feat(repo): agent-ready failure diagnostics (stack #146-#156)

.changeset/busy-mammals-clean.md

@@ -0,0 +1,5 @@
+---
+"@systemfsoftware/stryker-js-vitest-runner": minor
+---
+
+Every failed test reports a reproduce argv: vitest run <file> -t <full test name>, or vitest run <file> for a file-level failure.

.changeset/fine-pandas-sit.md

@@ -0,0 +1,5 @@
+---
+"@systemfsoftware/stryker-js": minor
+---
+
+Each failure record carries a one-step reproduction: the failing test's own runner command under the mutation-worker budget, stryker run --dryRunOnly when the runner gives none, stryker run --mutant <ids> for new survivors, or DoesNotReplay naming why and what stands in (worker kind, pid and exit code for crashed workers).

.changeset/green-kings-think.md

@@ -0,0 +1,5 @@
+---
+"@systemfsoftware/stryker-js": major
+---
+
+Every failed run ends in one FailureRecord: printed on stderr, emitted as the terminal stream event, and written to reports/mutation/failure.json. StageError and PrepareError are replaced by RunFailure, exit codes come from the failure catalog (a failing baseline test exits 5), failed baseline tests are named by project-relative file and line, and the cli.run span carries stryker.failure.code.

.changeset/humble-coats-teach.md

@@ -0,0 +1,5 @@
+---
+"@systemfsoftware/stryker-js": minor
+---
+
+A failed run's record reaches every surface: the Mutation Server Protocol answers engine failures with the record in error.data and keeps serving, MCP rerun_mutant returns a RerunEngineUnusable refusal carrying it and the new get_failure tool reads reports/mutation/failure.json, a configured sarif reporter writes a SARIF log whose invocation is executionSuccessful false with one notification per record, and the run log no longer prints an orphan 'Possible causes' block or the bundled parser's WASI ExperimentalWarning.

.changeset/moody-snails-watch.md

@@ -0,0 +1,5 @@
+---
+"@systemfsoftware/stryker-js-cli-contract": minor
+---
+
+RunFailed now carries a typed FailureRecord (catalog code, stage, structured evidence, cause chain, next action, replay capsule, trace id) instead of error/remediation/reason prose; the stream schema is 3.0. The package publishes contract/failure-catalog.json and FailureRecord renderers for terminal, markdown, GitHub annotations and SARIF.

.changeset/petite-insects-feel.md

@@ -0,0 +1,5 @@
+---
+"@systemfsoftware/stryker-js": minor
+---
+
+A dry-run-only run with incremental mode on publishes its dry-run coverage to the incremental file, keeping any verdicts already there, so a later run that reads the file through incrementalSources skips its own dry run. The run-inputs digest no longer covers dryRunOnly, so a preflight and the runs that reuse its coverage share one digest; existing caches are refused once with runInputsChanged.

.changeset/proud-bananas-work.md

@@ -0,0 +1,5 @@
+---
+"@systemfsoftware/stryker-js": minor
+---
+
+stryker annotate also renders the GitHub annotations of reports/mutation/failure.json when a run failed.

.changeset/quiet-clowns-remain.md

@@ -0,0 +1,5 @@
+---
+"@systemfsoftware/stryker-js-plugin-interface": minor
+---
+
+FailedTestResult gains an optional reproduce argv: the command that re-runs exactly that failing test.

.changeset/sixty-wombats-write.md

@@ -0,0 +1,5 @@
+---
+"@systemfsoftware/stryker-js-cli-contract": minor
+---
+
+FailedTestEvidence carries the runner's reproduce argv, and recordOf takes the decided capsule from its RecordContext instead of deriving it.

.changeset/small-results-greet.md

@@ -0,0 +1,5 @@
+---
+"@systemfsoftware/stryker-js-plugin-interface": major
+---
+
+FailedTestResult carries an optional project-relative source location and the raw stack; the unused startPosition field is removed.

.changeset/upset-chicken-hammer.md

@@ -0,0 +1,5 @@
+---
+"@systemfsoftware/stryker-js-vitest-runner": minor
+---
+
+Failed tests report the first project source frame as their location and keep the raw stack; file-level failures are named and identified by their project-relative path.

.changeset/witty-spies-try.md

@@ -0,0 +1,5 @@
+---
+"@systemfsoftware/stryker-js-plugin-interface": major
+---
+
+TestRunner.HitLimitReasonText now accepts only the canonical `Hit limit reached (count/limit)` shape; text that merely starts with the prefix no longer validates.

.changeset/young-memes-mix.md

@@ -0,0 +1,5 @@
+---
+"@systemfsoftware/stryker-js": patch
+---
+
+Timeout reasons that only start with `Hit limit reached` no longer classify as hit-limit timeouts; the timeout-fields property now draws every deciding input class under the mutation-worker budget.

.github/workflows/mutation.yml

@@ -3,6 +3,7 @@ name: Mutation
 on:
   push:
     branches: [main]
+  pull_request:
   workflow_dispatch:
 
 concurrency:
@@ -51,12 +52,96 @@ jobs:
           if-no-files-found: ignore
           retention-days: 1
 
-  mutation:
-    name: mutation · ${{ matrix.job.name }}
+  preflight:
+    name: preflight · one dry run per package
     needs: plan
     if: ${{ needs.plan.result == 'success' && needs.plan.outputs.jobs != '[]' }}
     runs-on: ubuntu-latest
+    timeout-minutes: 45
+    outputs:
+      coverage: ${{ steps.preflight.outputs.coverage }}
+    env:
+      OTEL_ENABLED: "true"
+      OTEL_SERVICE_NAME: stryker-js-mutation
+      OTEL_EXPORTER_OTLP_ENDPOINT: http://127.0.0.1:4318
+      OUT_DIR: mutation-telemetry
+    steps:
+      - uses: actions/checkout@v7
+      - uses: pnpm/action-setup@v6
+      - uses: actions/setup-node@v7
+        with:
+          node-version: 24
+          cache: pnpm
+      - run: pnpm install --frozen-lockfile
+      - uses: denoland/setup-deno@v2
+      - name: Build
+        run: pnpm exec turbo run build --output-logs=errors-only --log-order=stream ${{ join(fromJSON(needs.plan.outputs.jobs).*.filters, ' ') }}
+
+      - uses: actions/cache/restore@v6
+        with:
+          path: .stryker-incremental
+          key: stryker-incremental-${{ github.run_id }}
+          restore-keys: |
+            stryker-incremental-
+      - name: Place incremental files
+        run: if [ -d .stryker-incremental ]; then cp -a .stryker-incremental/. .; fi
+
+      - name: Install process-compose (pinned)
+        run: |
+          curl -sSfL -o /tmp/pc.tar.gz https://github.com/F1bonacc1/process-compose/releases/download/v1.122.0/process-compose_linux_amd64.tar.gz
+          echo '9b6dbc38324c0b0481f1cd1dd828ffdc78117129ec797678f4bf8c4023311281  /tmp/pc.tar.gz' | sha256sum -c -
+          sudo tar xzf /tmp/pc.tar.gz -C /usr/local/bin process-compose
+      - name: Start LGTM (blocks until probe-ready)
+        run: |
+          echo "TRACE_WINDOW_START=$(date +%s)" >> "$GITHUB_ENV"
+          pnpm lgtm:up
+
+      - id: preflight
+        name: Preflight
+        env:
+          JOBS: ${{ needs.plan.outputs.jobs }}
+        run: $MUTATION_JOB preflight --cap-seconds "$TARGET_SECONDS"
+
+      - uses: actions/upload-artifact@v7
+        with:
+          name: mutation-preflight
+          path: preflight/
+          if-no-files-found: ignore
+          retention-days: 1
+      - uses: actions/upload-artifact@v7
+        if: failure()
+        with:
+          name: mutation-preflight-evidence
+          path: preflight-evidence/
+          if-no-files-found: ignore
+          retention-days: 30
+
+      - name: Export telemetry
+        if: always()
+        continue-on-error: true
+        run: ./test/e2e/scripts/export-traces.ts
+      - name: Package telemetry
+        if: always()
+        run: if [ -d mutation-telemetry ]; then tar czf mutation-telemetry.tar.gz mutation-telemetry; fi
+      - uses: actions/upload-artifact@v7
+        if: always()
+        with:
+          name: mutation-telemetry-${{ github.run_id }}-preflight
+          path: mutation-telemetry.tar.gz
+          if-no-files-found: ignore
+          retention-days: 30
+
+  mutation:
+    name: mutation · ${{ matrix.job.name }}
+    needs: [plan, preflight]
+    if: ${{ needs.preflight.result == 'success' }}
+    runs-on: ubuntu-latest
     timeout-minutes: 75
+    env:
+      OTEL_ENABLED: "true"
+      OTEL_SERVICE_NAME: stryker-js-mutation
+      OTEL_EXPORTER_OTLP_ENDPOINT: http://127.0.0.1:4318
+      OUT_DIR: mutation-telemetry
     strategy:
       fail-fast: false
       matrix:
@@ -81,6 +166,21 @@ jobs:
             stryker-incremental-
       - name: Place incremental files
         run: if [ -d .stryker-incremental ]; then cp -a .stryker-incremental/. .; fi
+      - uses: actions/download-artifact@v8
+        if: ${{ needs.preflight.outputs.coverage == 'true' }}
+        with:
+          name: mutation-preflight
+          path: .
+
+      - name: Install process-compose (pinned)
+        run: |
+          curl -sSfL -o /tmp/pc.tar.gz https://github.com/F1bonacc1/process-compose/releases/download/v1.122.0/process-compose_linux_amd64.tar.gz
+          echo '9b6dbc38324c0b0481f1cd1dd828ffdc78117129ec797678f4bf8c4023311281  /tmp/pc.tar.gz' | sha256sum -c -
+          sudo tar xzf /tmp/pc.tar.gz -C /usr/local/bin process-compose
+      - name: Start LGTM (blocks until probe-ready)
+        run: |
+          echo "TRACE_WINDOW_START=$(date +%s)" >> "$GITHUB_ENV"
+          pnpm lgtm:up
 
       - name: Mutation
         env:
@@ -107,12 +207,27 @@ jobs:
           name: mutation-part-${{ matrix.job.id }}
           path: mutation-parts/
           if-no-files-found: ignore
-          retention-days: 1
+          retention-days: 30
+
+      - name: Export telemetry
+        if: always()
+        continue-on-error: true
+        run: ./test/e2e/scripts/export-traces.ts
+      - name: Package telemetry
+        if: always()
+        run: if [ -d mutation-telemetry ]; then tar czf mutation-telemetry.tar.gz mutation-telemetry; fi
+      - uses: actions/upload-artifact@v7
+        if: always()
+        with:
+          name: mutation-telemetry-${{ github.run_id }}-${{ matrix.job.id }}
+          path: mutation-telemetry.tar.gz
+          if-no-files-found: ignore
+          retention-days: 30
 
   timings:
     name: record mutation timings
     needs: [plan, mutation]
-    if: ${{ always() && needs.plan.result == 'success' }}
+    if: ${{ always() && needs.plan.result == 'success' && github.event_name != 'pull_request' }}
     runs-on: ubuntu-latest
     timeout-minutes: 5
     steps:
@@ -138,13 +253,14 @@ jobs:
 
   report:
     name: Mutation report
-    needs: [plan, mutation]
-    if: ${{ !cancelled() && needs.plan.result == 'success' && needs.plan.outputs.jobs != '[]' }}
+    needs: [plan, preflight, mutation]
+    if: ${{ !cancelled() && needs.preflight.result == 'success' }}
     runs-on: ubuntu-latest
     timeout-minutes: 25
     permissions:
       contents: read
       actions: write
+      security-events: write
     steps:
       - uses: actions/checkout@v7
       - uses: pnpm/action-setup@v6
@@ -168,7 +284,7 @@ jobs:
           path: .stryker-incremental
           merge-multiple: true
       - uses: actions/cache/save@v6
-        if: ${{ hashFiles('.stryker-incremental/**') != '' }}
+        if: ${{ github.event_name != 'pull_request' && hashFiles('.stryker-incremental/**') != '' }}
         with:
           path: .stryker-incremental
           key: stryker-incremental-${{ github.run_id }}
@@ -194,8 +310,53 @@ jobs:
           path: mutation-merged/*
           retention-days: 14
 
-      - name: Delete per-job artifacts
+      - name: Merge SARIF
         if: always()
+        run: $MUTATION_JOB sarif --parts mutation-parts --out mutation-sarif/mutation.sarif
+      - uses: github/codeql-action/upload-sarif@v4
+        if: ${{ always() && hashFiles('mutation-sarif/mutation.sarif') != '' }}
+        with:
+          sarif_file: mutation-sarif/mutation.sarif
+          category: mutation
+
+      - name: Survivor baseline
+        if: ${{ env.MUTATION_REPORT_COMPLETE == 'true' && github.ref == 'refs/heads/main' }}
+        run: |
+          mkdir -p mutation-gate/reports/mutation
+          cp mutation-merged/mutation-report.json mutation-gate/reports/mutation/mutation.json
+          cd mutation-gate
+          "$GITHUB_WORKSPACE/packages/stryker-js/node_modules/.bin/stryker" gate --baseline mutation-survivor-baseline.json --update-baseline
+      - uses: actions/upload-artifact@v7
+        if: ${{ env.MUTATION_REPORT_COMPLETE == 'true' && github.ref == 'refs/heads/main' }}
+        with:
+          name: mutation-survivor-baseline
+          path: mutation-gate/mutation-survivor-baseline.json
+          retention-days: 90
+
+      - name: Survivor baseline from main
+        if: ${{ github.event_name == 'pull_request' }}
+        env:
+          GH_TOKEN: ${{ github.token }}
+        run: |
+          mkdir -p mutation-gate/reports/mutation
+          id=$(gh api "repos/$GITHUB_REPOSITORY/actions/artifacts?name=mutation-survivor-baseline&per_page=100" \
+            --jq '[.artifacts[] | select(.expired | not) | select(.workflow_run.head_branch == "main")][0].id // empty') || id=
+          if [ -n "$id" ] && gh api "repos/$GITHUB_REPOSITORY/actions/artifacts/$id/zip" > mutation-gate/baseline.zip; then
+            unzip -o -q mutation-gate/baseline.zip -d mutation-gate
+          fi
+      - name: Gate new survivors
+        if: ${{ github.event_name == 'pull_request' }}
+        run: |
+          if [ ! -f mutation-gate/mutation-survivor-baseline.json ]; then
+            echo "::warning title=Mutation gate skipped::main has published no survivor baseline yet, so this run cannot tell a new survivor from an accepted one"
+            exit 0
+          fi
+          cp mutation-merged/mutation-report.json mutation-gate/reports/mutation/mutation.json
+          cd mutation-gate
+          "$GITHUB_WORKSPACE/packages/stryker-js/node_modules/.bin/stryker" gate --baseline mutation-survivor-baseline.json
+
+      - name: Delete per-job artifacts
+        if: ${{ always() && needs.mutation.result == 'success' }}
         continue-on-error: true
         env:
           GH_TOKEN: ${{ github.token }}