow_run_inventory_incomplete"
+        : "lookup_inventory_incomplete",
+      evidence,
+    );
+  const selection = newestRun(runs);
+  if (!selection.known)
+    return result("unknown", "workflow_run_order_unknown", evidence);
+  const latest = selection.latest;
+  if (
+    ambiguousRuns.some((candidate) => {
+      const order = latest ? compareRunEvidence(candidate, latest) : null;
+      return order === null || order >= 0;
+    })
+  )
+    return result("unknown", "workflow_run_identity_ambiguous", evidence);
+  if (latest !== undefined) {
+    const run = latest.run;
+    if (run.conclusion === "startup_failure")
+      return result("startup_failure", "workflow_startup_failure", evidence);
+    if (run.status === "waiting" || run.conclusion === "action_required")
+      return result(
+        "approval_required",
+        "workflow_waiting_for_approval_or_gate",
+        evidence,
+      );
+    if (["queued", "requested", "pending", "in_progress"].includes(run.status))
+      return result(
+        "queued_pending_executing",
+        `workflow_${run.status}`,
+        evidence,
+      );
+    if (run.status === "completed")
+      return result(
+        COMPLETED_CATEGORIES.get(run.conclusion) ?? "completed_other",
+        run.conclusion === "success"
+          ? "workflow_succeeded"
+          : `workflow_${run.conclusion ?? "unknown_conclusion"}`,
+        evidence,
+      );
+    return result("unknown", "workflow_status_unrecognized", evidence);
+  }
+  if (!pullRequestInventoryComplete)
+    return result("deferred", "lookup_inventory_incomplete", []);
+  if (pull.mergeable !== true)
+    return result("unknown", "pull_request_mergeability_unknown", []);
+  return result("missing", "no_pull_request_event_run_for_current_head", []);
+}
+
+/** Match only a pull_request event proven to belong to this immutable PR head. */
+export function matchRunToPullRequest({
+  run,
+  pull,
+  identityCounts,
+  inventoryComplete = true,
+}) {
+  if (run?.event !== "pull_request") return { kind: "none" };
+  if (!pull.identityValid || !isRecord(run)) return { kind: "none" };
+  const association = matchExplicitPullAssociation(run, pull);
+  if (association === "match")
+    return { kind: "match", basis: "event_pr_number_and_head_sha" };
+  if (association === "ambiguous") return { kind: "ambiguous" };
+  if (association === "other" || !sameHeadRepositoryAndRef(run, pull))
+    return { kind: "none" };
+  if (!uniqueHeadRepositoryAndRef(pull, identityCounts, inventoryComplete))
+    return { kind: "ambiguous" };
+  if (isSha(run.head_sha) && run.head_sha.toLowerCase() === pull.head.sha)
+    return { kind: "match", basis: "event_unique_head_ref_and_repository_id" };
+  return { kind: "none" };
+}
+
+function matchExplicitPullAssociation(run, pull) {
+  const associations = Array.isArray(run.pull_requests)
+    ? run.pull_requests
+    : [];
+  if (associations.length === 0) return "absent";
+  const matchingNumber = associations.filter(
+    (item) => item?.number === pull.number,
+  );
+  if (matchingNumber.length === 0) return "other";
+  const exactHead = matchingNumber.find(
+    (item) => item?.head?.sha === pull.head.sha,
+  );
+  if (exactHead === undefined) {
+    return matchingNumber.some(
+      (item) =>
+        typeof item?.head?.sha === "string" && item.head.sha !== pull.head.sha,
+    )
+      ? "other"
+      : "absent";
+  }
+  const associatedHeadRepositoryId = positiveInteger(exactHead.head?.repo?.id);
+  const runHeadRepositoryId = positiveInteger(run.head_repository?.id);
+  if (associatedHeadRepositoryId === null || runHeadRepositoryId === null)
+    return "ambiguous";
+  if (
+    associatedHeadRepositoryId !== pull.head.repositoryId ||
+    runHeadRepositoryId !== pull.head.repositoryId
+  )
+    return "other";
+  const baseRefMatches =
+    typeof exactHead.base?.ref !== "string" ||
+    exactHead.base.ref === pull.base.ref;
+  const baseRepositoryId = positiveInteger(exactHead.base?.repo?.id);
+  const baseRepositoryMatches =
+    baseRepositoryId === null || baseRepositoryId === pull.base.repositoryId;
+  return baseRefMatches && baseRepositoryMatches ? "match" : "other";
+}
+
+function sameHeadRepositoryAndRef(run, pull) {
+  return (
+    run.head_branch === pull.head.ref &&
+    run.head_repository?.id === pull.head.repositoryId
+  );
+}
+
+function uniqueHeadRepositoryAndRef(pull, identityCounts, inventoryComplete) {
+  if (!inventoryComplete) return false;
+  const key = headIdentityKey(pull.head.repositoryId, pull.head.ref);
+  const unknownRepositoryKey = headIdentityKey(null, pull.head.ref);
+  return (
+    identityCounts.get(key) === 1 &&
+    (identityCounts.get(unknownRepositoryKey) ?? 0) === 0
+  );
+}
+
+function normalizePullRequest(raw, identity) {
+  const number = positiveInteger(raw?.number);
+  const base = normalizeBase(raw?.base);
+  const head = normalizeHead(raw?.head);
+  const draft = raw?.draft === true;
+  return {
+    number,
+    draft,
+    createdAt: typeof raw?.created_at === "string" ? raw.created_at : null,
+    url: typeof raw?.html_url === "string" ? raw.html_url : null,
+    base,
+    head,
+    mergeable: typeof raw?.mergeable === "boolean" ? raw.mergeable : undefined,
+    identityValid: hasValidPullIdentity({ raw, identity, number, base, head }),
+  };
+}
+
+function normalizeBase(raw) {
+  return {
+    ref: typeof raw?.ref === "string" ? raw.ref : null,
+    repositoryId: positiveInteger(raw?.repo?.id),
+  };
+}
+
+function normalizeHead(raw) {
+  return {
+    repositoryId: positiveInteger(raw?.repo?.id),
+    repository:
+      typeof raw?.repo?.full_name === "string" ? raw.repo.full_name : null,
+    ref: typeof raw?.ref === "string" ? raw.ref : null,
+    sha:
+      typeof raw?.sha === "string" && isSha(raw.sha)
+        ? raw.sha.toLowerCase()
+        : null,
+  };
+}
+
+function hasValidPullIdentity({ raw, identity, number, base, head }) {
+  return (
+    number !== null &&
+    typeof raw?.draft === "boolean" &&
+    base.ref === identity.defaultBranch &&
+    (identity.repositoryId === null ||
+      base.repositoryId === identity.repositoryId) &&
+    head.repositoryId !== null &&
+    head.sha !== null &&
+    head.ref !== null
+  );
+}
+
+function countHeadIdentities(pulls) {
+  const counts = new Map();
+  for (const pull of pulls) {
+    if (typeof pull.head.ref !== "string") continue;
+    const key = headIdentityKey(pull.head.repositoryId, pull.head.ref);
+    counts.set(key, (counts.get(key) ?? 0) + 1);
+  }
+  return counts;
+}
+
+function headIdentityKey(repositoryId, ref) {
+  return `${String(repositoryId)}\u0000${ref}`;
+}
+
+function newestRun(runs) {
+  let latest = runs[0];
+  for (const candidate of runs) {
+    if (latest && (compareRunEvidence(candidate, latest) ?? 0) > 0)
+      latest = candidate;
+  }
+  const known = runs.every((candidate) => {
+    if (candidate === latest) return true;
+    const order = compareRunEvidence(candidate, latest);
+    return order !== null && order <= 0;
+  });
+  return { known, latest };
+}
+
+// Attempt ordering is meaningful within one run; different run IDs are not clocks.
+function compareRunEvidence(left, right) {
+  const leftRun = left.run;
+  const rightRun = right.run;
+  const id = positiveInteger(leftRun.id);
+  if (id !== null && id === positiveInteger(rightRun.id)) {
+    const leftAttempt = positiveInteger(leftRun.run_attempt);
+    const rightAttempt = positiveInteger(rightRun.run_attempt);
+    if (leftAttempt === null || rightAttempt === null) return null;
+    if (leftAttempt !== rightAttempt) return leftAttempt - rightAttempt;
+    return leftRun.status === rightRun.status &&
+      leftRun.conclusion === rightRun.conclusion
+      ? 0
+      : null;
+  }
+  // A retry may start long after its original creation; without an attempt
+  // clock, creation dates only order first attempts from different runs.
+  if (
+    positiveInteger(leftRun.run_attempt) !== 1 ||
+    positiveInteger(rightRun.run_attempt) !== 1
+  )
+    return null;
+  const time =
+    Date.parse(leftRun.created_at ?? "") -
+    Date.parse(rightRun.created_at ?? "");
+  return Number.isFinite(time) && time !== 0 ? time : null;
+}
+
+function result(category, reason, runs) {
+  return {
+    category,
+    reason,
+    runs: runs.map(({ run, basis }) => ({
+      id: positiveInteger(run.id),
+      attempt: positiveInteger(run.run_attempt),
+      status: typeof run.status === "string" ? run.status : null,
+      conclusion: typeof run.conclusion === "string" ? run.conclusion : null,
+      headSha: isSha(run.head_sha) ? run.head_sha.toLowerCase() : null,
+      createdAt: typeof run.created_at === "string" ? run.created_at : null,
+      url: typeof run.html_url === "string" ? run.html_url : null,
+      identityBasis: basis,
+    })),
+  };
+}
+
+async function listOpenPullRequests({ requestGet, identity }) {
+  const items = [];
+  let complete = false;
+  let pages = 0;
+  let error = null;
+  for (let page = 1; page <= MAX_PULL_REQUEST_PAGES; page += 1) {
+    try {
+      const endpoint = `repos/${identity.repository}/pulls?state=open&sort=created&direction=desc&per_page=${PAGE_SIZE}&page=${page}`;
+      const response = await requestGet(endpoint);
+      if (!Array.isArray(response))
+        throw new Error("pull_request_page_malformed");
+      pages += 1;
+      const normalized = response.map((raw) =>
+        normalizePullRequest(raw, identity),
+      );
+      if (
+        normalized.some(
+          (pull) => !Number.isFinite(Date.parse(pull.createdAt ?? "")),
+        )
+      )
+        throw new Error("pull_request_page_malformed");
+      items.push(...normalized);
+      if (response.length < PAGE_SIZE) {
+        complete = true;
+        break;
+      }
+    } catch (cause) {
+      error = safeErrorCode(cause);
+      break;
+    }
+  }
+  return { items, pages, complete, error };
+}
+
+async function listRecentWorkflowRuns({ requestGet, identity, cutoff }) {
+  const items = [];
+  let complete = false;
+  let pages = 0;
+  let error = null;
+  for (let page = 1; page <= MAX_WORKFLOW_RUN_PAGES; page += 1) {
+    try {
+      const cutoffDate = new Date(cutoff).toISOString().slice(0, 10);
+      const endpoint = `repos/${identity.repository}/actions/workflows/ci.yml/runs?event=pull_request&created=%3E%3D${cutoffDate}&per_page=${PAGE_SIZE}&page=${page}`;
+      const response = await requestGet(endpoint);
+      if (!isRecord(response) || !Array.isArray(response.workflow_runs))
+        throw new Error("workflow_run_page_malformed");
+      pages += 1;
+      if (!response.workflow_runs.every(isRecord))
+        throw new Error("workflow_run_page_malformed");
+      items.push(...response.workflow_runs);
+      if (response.workflow_runs.length < PAGE_SIZE) {
+        complete = true;
+        break;
+      }
+    } catch (cause) {
+      error = safeErrorCode(cause);
+      break;
+    }
+  }
+  return { items, pages, complete, error };
+}
+
+async function getPullRequestDetails({ requestGet, identity, pull }) {
+  if (pull.number === null)
+    return { detail: null, error: "pull_request_detail_identity_missing" };
+  try {
+    const response = await requestGet(
+      `repos/${identity.repository}/pulls/${pull.number}`,
+    );
+    if (!isRecord(response))
+      return { detail: null, error: "pull_request_detail_malformed" };
+    return {
+      detail: {
+        mergeable:
+          typeof response.mergeable === "boolean"
+            ? response.mergeable
+            : undefined,
+        headSha: isSha(response.head?.sha)
+          ? response.head.sha.toLowerCase()
+          : null,
+        headRef:
+          typeof response.head?.ref === "string" ? response.head.ref : null,
+        headRepositoryId: positiveInteger(response.head?.repo?.id),
+        baseRef:
+          typeof response.base?.ref === "string" ? response.base.ref : null,
+        baseRepositoryId: positiveInteger(response.base?.repo?.id),
+      },
+      error: null,
+    };
+  } catch (cause) {
+    return { detail: null, error: safeErrorCode(cause) };
+  }
+}
+
+function validateInputs({
+  repository,
+  repositoryId,
+  defaultBranch,
+  lookbackDays,
+  now,
+}) {
+  if (
+    typeof repository !== "string" ||
+    !/^[A-Za-z0-9_.-]+\/[A-Za-z0-9_.-]+$/u.test(repository)
+  )
+    throw new Error(
+      "CI_REPOSITORY must be an owner/name repository identifier",
+    );
+  if (
+    typeof defaultBranch !== "string" ||
+    defaultBranch.length === 0 ||
+    /[\r\n]/u.test(defaultBranch)
+  )
+    throw new Error("CI_DEFAULT_BRANCH must be a nonempty branch name");
+  const parsedRepositoryId =
+    repositoryId === undefined || repositoryId === null || repositoryId === ""
+      ? null
+      : positiveInteger(repositoryId);
+  if (
+    repositoryId !== undefined &&
+    repositoryId !== null &&
+    repositoryId !== "" &&
+    parsedRepositoryId === null
+  )
+    throw new Error("CI_REPOSITORY_ID must be a positive safe integer");
+  if (
+    !Number.isInteger(lookbackDays) ||
+    lookbackDays < 1 ||
+    lookbackDays > MAX_LOOKBACK_DAYS
+  )
+    throw new Error(
+      `lookbackDays must be an integer from 1 to ${String(MAX_LOOKBACK_DAYS)}`,
+    );
+  if (!Number.isFinite(now)) throw new Error("now must be a finite timestamp");
+  return {
+    repository,
+    repositoryId: parsedRepositoryId,
+    defaultBranch,
+    lookbackDays,
+  };
+}
+
+function positiveInteger(value) {
+  const number = typeof value === "number" ? value : Number(value);
+  return Number.isSafeInteger(number) && number > 0 ? number : null;
+}
+
+function isSha(value) {
+  return typeof value === "string" && SHA_PATTERN.test(value);
+}
+
+function isRecord(value) {
+  return typeof value === "object" && value !== null && !Array.isArray(value);
+}
+
+function safeErrorCode(cause) {
+  if (cause?.message === "lookup_time_budget_exhausted")
+    return "lookup_time_budget_exhausted";
+  if (cause?.message === "lookup_request_budget_exhausted")
+    return "lookup_request_budget_exhausted";
+  if (cause?.message === "github_api_malformed_json")
+    return "github_api_malformed_json";
+  if (cause?.message?.endsWith("_malformed")) return cause.message;
+  if (cause?.code === "ETIMEDOUT" || cause?.killed) return "github_api_timeout";
+  return "github_api_request_failed";
+}
+
+export async function writePrCiHealthReport(
+  report,
+  outputDirectory = OUTPUT_DIRECTORY,
+) {
+  const jsonPath = join(outputDirectory, "report.json");
+  const markdownPath = join(outputDirectory, "summary.md");
+  await mkdir(outputDirectory, { recursive: true });
+  await writeFile(jsonPath, `${JSON.stringify(report, null, 2)}\n`, "utf8");
+  await writeFile(markdownPath, renderMarkdownSummary(report), "utf8");
+  return { jsonPath, markdownPath };
+}
+
+export function renderMarkdownSummary(report) {
+  const rows = report.pullRequests.map(
+    (pull) =>
+      `| #${String(pull.number ?? "?")} | ${escapeCell(pull.head.repository ?? "unknown")} | ${escapeCell(pull.head.ref ?? "unknown")} | \`${pull.head.sha ?? "unknown"}\` | ${pull.ci.category} | ${escapeCell(pull.ci.reason)} |`,
+  );
+  return [
+    "# Pull request CI health snapshot",
+    "",
+    `Observed ${report.observedAt} for \`${report.repository}\`, target branch \`${escapeCell(report.defaultBranch)}\`; recent window ${String(report.lookbackDays)} days. Scan: **${report.scan.status}**.`,
+    "",
+    "This is read-only point-in-time telemetry. Missing is reported only when PR and workflow-run inventories are complete and the current immutable PR head identity is unambiguous. Incomplete lookups remain deferred or unknown.",
+    "",
+    `API pages: ${String(report.scan.pullRequestPages)} PR, ${String(report.scan.workflowRunPages)} workflow runs; ${String(report.scan.apiRequests)} bounded requests. Errors: ${report.scan.errors.length === 0 ? "none" : report.scan.errors.join(", ")}.`,
+    "",
+    "| PR | Head repository | Head ref | Head SHA | CI classification | Evidence |",
+    "| ---: | --- | --- | --- | --- | --- |",
+    ...rows,
+    "",
+    "Categories: missing, startup failure, queued/pending/executing, approval required, completed outcomes, merge conflict, unknown, and deferred.",
+    "",
+  ].join("\n");
+}
+
+function escapeCell(value) {
+  return String(value)
+    .replaceAll("|", "\\|")
+    .replaceAll("\n", " ")
+    .replaceAll("\r", " ");
+}
+
+if (
+  process.argv[1] !== undefined &&
+  resolve(process.argv[1]) === fileURLToPath(import.meta.url)
+) {
+  try {
+    const report = await collectPrCiHealth({
+      repository: process.env.CI_REPOSITORY,
+      repositoryId: process.env.CI_REPOSITORY_ID,
+      defaultBranch: process.env.CI_DEFAULT_BRANCH,
+    });
+    const paths = await writePrCiHealthReport(
+      report,
+      process.env.CI_PR_HEALTH_DIRECTORY ?? OUTPUT_DIRECTORY,
+    );
+    if (process.env.GITHUB_STEP_SUMMARY)
+      await appendFile(
+        process.env.GITHUB_STEP_SUMMARY,
+        renderMarkdownSummary(report),
+        "utf8",
+      );
+    process.stdout.write(
+      `Wrote ${report.pullRequests.length} recent PR CI records to ${paths.jsonPath}\n`,
+    );
+  } catch (cause) {
+    process.stderr.write(
+      `PR CI health telemetry failed: ${cause instanceof Error ? cause.message : "unknown_error"}\n`,
+    );
+    process.exitCode = 1;
+  }
+}
@@ -2,6 +2,7 @@ import { execFileSync } from "node:child_process";
 import { appendFileSync, mkdirSync, writeFileSync } from "node:fs";
 import { resolve } from "node:path";
 import { fileURLToPath } from "node:url";
+import { normalizeWorkflowRun } from "./workflow-run.mjs";

 const ARTIFACT_NAME = "ci-full-qualification-v1";
 const MAX_AGE_MS = 24 * 60 * 60 * 1000;
@@ -119,8 +120,8 @@ function admit() {
    // Artifact metadata narrows the search before run lookups. Never scan an
    // unbounded history or treat a missing receipt as a successful qualification.
    for (const artifact of candidates.slice(0, 3)) {
-      const run = api(
-        `repos/${repository}/actions/runs/${artifact.workflow_run.id}`,
+      const run = normalizeWorkflowRun(
+        api(`repos/${repository}/actions/runs/${artifact.workflow_run.id}`),
      );
      // An old run can upload a new artifact when only its gate is rerun.
      // Bound the original run age as well so retries cannot renew stale proof.
@@ -133,7 +134,7 @@ function admit() {
        run.id === artifact.workflow_run.id &&
        run.status === "completed" &&
        run.conclusion === "success" &&
-        run.path === ".github/workflows/ci.yml" &&
+        run.workflowPath?.file === ".github/workflows/ci.yml" &&
        ["schedule", "workflow_dispatch"].includes(run.event) &&
        run.head_sha === head &&
        run.head_branch === branch &&
@@ -30,7 +30,12 @@ if (outputs.tier === "full" && selectedScopes.length !== CI_FLAGS.length)
  throw new Error("Full qualification requires every canonical CI scope");
 const required = reused ? ["changes"] : ["changes", "static", "test"];
 const lanes = {
-  code: ["build", "test-shard"],
+  code: [
+    "build",
+    "test-shard",
+    ...(outputs.tier === "fast" ? ["test-source"] : []),
+  ],
+  reverse_tools: ["reverse-tools-quality"],
  runtime: ["windows-native-build", "package-e2e"],
  package: ["windows-curated"],
  sqlite: ["sqlite-runtime"],
@@ -2,6 +2,62 @@
 // Shared build inputs are classified once in plan.mjs rather than repeated in
 // every workflow.
 export const SCOPES = {
+  reverse_tools: [
+    "src/application/reverse/**",
+    "src/application/binary/BinarySession.ts",
+    "src/cutter/**",
+    "src/gdb/**",
+    "src/frida/frida-bindings.d.ts",
+    "src/objdump/**",
+    "src/cli/reverseEngineeringCommands.ts",
+    "src/process/ownedDebuggerSession.ts",
+    "src/rizin/**",
+    "src/contracts/*ToolContracts.ts",
+    "src/server/registerCutterTools.ts",
+    "src/server/externalMutationAdmission.ts",
+    "src/server/registerGdbTools.ts",
+    "src/server/registerReverseEngineeringTools.ts",
+    "src/server/providerEnvironment.ts",
+    "src/process/ExecFileOutput.ts",
+    "src/process/ExecFileOutput.test.ts",
+    "tests/composition/analysis-sessions/binarySession.test.ts",
+    "tests/boundary/filesystem/directAnalysisSnapshotReplay.test.ts",
+    "tests/boundary/filesystem/reverseToolsMutationConfig.test.ts",
+    "src/server/registerRizinDebugTools.ts",
+    "tests/boundary/providers/CutterBridge*.test.ts",
+    "tests/boundary/providers/cutterBridgeProducer.test.ts",
+    "tests/boundary/providers/GdbSession*.test.ts",
+    "tests/boundary/providers/GdbCommandQueue.test.ts",
+    "tests/boundary/providers/gdbTestHelpers.ts",
+    "tests/boundary/providers/RizinDebugSessionCoverage.part*.test.ts",
+    "tests/boundary/providers/RizinCommandTimeout.test.ts",
+    "tests/boundary/providers/RizinTimeoutEvidence.test.ts",
+    "tests/boundary/providers/RizinDebugSession*.test.ts",
+    "tests/boundary/providers/RizinInferiorTty.test.ts",
+    "tests/boundary/providers/RizinTestInferiorTty.ts",
+    "tests/boundary/process/ownedDebuggerSession.test.ts",
+    "tests/boundary/cli/gdbStartupCleanupCli.test.ts",
+    "tests/boundary/cli/reverseEngineeringCommands.test.ts",
+    "tests/boundary/mcp/cutter*.test.ts",
+    "tests/boundary/mcp/gdbCancellationMcp.test.ts",
+    "tests/boundary/mcp/debuggerCancellationMcp.test.ts",
+    "tests/boundary/mcp/rizinDebuggerCancellationMcp.test.ts",
+    "tests/boundary/mcp/reverseEngineeringPartialMcp.test.ts",
+    "tests/boundary/mcp/workflowSnapshotReplayMcp.test.ts",
+    "tests/boundary/mcp/reverseEngineeringEnvironmentMcp.test.ts",
+    "tests/support/reverseEngineering*.ts",
+    "tests/support/ownedRizinCommandEvidenceTests.ts",
+    "tests/support/reverseDebuggerMcpTestHelpers.ts",
+    "tests/processOwnershipRunner.ts",
+    "tests/composition/reverseEngineering*.test.ts",
+    "vitest.reverse-tools.config.ts",
+    "vitest.reverse-tools-semantics.config.ts",
+    "stryker.config.mjs",
+    "stryker.semantics.config.mjs",
+    "tsconfig.reverse-tools-mutation.json",
+    "scripts/prepare-reverse-tools-mutation.mjs",
+    "scripts/lib/reverse-tools-mutation-targets.mjs",
+  ],
  // Apple artifact CLI/MCP lane on macOS, including its focused regressions.
  apple: [
    "src/native/Apple*",
@@ -133,6 +133,8 @@ try {
    for (const clone of report.duplicates)
      for (const file of [clone.firstFile, clone.secondFile])
        file.name = relative(process.cwd(), file.name).split(sep).join("/");
+    // Record the actual checker result, including caller-selected thresholds.
+    report.reaCheck = { exitCode: result.status };
    await writeFile(reportPath, JSON.stringify(report, null, 2));
    const { sources, clones, percentage } = report.statistics.total;
    console.log(
@@ -0,0 +1,18 @@
+export interface StaticTaskContext {
+  readonly version: 1;
+  readonly sourceSha: string;
+  readonly sourceClean: boolean;
+  readonly prHeadSha?: string;
+  readonly nodeVersion: string;
+  readonly npmVersion: string;
+  readonly turboVersion: string;
+  readonly platform: string;
+  readonly architecture: string;
+  readonly taskHashes: Readonly<Record<string, string>>;
+}
+export function parseStaticTaskContext(value: unknown): StaticTaskContext;
+export function recordStaticTaskContext(options?: {
+  root?: string;
+  npmCli?: string;
+  outputDirectory?: string;
+}): Promise<StaticTaskContext>;
@@ -0,0 +1,104 @@
+import { execFile } from "node:child_process";
+import { mkdir, readFile, writeFile } from "node:fs/promises";
+import { join, resolve } from "node:path";
+import { fileURLToPath } from "node:url";
+import { promisify } from "node:util";
+import { z } from "zod";
+
+const execute = promisify(execFile);
+const contextSchema = z.object({
+  version: z.literal(1),
+  sourceSha: z.string().regex(/^[a-f0-9]{40}$/u),
+  sourceClean: z.boolean(),
+  prHeadSha: z
+    .string()
+    .regex(/^[a-f0-9]{40}$/u)
+    .optional(),
+  nodeVersion: z.string().min(1),
+  npmVersion: z.string().min(1),
+  turboVersion: z.string().min(1),
+  platform: z.string().min(1),
+  architecture: z.string().min(1),
+  taskHashes: z.record(z.string(), z.string().regex(/^[a-f0-9]{16}$/u)),
+});
+
+/** Parse the producing check graph's receipt; incomplete provenance is unknown. */
+export const parseStaticTaskContext = (value) => contextSchema.parse(value);
+
+/** Retain only named task hashes and tool/source identity, never the environment. */
+export async function recordStaticTaskContext({
+  root = process.cwd(),
+  npmCli = process.env.npm_execpath,
+  outputDirectory = join(root, ".cache/static-checks"),
+} = {}) {
+  if (typeof npmCli !== "string" || !npmCli)
+    throw new Error(
+      "Run static context recording through npm run ci:static-context",
+    );
+  const options = {
+    cwd: root,
+    encoding: "utf8",
+    timeout: 30_000,
+    maxBuffer: 32 * 1024 * 1024,
+  };
+  const [graph, source, status, npmVersion, turbo] = await Promise.all([
+    execute(
+      process.execPath,
+      [npmCli, "run", "--silent", "check:ci:tasks", "--", "--dry=json"],
+      options,
+    ),
+    execute("git", ["rev-parse", "HEAD"], options),
+    execute(
+      "git",
+      ["status", "--porcelain", "--untracked-files=normal"],
+      options,
+    ),
+    execute(process.execPath, [npmCli, "--version"], options),
+    readFile(join(root, "node_modules/turbo/package.json"), "utf8"),
+  ]);
+  const tasks = z
+    .object({
+      tasks: z
+        .array(
+          z.object({
+            task: z.string().min(1),
+            hash: z.string().regex(/^[a-f0-9]{16}$/u),
+          }),
+        )
+        .min(1),
+    })
+    .parse(JSON.parse(graph.stdout)).tasks;
+  if (new Set(tasks.map(({ task }) => task)).size !== tasks.length)
+    throw new Error("Static graph contains duplicate task identities");
+  const context = parseStaticTaskContext({
+    version: 1,
+    sourceSha: source.stdout.trim(),
+    sourceClean: status.stdout.length === 0,
+    ...(process.env.CI_PR_HEAD_SHA
+      ? { prHeadSha: process.env.CI_PR_HEAD_SHA }
+      : {}),
+    nodeVersion: process.version,
+    npmVersion: npmVersion.stdout.trim(),
+    turboVersion: z.object({ version: z.string() }).parse(JSON.parse(turbo))
+      .version,
+    platform: process.platform,
+    architecture: process.arch,
+    taskHashes: Object.fromEntries(tasks.map(({ task, hash }) => [task, hash])),
+  });
+  if (process.env.GITHUB_SHA && process.env.GITHUB_SHA !== context.sourceSha)
+    throw new Error(
+      "Static context source does not match the workflow revision",
+    );
+  await mkdir(outputDirectory, { recursive: true });
+  await writeFile(
+    join(outputDirectory, "static-task-context.json"),
+    JSON.stringify(context, null, 2) + "\n",
+  );
+  return context;
+}
+
+if (
+  process.argv[1] &&
+  resolve(process.argv[1]) === fileURLToPath(import.meta.url)
+)
+  await recordStaticTaskContext();
@@ -0,0 +1,46 @@
+import { spawn } from "node:child_process";
+import { resolve } from "node:path";
+
+import {
+  BUILT_TEST_PROJECTS,
+  SOURCE_TEST_PROJECTS,
+} from "../lib/test-projects.mjs";
+
+const [group, ...arguments_] = process.argv.slice(2);
+const projects = { source: SOURCE_TEST_PROJECTS, built: BUILT_TEST_PROJECTS }[
+  group
+];
+if (!projects) throw new Error("Expected source or built test group");
+// The group owns project selection; callers can request reporters or shards.
+if (
+  arguments_.some((argument) => /^--(?:project|exclude)(?:=|$)/u.test(argument))
+)
+  throw new Error("Test groups cannot override their project inventory");
+const child = spawn(
+  process.execPath,
+  [
+    resolve("node_modules/vitest/vitest.mjs"),
+    "run",
+    ...projects.flatMap(({ name }) => ["--project", name]),
+    ...arguments_,
+  ],
+  { stdio: "inherit" },
+);
+let shutdownSignal;
+for (const signal of ["SIGINT", "SIGTERM"])
+  process.on(signal, () => {
+    shutdownSignal ??= signal;
+    child.kill(signal);
+  });
+child.once("error", (error) => {
+  console.error(error.message);
+  process.exitCode = 1;
+});
+child.once("close", (code) => {
+  process.exitCode =
+    shutdownSignal === "SIGINT"
+      ? 130
+      : shutdownSignal === "SIGTERM"
+        ? 143
+        : (code ?? 1);
+});
@@ -2,6 +2,7 @@ import { execFileSync } from "node:child_process";
 import { appendFileSync } from "node:fs";
 import { resolve } from "node:path";
 import { fileURLToPath } from "node:url";
+import { normalizeWorkflowRun } from "./workflow-run.mjs";

 const ARTIFACT_NAME = "test-timings";
 const MAX_AGE_MS = 30 * 24 * 60 * 60 * 1000;
@@ -68,8 +69,8 @@ function findTimingRun({ repository, branch, repositoryId, now = Date.now() }) {
    .slice(0, MAX_CANDIDATES);

  for (const artifact of candidates) {
-    const run = api(
-      `repos/${repository}/actions/runs/${artifact.workflow_run.id}`,
+    const run = normalizeWorkflowRun(
+      api(`repos/${repository}/actions/runs/${artifact.workflow_run.id}`),
    );
    const runAge = now - Date.parse(run?.created_at);
    if (
@@ -80,7 +81,7 @@ function findTimingRun({ repository, branch, repositoryId, now = Date.now() }) {
      run.id === artifact.workflow_run.id &&
      run.status === "completed" &&
      run.conclusion === "success" &&
-      run.path === ".github/workflows/ci.yml" &&
+      run.workflowPath?.file === ".github/workflows/ci.yml" &&
      ["schedule", "workflow_dispatch"].includes(run.event) &&
      run.head_branch === branch &&
      run.head_repository?.full_name === repository
@@ -41,9 +41,20 @@ const files = report.files.map(({ name, durationMs, state }) => {
    throw new Error(`Test report path is outside the checkout: ${name}`);
  return { name: path.split(sep).join("/"), durationMs, state };
 });
-const complete = process.env.SHARD_RESULT === "success";
+// Keep group outcomes intact: an unavailable group is not a completed failure.
+const groupResults = [
+  process.env.SHARD_RESULT,
+  ...(process.env.SOURCE_REQUIRED === "true"
+    ? [process.env.SOURCE_RESULT]
+    : []),
+];
+const groupsCompleted = groupResults.every((state) =>
+  ["success", "failure"].includes(state),
+);
 const passed =
-  complete && process.env.MERGE_RESULT === "success" && report.success;
+  groupResults.every((state) => state === "success") &&
+  process.env.MERGE_RESULT === "success" &&
+  report.success;
 // Successful shards and merged reports must account for the complete portable
 // inventory. Coverage thresholds alone cannot prove that every file ran.
 const verifyInventory = process.env.CI_VERIFY_TEST_INVENTORY === "true";
@@ -69,7 +80,7 @@ if (process.env.CI_TEST_SELECTION) {
      comparison.complete &&
      report.unhandledErrors === 0 &&
      ["passed", "failed"].includes(report.reason) &&
-      ["success", "failure"].includes(process.env.SHARD_RESULT) &&
+      groupsCompleted &&
      ["success", "failure"].includes(process.env.MERGE_RESULT),
    executionSha: process.env.GITHUB_SHA ?? null,
    prHeadSha: process.env.CI_PR_HEAD_SHA ?? null,
@@ -91,7 +102,9 @@ if (passed && verifyInventory) {
    );
 }
 const summary = [
-  complete ? "## Merged test diagnostics" : "## Partial test diagnostics",
+  groupsCompleted
+    ? "## Merged test diagnostics"
+    : "## Partial test diagnostics",
  "",
  ...(selection
    ? [
@@ -99,11 +112,14 @@ const summary = [
        "",
        `Plan: ${selection.plan.mode}. Proposed files: ${selection.selected.length}; omitted: ${selection.omitted.length}.`,
        `Full-run comparison: ${selection.complete ? "complete" : "incomplete"}; failures outside proposed selection: ${selection.missedFailures.length}.`,
-        "The complete suite still ran. This report cannot enable omissions or qualify a revision. Skipped, missing, and interrupted observations do not establish parity.",
+        "This observational report cannot enable omissions or qualify a revision. Skipped, missing, and interrupted observations do not establish parity.",
        "",
      ]
    : []),
  `Shard result: ${process.env.SHARD_RESULT ?? "unknown"}; merge result: ${process.env.MERGE_RESULT ?? "unknown"}.`,
+  ...(process.env.SOURCE_REQUIRED === "true"
+    ? [`Source result: ${process.env.SOURCE_RESULT ?? "unknown"}.`]
+    : []),
  passed
    ? "Successful timing history is available."
    : "This report is diagnostic only and is not a successful timing baseline.",
@@ -0,0 +1,10 @@
+export interface WorkflowPath {
+  readonly file: string;
+  readonly ref: string | null;
+}
+export interface NormalizedWorkflowRun {
+  readonly workflowPath: WorkflowPath | null;
+}
+export function normalizeWorkflowRun<T extends Record<string, unknown>>(
+  value: T,
+): Omit<T, "workflowPath"> & NormalizedWorkflowRun;
@@ -0,0 +1,15 @@
+/** Keep the API path as evidence and parse its optional ref at the boundary. */
+export function normalizeWorkflowRun(value) {
+  if (typeof value !== "object" || value === null || Array.isArray(value))
+    throw new Error("workflow_run_malformed");
+  const match =
+    typeof value.path === "string" && !/[\u0000-\u001f\u007f]/u.test(value.path)
+      ? /^(\.github\/workflows\/[^/\\@\s]+\.ya?ml)(?:@([^\s]+))?$/u.exec(
+          value.path,
+        )
+      : null;
+  return {
+    ...value,
+    workflowPath: match ? { file: match[1], ref: match[2] ?? null } : null,
+  };
+}
@@ -5,6 +5,7 @@ import { fileURLToPath } from "node:url";
 import { promisify } from "node:util";
 import { TraceMap, eachMapping } from "@jridgewell/trace-mapping";
 import { CdpBrowserProvider } from "../../dist/browser/CdpBrowserProvider.js";
+import { listBrowserTargetsInputSchema } from "../../dist/domain/browserObservation.js";
 import { CdpConnection } from "../../dist/browser/CdpConnection.js";
 import { parseEvidence } from "../../dist/domain/evidence.js";
 import { traceSourceMap } from "../../dist/javascript/sourceMaps/TraceSourceMap.js";
@@ -89,10 +90,12 @@ async function verifyCase(executable, entrypoint, scenario) {
    fixture = await startRuntimeFixtureBrowser(executable, site.origin);
    const endpoint = `http://127.0.0.1:${fixture.port}`;
    const provider = new CdpBrowserProvider();
-    const targets = await provider.listTargets({
-      cdp_endpoint: endpoint,
-      allowed_origins: [site.origin],
-    });
+    const targets = await provider.listTargets(
+      listBrowserTargetsInputSchema.parse({
+        cdp_endpoint: endpoint,
+        allowed_origins: [site.origin],
+      }),
+    );
    if (!targets.ok) throw targets.error;
    const target = targets.value.targets[0];
    assert.ok(target, "Owned fixture target missing");
@@ -8,6 +8,7 @@ import { promisify } from "node:util";
 import { Client } from "@modelcontextprotocol/client";
 import { StdioClientTransport } from "@modelcontextprotocol/client/stdio";
 import { CdpBrowserProvider } from "../../dist/browser/CdpBrowserProvider.js";
+import { listBrowserTargetsInputSchema } from "../../dist/domain/browserObservation.js";
 import { CdpConnection } from "../../dist/browser/CdpConnection.js";
 import { parseEvidence } from "../../dist/domain/evidence.js";
 import { webExecutionSchema } from "../../dist/domain/webExecution.js";
@@ -45,10 +46,12 @@ export async function verifyBrowserRuntime(
  try {
    fixture = await startRuntimeFixtureBrowser(executable, site.origin);
    const endpoint = `http://127.0.0.1:${fixture.port}`;
-    const targets = await new CdpBrowserProvider().listTargets({
-      cdp_endpoint: endpoint,
-      allowed_origins: [site.origin],
-    });
+    const targets = await new CdpBrowserProvider().listTargets(
+      listBrowserTargetsInputSchema.parse({
+        cdp_endpoint: endpoint,
+        allowed_origins: [site.origin],
+      }),
+    );
    if (!targets.ok) throw targets.error;
    const target = targets.value.targets[0];
    assert.ok(target, "owned fixture page target missing");
@@ -19,6 +19,11 @@ export const toolFamilyCatalog = (sources) => {
      contracts: [
        ...sources.nativeContracts.NATIVE_TOOL_CONTRACTS,
        ...sources.binaryDiagnosticsContracts.BINARY_DIAGNOSTICS_TOOL_CONTRACTS,
+        ...sources.cutterContracts.CUTTER_TOOL_CONTRACTS,
+        ...sources.reverseEngineeringContracts
+          .REVERSE_ENGINEERING_TOOL_CONTRACTS,
+        ...sources.gdbContracts.GDB_TOOL_CONTRACTS,
+        ...sources.rizinDebugContracts.RIZIN_DEBUG_TOOL_CONTRACTS,
      ],
    },
    {
@@ -164,6 +169,32 @@ export const providerCatalog = (sources) => {
    activeContracts,
  } = electronContractSlices(sources);
  return [
+    {
+      identity: sources.cutterProvider.CUTTER_PROVIDER_IDENTITY,
+      contracts: sources.cutterContracts.CUTTER_TOOL_CONTRACTS,
+    },
+    {
+      identity: sources.objdumpProvider.OBJDUMP_PROVIDER_IDENTITY,
+      contracts:
+        sources.reverseEngineeringContracts.REVERSE_ENGINEERING_TOOL_CONTRACTS.filter(
+          ({ name }) => name === "inspect_with_objdump",
+        ),
+    },
+    {
+      identity: sources.rizinProvider.RIZIN_PROVIDER_IDENTITY,
+      contracts:
+        sources.reverseEngineeringContracts.REVERSE_ENGINEERING_TOOL_CONTRACTS.filter(
+          ({ name }) => name === "execute_rizin_command",
+        ),
+    },
+    {
+      identity: sources.gdbProvider.GDB_PROVIDER_IDENTITY,
+      contracts: sources.gdbContracts.GDB_TOOL_CONTRACTS,
+    },
+    {
+      identity: sources.rizinDebugProvider.RIZIN_DEBUGGER_PROVIDER_IDENTITY,
+      contracts: sources.rizinDebugContracts.RIZIN_DEBUG_TOOL_CONTRACTS,
+    },
    {
      identity: sources.fridaProvider.FRIDA_PROVIDER_IDENTITY,
      contracts: sources.fridaContracts.FRIDA_TOOL_CONTRACTS,
@@ -43,6 +43,16 @@ const SOURCE_PATHS = {
  sessionContracts: "dist/contracts/sessionToolContracts.js",
  binaryDiagnosticsContracts:
    "dist/contracts/native/binaryDiagnosticsToolContracts.js",
+  cutterContracts: "dist/contracts/cutterToolContracts.js",
+  cutterProvider: "dist/cutter/CutterBridgeClient.js",
+  reverseEngineeringContracts:
+    "dist/contracts/reverseEngineeringToolContracts.js",
+  gdbContracts: "dist/contracts/gdbToolContracts.js",
+  gdbProvider: "dist/gdb/GdbSessionManager.js",
+  rizinDebugContracts: "dist/contracts/rizinDebugToolContracts.js",
+  objdumpProvider: "dist/objdump/ObjdumpCommand.js",
+  rizinProvider: "dist/rizin/RizinCommand.js",
+  rizinDebugProvider: "dist/rizin/RizinDebugSessionManager.js",
  fridaContracts: "dist/contracts/fridaToolContracts.js",
  fridaProvider: "dist/frida/FridaInstrumentationManager.js",
  pwntoolsProvider: "dist/native/pwntools/PwntoolsRelease.js",
@@ -1,17 +1,4 @@
-const sourceProjects = [
-  "domain",
-  "services",
-  "adapters",
-  "composition",
-  "conformance",
-  "evaluation",
-];
-
-const sourceTest = (path) =>
-  path.startsWith("src/") ||
-  ["composition", "conformance", "evaluation"].some((project) =>
-    path.startsWith(`tests/${project}/`),
-  );
+import { SOURCE_TEST_PROJECTS, projectForTest } from "./test-projects.mjs";

 /** Parse explicit test paths and an optional branch-comparison base. */
 export const parseDevelopmentTestRequest = (mode, arguments_) => {
@@ -35,10 +22,9 @@ export const parseDevelopmentTestRequest = (mode, arguments_) => {
        path.split("/").some((part) => part === ".." || part === ".")
      )
        throw new Error(`Expected an exact repository test path: ${argument}`);
-      if (mode !== "focused" && !sourceTest(path))
-        throw new Error(
-          `Use test:focused for compiled boundary tests: ${path}`,
-        );
+      const project = projectForTest(path);
+      if (mode !== "focused" && project.group !== "source")
+        throw new Error(`Use test:focused for built test projects: ${path}`);
      paths.push(path);
    }
  }
@@ -72,9 +58,14 @@ export const developmentTestPlan = (request, baseCommit) => {
    !/^[a-f0-9]{40,64}$/u.test(baseCommit ?? "")
  )
    throw new Error("Changed tests require a resolved Git merge base");
-  const projects = request.mode === "focused" ? [] : sourceProjects;
+  const projects =
+    request.mode === "focused"
+      ? []
+      : SOURCE_TEST_PROJECTS.map(({ name }) => name);
  return {
-    needsBuild: request.paths.some((path) => !sourceTest(path)),
+    needsBuild: request.paths.some(
+      (path) => projectForTest(path).group === "built",
+    ),
    artifactTasks: Object.entries(artifactConsumers).flatMap(
      ([task, consumers]) =>
        request.paths.some((path) => consumers.includes(path)) ? [task] : [],
@@ -215,6 +215,27 @@ async function verifyNativeValueAdapters(
    const functionEvidence = parseEvidence(analyzed.structuredContent);
    assert.equal(functionEvidence.operation, "analyze_function");
    assert.equal(functionEvidence.provider.id, "ghidra");
+    assert.deepEqual(
+      functionEvidence.normalized_result.facet_availability,
+      Object.fromEntries(
+        [
+          "pseudocode",
+          "assembly",
+          "comments",
+          "callers",
+          "callees",
+          "references",
+          "strings_names",
+          "cfg",
+        ].map((facet) => [facet, { available: true }]),
+      ),
+    );
+    assert.equal(functionEvidence.relationship_coverage?.status, "complete");
+    assert.ok(
+      functionEvidence.relationship_coverage.diagnostics.some((diagnostic) =>
+        diagnostic.includes("indirect calls remain unknown"),
+      ),
+    );
    const functionProfile = analysisProfileSchema.parse(
      functionEvidence.analysis_profile,
    );
@@ -235,7 +256,11 @@ async function verifyNativeValueAdapters(
    );
    assert.notEqual(comparison.isError, true, JSON.stringify(comparison));
    const comparisonEvidence = parseEvidence(comparison.structuredContent);
-    assert.equal(comparisonEvidence.normalized_result.status, "unchanged");
+    assert.equal(
+      comparisonEvidence.normalized_result.status,
+      "unchanged",
+      JSON.stringify(comparisonEvidence.normalized_result.dimensions),
+    );
    assert.deepEqual(
      comparisonEvidence.normalized_result,
      comparison.structuredContent?.normalized_result,
@@ -243,6 +268,7 @@ async function verifyNativeValueAdapters(
    assert.ok(
      comparisonEvidence.evidence_links.includes(functionEvidence.evidence_id),
    );
+    await verifyUnresolvedCallEvidence(client);
    const bundle = await client.callTool({
      name: "get_evidence_bundle",
      arguments: {},
@@ -303,3 +329,63 @@ async function verifyNativeValueAdapters(
    unknowns: cliGraph.unknowns.length,
  };
 }
+
+async function verifyUnresolvedCallEvidence(client) {
+  const inventory = await client.callTool({
+    name: "list_procedures",
+    arguments: {},
+  });
+  assert.notEqual(inventory.isError, true, JSON.stringify(inventory));
+  const procedures = parseEvidence(
+    inventory.structuredContent,
+  ).normalized_result;
+  const indirect = procedures.find((item) =>
+    item.value.endsWith("rea_ghidra_inventory_indirect"),
+  );
+  const leaf = procedures.find((item) =>
+    item.value.endsWith("rea_ghidra_inventory_leaf"),
+  );
+  assert.ok(
+    indirect && leaf,
+    "Source fixture lacks the unresolved callback and its possible target",
+  );
+  const analyzed = await client.callTool(
+    { name: "analyze_function", arguments: { procedure: indirect.address } },
+    { timeout: 180000 },
+  );
+  assert.notEqual(analyzed.isError, true, JSON.stringify(analyzed));
+  const evidence = parseEvidence(analyzed.structuredContent);
+  assert.equal(evidence.relationship_coverage?.status, "partial");
+  assert.ok(evidence.normalized_result.unresolved_calls.length > 0);
+  const traced = await client.callTool(
+    {
+      name: "trace_call_path",
+      arguments: {
+        start: indirect.address,
+        goal: leaf.address,
+        direction: "forward",
+      },
+    },
+    { timeout: 180000 },
+  );
+  assert.notEqual(traced.isError, true, JSON.stringify(traced));
+  const trace = parseEvidence(traced.structuredContent).normalized_result;
+  assert.equal(trace.goal_status, "unknown");
+  assert.equal(trace.truncated, true);
+  assert.ok(trace.residual_unknowns.length > 0);
+  const compared = await client.callTool(
+    {
+      name: "compare_functions",
+      arguments: {
+        left: analyzed.structuredContent,
+        right: analyzed.structuredContent,
+      },
+    },
+    { timeout: 180000 },
+  );
+  assert.notEqual(compared.isError, true, JSON.stringify(compared));
+  assert.equal(
+    parseEvidence(compared.structuredContent).normalized_result.status,
+    "unknown",
+  );
+}
@@ -131,6 +131,25 @@ export async function verifyGhidraBoundaries(
  const externalDossier = await call("analyze_function", {
    procedure: external.address,
  });
+  assert.equal(externalDossier.pseudocode, "");
+  assert.deepEqual(externalDossier.assembly, []);
+  assert.deepEqual(externalDossier.callees, []);
+  assert.deepEqual(externalDossier.facet_availability.pseudocode, {
+    available: false,
+    reason: "Ghidra does not decompile external functions.",
+  });
+  for (const facet of [
+    "assembly",
+    "comments",
+    "callers",
+    "callees",
+    "references",
+    "strings_names",
+    "cfg",
+  ])
+    assert.deepEqual(externalDossier.facet_availability[facet], {
+      available: true,
+    });
  const externalEntry = await call("resolve_containing_procedure", {
    address: external.address,
  });
@@ -10,6 +10,25 @@ interface StringExpectation {
  readonly recognitionOptional?: boolean;
 }

+/** Verify native fragments against a complete, independently read source extent. */
+export function verifyHopperStringFragments(
+  call: VerifierCall,
+  inventory: readonly z.output<typeof analysisStringSchema>[],
+  expectation: {
+    readonly value: string;
+    readonly symbol: string;
+    readonly recognitionOptional?: boolean;
+  },
+): Promise<{
+  readonly sourceBytes: number;
+  readonly recognition: {
+    readonly address: string;
+    readonly state: "observed" | "not_observed";
+    readonly decodedObject: "verified" | "unverified";
+  };
+  readonly fragments: readonly z.output<typeof analysisStringSchema>[];
+}>;
+
 /** Verify source bytes, observed decoding, address filters and literal searches. */
 export function verifyHopperStringObjects(
  call: VerifierCall,
@@ -106,27 +106,18 @@ async function verifySearchModes(call, address) {
    [{ address, value: "REA_Search_Upper" }],
  );
  const strings = await call("list_strings");
-  const linux = process.platform === "linux";
-  const fullLiteral = "REA_LONG_LITERAL_" + "x".repeat(4096) + "needle";
-  // Recognition is a provider/loader capability, not an adapter guarantee.
-  // Always retain the original size in raw-byte coverage; Linux additionally
-  // requires an extent observed to produce native fragmented string objects.
-  const fullBytes = Buffer.from(fullLiteral + "\0");
-  const raw = await call("read_bytes", {
-    address: `${fixtureSymbolPrefix}rea_long_literal`,
-    length: fullBytes.length,
+  const fullLiteral = await verifyHopperStringFragments(call, strings, {
+    symbol: `${fixtureSymbolPrefix}rea_long_literal`,
+    value: "REA_LONG_LITERAL_" + "x".repeat(4096) + "needle",
+    recognitionOptional: true,
  });
-  assert.equal(raw.complete, true);
-  assert.equal(raw.bytes_hex, fullBytes.toString("hex"));
-  const long = linux
-    ? "REA_FRAGMENTED_LITERAL_" + "x".repeat(512) + "fragment_tail"
-    : fullLiteral;
-  const longItem = strings.find(
-    (item) => long.startsWith(item.value) && item.value.startsWith("REA_"),
-  );
-  assert.ok(longItem, "fixture omitted its real long literal");
-  if (!linux) assert.equal(longItem.address, raw.address);
-  assert.ok(long.startsWith(longItem.value));
+  const fragmentedLiteral = await verifyHopperStringFragments(call, strings, {
+    symbol: `${fixtureSymbolPrefix}rea_fragmented_literal`,
+    value: "REA_FRAGMENTED_LITERAL_" + "x".repeat(512) + "fragment_tail",
+  });
+  const [longItem] = fragmentedLiteral.fragments;
+  assert.ok(longItem, "fixture omitted its typed fragmented literal");
+  assert.ok(fragmentedLiteral.fragments.length > 1);
  assert.ok(longItem.provider_value.endsWith("…"));
  assert.equal(longItem.string.termination, "missing");
  const literalProcedure = (await call("list_procedures")).find((item) =>
@@ -154,33 +145,13 @@ async function verifySearchModes(call, address) {
    "function dossier discarded typed string provenance",
  );

-  const fragments = strings.filter((item) => {
-    const offset = BigInt(item.address) - BigInt(longItem.address);
-    return offset >= 0n && offset < BigInt(Buffer.byteLength(long));
-  });
-  assert.equal(fragments.map((item) => item.value).join(""), long);
-  for (const fragment of fragments) {
-    assert.equal(
-      Buffer.byteLength(fragment.value),
-      fragment.string.byte_length -
-        (fragment.string.termination === "missing" ? 0 : 1),
-    );
-    const bytes = await call("read_bytes", {
-      address: fragment.address,
-      length: fragment.string.byte_length,
-    });
-    const expected = Buffer.from(
-      fragment.value + (fragment.string.termination === "missing" ? "" : "\0"),
-    );
-    assert.equal(bytes.bytes_hex, expected.toString("hex"));
-  }
  assert.deepEqual(
    await call("search_strings", {
-      pattern: linux ? "fragment_tail" : "needle",
+      pattern: "fragment_tail",
      mode: "literal",
      case_sensitive: true,
    }),
-    [fragments.at(-1)],
+    [fragmentedLiteral.fragments.at(-1)],
  );
  const expected = strings.filter((item) =>
    /^REA_C_(?:ENTRY|LEAF)$/u.test(item.value),
@@ -201,7 +172,7 @@ async function verifySearchModes(call, address) {
        value: "REA_UTF8_é_😀",
        encoding: "utf-8",
        symbol: `${fixtureSymbolPrefix}rea_utf8_literal`,
-        recognitionOptional: linux,
+        recognitionOptional: true,
      },
      { value: 'REA_ESCAPED_"\\line\nend\t\r', encoding: "utf-8" },
      { value: "REA_LITERAL_BACKSLASH_\\n", encoding: "utf-8" },
@@ -209,33 +180,90 @@ async function verifySearchModes(call, address) {
        value: "REA_LITERAL_…",
        encoding: "utf-8",
        symbol: `${fixtureSymbolPrefix}rea_ellipsis_literal`,
-        recognitionOptional: linux,
+        recognitionOptional: true,
      },
      {
        value: "REA_LATIN1_ÿ",
        encoding: "latin-1",
        symbol: `${fixtureSymbolPrefix}rea_latin1_literal`,
-        recognitionOptional: linux,
+        recognitionOptional: true,
      },
    ],
    strings,
  );
  return {
-    fullLiteralSourceBytes: fullBytes.length,
-    fragmentedLiteralSourceBytes: Buffer.byteLength(long),
-    fullLiteralRecognition: {
-      address: raw.address,
-      state: strings.some((item) => item.address === raw.address)
-        ? "observed"
-        : "not_observed",
-      // Linux validates the original through raw bytes; its required decoded
-      // fragment/dossier checks exercise the separate supported-size literal.
-      decodedObject: linux ? "unverified" : "verified",
-    },
+    fullLiteralSourceBytes: fullLiteral.sourceBytes,
+    fragmentedLiteralSourceBytes: fragmentedLiteral.sourceBytes,
+    fullLiteralRecognition: fullLiteral.recognition,
    ...objects,
  };
 }

+/** Bind native fragments to exact source bytes without assuming loader recognition. */
+export async function verifyHopperStringFragments(
+  call,
+  strings,
+  { symbol, value, recognitionOptional = false },
+) {
+  const expectedBytes = Buffer.from(value + "\0");
+  const raw = await call("read_bytes", {
+    address: symbol,
+    length: expectedBytes.length,
+  });
+  assert.equal(raw.complete, true);
+  assert.equal(raw.bytes_hex, expectedBytes.toString("hex"));
+  const start = BigInt(raw.address);
+  const end = start + BigInt(expectedBytes.length);
+  const fragments = strings.filter((item) => {
+    const address = BigInt(item.address);
+    return address >= start && address < end;
+  });
+  if (fragments.length === 0 && recognitionOptional)
+    return {
+      sourceBytes: expectedBytes.length,
+      recognition: {
+        address: raw.address,
+        state: "not_observed",
+        decodedObject: "unverified",
+      },
+      fragments,
+    };
+  assert.ok(fragments.length > 0, "fixture omitted its typed string fragments");
+  let next = start;
+  for (const [index, fragment] of fragments.entries()) {
+    assert.equal(
+      BigInt(fragment.address),
+      next,
+      "string fragments have a gap, overlap or wrong source identity",
+    );
+    const terminal = index === fragments.length - 1;
+    assert.equal(
+      fragment.string.termination,
+      terminal ? "present_or_not_required" : "missing",
+    );
+    const expected = Buffer.from(fragment.value + (terminal ? "\0" : ""));
+    assert.equal(fragment.string.byte_length, expected.length);
+    const bytes = await call("read_bytes", {
+      address: fragment.address,
+      length: expected.length,
+    });
+    assert.equal(bytes.complete, true);
+    assert.equal(bytes.bytes_hex, expected.toString("hex"));
+    next += BigInt(expected.length);
+  }
+  assert.equal(next, end, "string fragments do not cover the source extent");
+  assert.equal(fragments.map((item) => item.value).join(""), value);
+  return {
+    sourceBytes: expectedBytes.length,
+    recognition: {
+      address: raw.address,
+      state: "observed",
+      decodedObject: "verified",
+    },
+    fragments,
+  };
+}
+
 /** Compare native string identities and values with the source fixture's bytes. */
 export async function verifyHopperStringObjects(
  call,
@@ -327,14 +327,57 @@ export const analyzeFixtureProcedure = async (
  options,
  procedure,
  normalizedResult,
-) =>
-  normalizedResult(
+) => {
+  const reply = await client.callTool(
+    {
+      name: "analyze_function",
+      arguments: { procedure: procedure.address },
+    },
+    options,
+  );
+  const dossier = normalizedResult(reply, "analyze_function");
+  assert.deepEqual(reply.structuredContent.relationship_coverage, {
+    status: "complete",
+  });
+  for (const facet of [
+    "pseudocode",
+    "assembly",
+    "comments",
+    "callers",
+    "callees",
+    "references",
+    "strings_names",
+    "cfg",
+  ])
+    assert.deepEqual(dossier.facet_availability[facet], { available: true });
+  const comparison = normalizedResult(
    await client.callTool(
      {
-        name: "analyze_function",
-        arguments: { procedure: procedure.address },
+        name: "compare_functions",
+        arguments: {
+          left: reply.structuredContent,
+          right: reply.structuredContent,
+        },
      },
      options,
    ),
-    "analyze_function",
+    "compare_functions",
  );
+  assert.deepEqual(
+    Object.fromEntries(
+      comparison.dimensions.map(({ dimension, status }) => [dimension, status]),
+    ),
+    {
+      identity: "unchanged",
+      pseudocode: "unchanged",
+      assembly: "unchanged",
+      comments: "unchanged",
+      calls: "unchanged",
+      references: "unknown",
+      strings_names: "unchanged",
+      cfg: "unchanged",
+    },
+    JSON.stringify(comparison.dimensions),
+  );
+  return dossier;
+};
@@ -0,0 +1,182 @@
+import { readFileSync } from "node:fs";
+import ts from "typescript6";
+
+const compact = (text) => text.replace(/\s+/gu, "");
+const nameOf = (node, source) => node.name?.getText(source);
+
+const methodOwner = (node, source) => {
+  for (let parent = node; parent !== undefined; parent = parent.parent) {
+    if (ts.isMethodDeclaration(parent) || ts.isConstructorDeclaration(parent))
+      return {
+        owner: nameOf(parent.parent, source),
+        body: parent.body,
+        method: ts.isConstructorDeclaration(parent)
+          ? "constructor"
+          : nameOf(parent, source),
+      };
+  }
+  return undefined;
+};
+
+const withinOwner = (node, selector, source) => {
+  if (selector.owner === undefined) return true;
+  const owner = methodOwner(node, source);
+  return (
+    owner?.owner === selector.owner &&
+    (selector.method === undefined || owner.method === selector.method)
+  );
+};
+
+const contains = (node, predicate) =>
+  predicate(node) ||
+  ts.forEachChild(node, (child) => contains(child, predicate)) === true;
+
+const inBlockWithBinding = (node, binding) => {
+  if (binding === undefined) return true;
+  let block = node.parent;
+  while (block !== undefined && !ts.isBlock(block)) block = block.parent;
+  return block?.statements.some(
+    (statement) =>
+      ts.isVariableStatement(statement) &&
+      statement.declarationList.declarations.some(
+        (declaration) =>
+          ts.isIdentifier(declaration.name) &&
+          declaration.name.text === binding,
+      ),
+  );
+};
+
+const isCommandHandler = (node, command) => {
+  if (!ts.isArrowFunction(node)) return false;
+  const property = node.parent;
+  if (
+    !ts.isPropertyAssignment(property) ||
+    !ts.isIdentifier(property.name) ||
+    property.name.text !== "run"
+  )
+    return false;
+  const call = property.parent.parent;
+  if (
+    !ts.isCallExpression(call) ||
+    !ts.isPropertyAccessExpression(call.expression) ||
+    !ts.isIdentifier(call.expression.expression) ||
+    call.expression.expression.text !== "cli" ||
+    call.expression.name.text !== "command"
+  )
+    return false;
+  const selected = call.arguments[0];
+  return (
+    selected !== undefined &&
+    ts.isPropertyAccessExpression(selected) &&
+    ts.isIdentifier(selected.expression) &&
+    selected.expression.text === "CLI_COMMANDS" &&
+    selected.name.text === command
+  );
+};
+
+const matches = (node, selector, source) => {
+  if (!withinOwner(node, selector, source)) return false;
+  if (
+    selector.direct === true &&
+    node.parent !== methodOwner(node, source)?.body
+  )
+    return false;
+  switch (selector.kind) {
+    case "expression":
+      return (
+        ts.isBinaryExpression(node) &&
+        compact(node.getText(source)) === compact(selector.text)
+      );
+    case "declaration":
+      return (
+        ts.isVariableDeclaration(node) &&
+        nameOf(node, source) === selector.name &&
+        inBlockWithBinding(node, selector.withinBlockWithBinding)
+      );
+    case "method":
+      return (
+        ts.isMethodDeclaration(node) && nameOf(node, source) === selector.name
+      );
+    case "constructor":
+      return ts.isConstructorDeclaration(node);
+    case "guard":
+      return (
+        ts.isIfStatement(node) &&
+        compact(node.expression.getText(source)) === compact(selector.condition)
+      );
+    case "call":
+      return (
+        ts.isCallExpression(node) &&
+        compact(node.expression.getText(source)) === compact(selector.callee) &&
+        (selector.event === undefined ||
+          (node.arguments[0] !== undefined &&
+            ts.isStringLiteral(node.arguments[0]) &&
+            node.arguments[0].text === selector.event))
+      );
+    case "try":
+      return (
+        ts.isTryStatement(node) &&
+        contains(node.tryBlock, (child) =>
+          selector.assignment !== undefined
+            ? ts.isBinaryExpression(child) &&
+              child.operatorToken.kind === ts.SyntaxKind.EqualsToken &&
+              ts.isIdentifier(child.left) &&
+              child.left.text === selector.assignment
+            : ts.isCallExpression(child) &&
+              compact(child.expression.getText(source)) ===
+                compact(selector.callee),
+        )
+      );
+    case "command-handler":
+      return isCommandHandler(node, selector.command);
+    default:
+      throw new Error(
+        `Unknown mutation selector kind: ${String(selector.kind)}`,
+      );
+  }
+};
+
+/** Parse a source once and resolve one owned region for every requested selector. */
+export const mutationRanges = (path, selectors) => {
+  if (!Array.isArray(selectors) || selectors.length === 0)
+    throw new Error(`Mutation selectors must be nonempty: ${path}`);
+  for (const selector of selectors)
+    if (
+      ["method", "constructor", "guard", "call", "try"].includes(
+        selector.kind,
+      ) &&
+      typeof selector.owner !== "string"
+    )
+      throw new Error(
+        `Mutation selector requires a named class owner: ${path}`,
+      );
+  const source = ts.createSourceFile(
+    path,
+    readFileSync(path, "utf8"),
+    ts.ScriptTarget.Latest,
+    true,
+  );
+  if (source.parseDiagnostics.length !== 0)
+    throw new Error(
+      `Cannot select mutations from malformed TypeScript: ${path}`,
+    );
+  const found = selectors.map(() => []);
+  const visit = (node) => {
+    for (const [index, selector] of selectors.entries())
+      if (matches(node, selector, source)) found[index].push(node);
+    ts.forEachChild(node, visit);
+  };
+  visit(source);
+  return selectors.map((selector, index) => {
+    const candidates = found[index];
+    if (candidates.length !== 1)
+      throw new Error(
+        `Expected one mutation target in ${path}: ${JSON.stringify(selector)}; found ${candidates.length}`,
+      );
+    const node = candidates[0];
+    const first =
+      source.getLineAndCharacterOfPosition(node.getStart(source)).line + 1;
+    const last = source.getLineAndCharacterOfPosition(node.getEnd()).line + 1;
+    return `${path}:${first}-${last}`;
+  });
+};
--- a/Show More
+++ b/Show More