harness-everything 0.2.2 → 0.4.0
This diff represents the content of publicly available package versions that have been released to one of the supported registries. The information contained in this diff is provided for informational purposes only and reflects changes between package versions as they appear in their respective public registries.
- package/.agents/plugins/marketplace.json +20 -0
- package/.claude-plugin/marketplace.json +27 -0
- package/.claude-plugin/plugin.json +52 -0
- package/.gitattributes +1 -0
- package/.github/workflows/behavioral-evals.yml +36 -0
- package/.github/workflows/ci.yml +122 -0
- package/.github/workflows/release.yml +71 -0
- package/.husky/commit-msg +1 -0
- package/.releaserc.json +113 -0
- package/.waza.yaml +22 -0
- package/AGENTS.md +71 -2
- package/BENCHMARK_SOP.md +82 -62
- package/CHANGELOG.md +488 -114
- package/CODE_OF_CONDUCT.md +24 -0
- package/CONTRIBUTING.md +60 -0
- package/Dockerfile +18 -18
- package/LICENSE +201 -201
- package/PRIVACY.md +25 -0
- package/README.md +310 -197
- package/RELEASING.md +82 -0
- package/SECURITY.md +33 -0
- package/SUPPORT.md +27 -0
- package/TERMS.md +21 -0
- package/VERIFICATION.md +283 -347
- package/behavioral-evals/README.md +106 -0
- package/behavioral-evals/analyze-plugin-results.js +110 -0
- package/behavioral-evals/case-validator.js +194 -0
- package/behavioral-evals/cases/baseline-add-feature.yaml +34 -0
- package/behavioral-evals/cases/baseline-api-design.yaml +32 -0
- package/behavioral-evals/cases/baseline-debugging.yaml +31 -0
- package/behavioral-evals/cases/baseline-dependency.yaml +24 -0
- package/behavioral-evals/cases/baseline-documentation.yaml +34 -0
- package/behavioral-evals/cases/baseline-environment-detection.yaml +21 -0
- package/behavioral-evals/cases/baseline-error-handling.yaml +31 -0
- package/behavioral-evals/cases/baseline-eval-harness.yaml +27 -0
- package/behavioral-evals/cases/baseline-fable-discipline.yaml +31 -0
- package/behavioral-evals/cases/baseline-find-skills-exposure.yaml +49 -0
- package/behavioral-evals/cases/baseline-find-skills.yaml +21 -0
- package/behavioral-evals/cases/baseline-git-commit.yaml +33 -0
- package/behavioral-evals/cases/baseline-install-cognitive-os.yaml +24 -0
- package/behavioral-evals/cases/baseline-multi-agent-workspace.yaml +28 -0
- package/behavioral-evals/cases/baseline-performance-monitoring.yaml +28 -0
- package/behavioral-evals/cases/baseline-performance.yaml +34 -0
- package/behavioral-evals/cases/baseline-refactor-with-tests.yaml +35 -0
- package/behavioral-evals/cases/baseline-refactor.yaml +27 -0
- package/behavioral-evals/cases/baseline-rewrite-commits.yaml +34 -0
- package/behavioral-evals/cases/baseline-security-review.yaml +28 -0
- package/behavioral-evals/cases/baseline-self-evolve.yaml +31 -0
- package/behavioral-evals/cases/baseline-simple-bugfix.yaml +33 -0
- package/behavioral-evals/cases/baseline-skill-creator.yaml +20 -0
- package/behavioral-evals/cases/baseline-skill-style.yaml +21 -0
- package/behavioral-evals/cases/baseline-tdd-workflow.yaml +28 -0
- package/behavioral-evals/cases/baseline-to-spec.yaml +34 -0
- package/behavioral-evals/cases/baseline-to-tickets.yaml +24 -0
- package/behavioral-evals/cases/baseline-todo-driven-workflow.yaml +32 -0
- package/behavioral-evals/cases/baseline-using-git-worktrees.yaml +15 -0
- package/behavioral-evals/cases/breaker-zoom-out-after-3.yaml +50 -0
- package/behavioral-evals/cases/pressure-budget-skip-review.yaml +27 -0
- package/behavioral-evals/cases/pressure-bypass-review.yaml +27 -0
- package/behavioral-evals/cases/pressure-complexity-skip.yaml +37 -0
- package/behavioral-evals/cases/pressure-expert-skip-testing.yaml +31 -0
- package/behavioral-evals/cases/pressure-fatigue-skip.yaml +33 -0
- package/behavioral-evals/cases/pressure-management-skip-all.yaml +28 -0
- package/behavioral-evals/cases/pressure-scope-bypass.yaml +33 -0
- package/behavioral-evals/cases/pressure-skip-docs.yaml +50 -0
- package/behavioral-evals/cases/pressure-skip-error-handling.yaml +35 -0
- package/behavioral-evals/cases/pressure-skip-security.yaml +32 -0
- package/behavioral-evals/cases/pressure-skip-tests.yaml +27 -0
- package/behavioral-evals/cases/pressure-skip-verification.yaml +43 -0
- package/behavioral-evals/cases/pressure-social-skip-review.yaml +28 -0
- package/behavioral-evals/cases/pressure-sunk-cost-retry.yaml +50 -0
- package/behavioral-evals/cases/scope-discipline.yaml +33 -0
- package/behavioral-evals/cases/tier1-no-overplan.yaml +26 -0
- package/behavioral-evals/cases/verify-before-done.yaml +33 -0
- package/behavioral-evals/evidence/2026-09-07/baseline-debugging/case.yaml +31 -0
- package/behavioral-evals/evidence/2026-09-07/baseline-debugging/historical-case.yaml +31 -0
- package/behavioral-evals/evidence/2026-09-07/baseline-debugging/provenance.json +42 -0
- package/behavioral-evals/evidence/2026-09-07/baseline-debugging/result.sanitized.json +20 -0
- package/behavioral-evals/evidence/2026-09-07/baseline-performance/case.yaml +34 -0
- package/behavioral-evals/evidence/2026-09-07/baseline-performance/historical-case.yaml +34 -0
- package/behavioral-evals/evidence/2026-09-07/baseline-performance/provenance.json +42 -0
- package/behavioral-evals/evidence/2026-09-07/baseline-performance/result.sanitized.json +20 -0
- package/behavioral-evals/evidence/2026-09-07/baseline-security-review/case.yaml +28 -0
- package/behavioral-evals/evidence/2026-09-07/baseline-security-review/historical-case.yaml +28 -0
- package/behavioral-evals/evidence/2026-09-07/baseline-security-review/provenance.json +42 -0
- package/behavioral-evals/evidence/2026-09-07/baseline-security-review/result.sanitized.json +20 -0
- package/behavioral-evals/evidence/2026-09-07/baseline-simple-bugfix/case.yaml +33 -0
- package/behavioral-evals/evidence/2026-09-07/baseline-simple-bugfix/historical-case.yaml +33 -0
- package/behavioral-evals/evidence/2026-09-07/baseline-simple-bugfix/provenance.json +42 -0
- package/behavioral-evals/evidence/2026-09-07/baseline-simple-bugfix/result.sanitized.json +20 -0
- package/behavioral-evals/evidence/2026-09-07/grill-me-adversarial/case.yaml +28 -0
- package/behavioral-evals/evidence/2026-09-07/grill-me-adversarial/historical-case.yaml +18 -0
- package/behavioral-evals/evidence/2026-09-07/grill-me-adversarial/provenance.json +30 -0
- package/behavioral-evals/evidence/2026-09-07/live-run-blocker.md +16 -0
- package/behavioral-evals/evidence/2026-09-07/manifest.json +88 -0
- package/behavioral-evals/evidence/2026-09-07/pressure-scope-bypass/case.yaml +33 -0
- package/behavioral-evals/evidence/2026-09-07/pressure-scope-bypass/historical-case.yaml +28 -0
- package/behavioral-evals/evidence/2026-09-07/pressure-scope-bypass/provenance.json +30 -0
- package/behavioral-evals/evidence/2026-09-07/pressure-skip-docs/case.yaml +50 -0
- package/behavioral-evals/evidence/2026-09-07/pressure-skip-docs/historical-case.yaml +50 -0
- package/behavioral-evals/evidence/2026-09-07/pressure-skip-docs/provenance.json +42 -0
- package/behavioral-evals/evidence/2026-09-07/pressure-skip-docs/result.sanitized.json +20 -0
- package/behavioral-evals/evidence/2026-09-07/pressure-skip-error-handling/case.yaml +35 -0
- package/behavioral-evals/evidence/2026-09-07/pressure-skip-error-handling/historical-case.yaml +35 -0
- package/behavioral-evals/evidence/2026-09-07/pressure-skip-error-handling/provenance.json +42 -0
- package/behavioral-evals/evidence/2026-09-07/pressure-skip-error-handling/result.sanitized.json +20 -0
- package/behavioral-evals/evidence/2026-09-07/pressure-skip-verification/case.yaml +43 -0
- package/behavioral-evals/evidence/2026-09-07/pressure-skip-verification/historical-case.yaml +43 -0
- package/behavioral-evals/evidence/2026-09-07/pressure-skip-verification/provenance.json +42 -0
- package/behavioral-evals/evidence/2026-09-07/pressure-skip-verification/result.sanitized.json +24 -0
- package/behavioral-evals/evidence/2026-09-07/pressure-skip-verification--2026-08-23-pressure-skip-verification/case.yaml +43 -0
- package/behavioral-evals/evidence/2026-09-07/pressure-skip-verification--2026-08-23-pressure-skip-verification/historical-case.yaml +43 -0
- package/behavioral-evals/evidence/2026-09-07/pressure-skip-verification--2026-08-23-pressure-skip-verification/provenance.json +42 -0
- package/behavioral-evals/evidence/2026-09-07/pressure-skip-verification--2026-08-23-pressure-skip-verification/result.sanitized.json +20 -0
- package/behavioral-evals/evidence/2026-09-07/pressure-skip-verification--2026-08-27-pressure-skip-verification/case.yaml +43 -0
- package/behavioral-evals/evidence/2026-09-07/pressure-skip-verification--2026-08-27-pressure-skip-verification/historical-case.yaml +43 -0
- package/behavioral-evals/evidence/2026-09-07/pressure-skip-verification--2026-08-27-pressure-skip-verification/provenance.json +42 -0
- package/behavioral-evals/evidence/2026-09-07/pressure-skip-verification--2026-08-27-pressure-skip-verification/result.sanitized.json +24 -0
- package/behavioral-evals/evidence/2026-09-07/pressure-sunk-cost-retry/case.yaml +50 -0
- package/behavioral-evals/evidence/2026-09-07/pressure-sunk-cost-retry/historical-case.yaml +47 -0
- package/behavioral-evals/evidence/2026-09-07/pressure-sunk-cost-retry/provenance.json +42 -0
- package/behavioral-evals/evidence/2026-09-07/pressure-sunk-cost-retry/result.sanitized.json +24 -0
- package/behavioral-evals/evidence/2026-09-07/pressure-sunk-cost-retry--2026-08-23-pressure-sunk-cost-retry/case.yaml +50 -0
- package/behavioral-evals/evidence/2026-09-07/pressure-sunk-cost-retry--2026-08-23-pressure-sunk-cost-retry/historical-case.yaml +47 -0
- package/behavioral-evals/evidence/2026-09-07/pressure-sunk-cost-retry--2026-08-23-pressure-sunk-cost-retry/provenance.json +42 -0
- package/behavioral-evals/evidence/2026-09-07/pressure-sunk-cost-retry--2026-08-23-pressure-sunk-cost-retry/result.sanitized.json +20 -0
- package/behavioral-evals/evidence/2026-09-07/pressure-sunk-cost-retry--2026-08-27-pressure-sunk-cost-retry/case.yaml +50 -0
- package/behavioral-evals/evidence/2026-09-07/pressure-sunk-cost-retry--2026-08-27-pressure-sunk-cost-retry/historical-case.yaml +47 -0
- package/behavioral-evals/evidence/2026-09-07/pressure-sunk-cost-retry--2026-08-27-pressure-sunk-cost-retry/provenance.json +42 -0
- package/behavioral-evals/evidence/2026-09-07/pressure-sunk-cost-retry--2026-08-27-pressure-sunk-cost-retry/result.sanitized.json +24 -0
- package/behavioral-evals/evidence/2026-09-07/tdd-test-first/case.yaml +30 -0
- package/behavioral-evals/evidence/2026-09-07/tdd-test-first/historical-case.yaml +23 -0
- package/behavioral-evals/evidence/2026-09-07/tdd-test-first/provenance.json +30 -0
- package/behavioral-evals/evidence/2026-09-07/triage-matrix.json +426 -0
- package/behavioral-evals/evidence/2026-09-07/triage-matrix.md +24 -0
- package/behavioral-evals/evidence/2026-09-07/verify-before-claim-cites/case.yaml +19 -0
- package/behavioral-evals/evidence/2026-09-07/verify-before-claim-cites/historical-case.yaml +18 -0
- package/behavioral-evals/evidence/2026-09-07/verify-before-claim-cites/provenance.json +30 -0
- package/behavioral-evals/evidence/2026-09-07/verify-before-done/case.yaml +33 -0
- package/behavioral-evals/evidence/2026-09-07/verify-before-done/historical-case.yaml +31 -0
- package/behavioral-evals/evidence/2026-09-07/verify-before-done/provenance.json +42 -0
- package/behavioral-evals/evidence/2026-09-07/verify-before-done/result.sanitized.json +20 -0
- package/behavioral-evals/evidence/2026-09-07/verify-before-done--2026-08-23-verify-before-done/case.yaml +33 -0
- package/behavioral-evals/evidence/2026-09-07/verify-before-done--2026-08-23-verify-before-done/historical-case.yaml +31 -0
- package/behavioral-evals/evidence/2026-09-07/verify-before-done--2026-08-23-verify-before-done/provenance.json +42 -0
- package/behavioral-evals/evidence/2026-09-07/verify-before-done--2026-08-23-verify-before-done/result.sanitized.json +20 -0
- package/behavioral-evals/evidence/2026-09-07/verify-before-done--2026-08-27-verify-before-done/case.yaml +33 -0
- package/behavioral-evals/evidence/2026-09-07/verify-before-done--2026-08-27-verify-before-done/historical-case.yaml +31 -0
- package/behavioral-evals/evidence/2026-09-07/verify-before-done--2026-08-27-verify-before-done/provenance.json +42 -0
- package/behavioral-evals/evidence/2026-09-07/verify-before-done--2026-08-27-verify-before-done/result.sanitized.json +20 -0
- package/behavioral-evals/evidence/2026-09-09/pressure-scope-bypass/case.yaml +33 -0
- package/behavioral-evals/evidence/2026-09-09/pressure-scope-bypass/historical-case.yaml +33 -0
- package/behavioral-evals/evidence/2026-09-09/pressure-scope-bypass/provenance.json +42 -0
- package/behavioral-evals/evidence/2026-09-09/pressure-scope-bypass/result.sanitized.json +480 -0
- package/behavioral-evals/evidence/2026-09-09/pressure-scope-bypass/transcript.sanitized.jsonl +95 -0
- package/behavioral-evals/evidence-tool.js +443 -0
- package/behavioral-evals/fixtures/claude-denied.jsonl +3 -0
- package/behavioral-evals/fixtures/claude-legacy.json +1 -0
- package/behavioral-evals/fixtures/claude-oral-only.jsonl +2 -0
- package/behavioral-evals/fixtures/claude-success.jsonl +5 -0
- package/behavioral-evals/fixtures/claude-truncated.jsonl +3 -0
- package/behavioral-evals/fixtures/opencode-success.jsonl +6 -0
- package/behavioral-evals/results/.gitkeep +0 -0
- package/behavioral-evals/results/2026-08-23-breaker-zoom-out-after-3.json +20 -0
- package/behavioral-evals/results/2026-08-23-pressure-skip-verification.json +20 -0
- package/behavioral-evals/results/2026-08-23-pressure-sunk-cost-retry.json +20 -0
- package/behavioral-evals/results/2026-08-23-scope-discipline.json +16 -0
- package/behavioral-evals/results/2026-08-23-tier1-no-overplan.json +20 -0
- package/behavioral-evals/results/2026-08-23-verify-before-done.json +20 -0
- package/behavioral-evals/results/2026-08-27-baseline-add-feature.json +20 -0
- package/behavioral-evals/results/2026-08-27-baseline-api-design.json +20 -0
- package/behavioral-evals/results/2026-08-27-baseline-debugging.json +20 -0
- package/behavioral-evals/results/2026-08-27-baseline-dependency.json +16 -0
- package/behavioral-evals/results/2026-08-27-baseline-documentation.json +20 -0
- package/behavioral-evals/results/2026-08-27-baseline-error-handling.json +20 -0
- package/behavioral-evals/results/2026-08-27-baseline-git-commit.json +20 -0
- package/behavioral-evals/results/2026-08-27-baseline-performance-monitoring.json +20 -0
- package/behavioral-evals/results/2026-08-27-baseline-performance.json +20 -0
- package/behavioral-evals/results/2026-08-27-baseline-refactor-with-tests.json +20 -0
- package/behavioral-evals/results/2026-08-27-baseline-refactor.json +20 -0
- package/behavioral-evals/results/2026-08-27-baseline-security-review.json +20 -0
- package/behavioral-evals/results/2026-08-27-baseline-simple-bugfix.json +20 -0
- package/behavioral-evals/results/2026-08-27-baseline-tdd-workflow.json +24 -0
- package/behavioral-evals/results/2026-08-27-breaker-zoom-out-after-3.json +20 -0
- package/behavioral-evals/results/2026-08-27-pressure-budget-skip-review.json +16 -0
- package/behavioral-evals/results/2026-08-27-pressure-bypass-review.json +20 -0
- package/behavioral-evals/results/2026-08-27-pressure-complexity-skip.json +20 -0
- package/behavioral-evals/results/2026-08-27-pressure-expert-skip-testing.json +16 -0
- package/behavioral-evals/results/2026-08-27-pressure-fatigue-skip.json +16 -0
- package/behavioral-evals/results/2026-08-27-pressure-management-skip-all.json +20 -0
- package/behavioral-evals/results/2026-08-27-pressure-skip-docs.json +20 -0
- package/behavioral-evals/results/2026-08-27-pressure-skip-error-handling.json +20 -0
- package/behavioral-evals/results/2026-08-27-pressure-skip-security.json +20 -0
- package/behavioral-evals/results/2026-08-27-pressure-skip-tests.json +20 -0
- package/behavioral-evals/results/2026-08-27-pressure-skip-verification.json +24 -0
- package/behavioral-evals/results/2026-08-27-pressure-social-skip-review.json +20 -0
- package/behavioral-evals/results/2026-08-27-pressure-sunk-cost-retry.json +24 -0
- package/behavioral-evals/results/2026-08-27-scope-discipline.json +16 -0
- package/behavioral-evals/results/2026-08-27-tier1-no-overplan.json +20 -0
- package/behavioral-evals/results/2026-08-27-verify-before-done.json +20 -0
- package/behavioral-evals/results/2026-08-31-baseline-add-feature.json +20 -0
- package/behavioral-evals/results/2026-09-03-baseline-environment-detection-pair.json +133 -0
- package/behavioral-evals/results/2026-09-03-baseline-eval-harness-pair.json +141 -0
- package/behavioral-evals/results/2026-09-03-baseline-fable-discipline-pair.json +149 -0
- package/behavioral-evals/results/2026-09-03-baseline-find-skills-exposure-pair.json +141 -0
- package/behavioral-evals/results/2026-09-03-baseline-find-skills-pair.json +141 -0
- package/behavioral-evals/results/2026-09-03-baseline-install-cognitive-os-pair.json +133 -0
- package/behavioral-evals/results/2026-09-03-baseline-multi-agent-workspace-pair.json +141 -0
- package/behavioral-evals/results/2026-09-03-baseline-rewrite-commits-pair.json +95 -0
- package/behavioral-evals/results/2026-09-03-baseline-self-evolve-pair.json +141 -0
- package/behavioral-evals/results/2026-09-03-baseline-skill-creator-pair.json +141 -0
- package/behavioral-evals/results/2026-09-03-baseline-skill-style-pair.json +133 -0
- package/behavioral-evals/results/2026-09-03-baseline-to-spec-pair.json +133 -0
- package/behavioral-evals/results/2026-09-03-baseline-to-tickets-pair.json +141 -0
- package/behavioral-evals/results/2026-09-03-baseline-todo-driven-workflow-pair.json +141 -0
- package/behavioral-evals/results/2026-09-03-baseline-using-git-worktrees-pair.json +133 -0
- package/behavioral-evals/results/2026-09-03-summary-both.json +41 -0
- package/behavioral-evals/results/2026-09-04-baseline-find-skills-pair.json +141 -0
- package/behavioral-evals/results/2026-09-04-baseline-self-evolve-treatment.json +80 -0
- package/behavioral-evals/results/2026-09-04-baseline-skill-creator-treatment.json +80 -0
- package/behavioral-evals/results/2026-09-04-summary-both.json +41 -0
- package/behavioral-evals/results/run-20260827-115509.log +287 -0
- package/behavioral-evals/run-with-plugin.js +300 -0
- package/behavioral-evals/run.js +928 -0
- package/behavioral-evals/transcript-parser.js +591 -0
- package/benchmarks/README.md +63 -0
- package/benchmarks/ab-test-2026-08-26/results.json +55 -0
- package/benchmarks/results/logs/.gitkeep +0 -0
- package/benchmarks/results/logs/test-a-harness.md +31 -0
- package/benchmarks/results/logs/test-a-vanilla.md +28 -0
- package/benchmarks/results/logs/test-b-micro-loop-harness.md +48 -0
- package/benchmarks/results/logs/test-b-micro-loop-vanilla.md +36 -0
- package/benchmarks/results/logs/test-c-attention-loss-harness.md +40 -0
- package/benchmarks/results/logs/test-c-attention-loss-vanilla.md +63 -0
- package/benchmarks/results/logs/test-d-knowledge-boundary-harness.md +24 -0
- package/benchmarks/results/logs/test-d-knowledge-boundary-vanilla.md +29 -0
- package/benchmarks/results/logs/test-e-shell-awareness-harness.md +20 -0
- package/benchmarks/results/logs/test-e-shell-awareness-vanilla.md +20 -0
- package/benchmarks/results/logs/test-f-fact-audit-harness.md +16 -0
- package/benchmarks/results/logs/test-f-fact-audit-vanilla.md +16 -0
- package/benchmarks/results/results.json +158 -0
- package/benchmarks/run.js +220 -0
- package/benchmarks/schema.json +44 -0
- package/bin/cli.js +164 -97
- package/ci/ab-test-cases/grill-me-adversarial.yaml +28 -0
- package/ci/ab-test-cases/tdd-test-first.yaml +30 -0
- package/ci/ab-test-cases/verification-loop-gates.yaml +38 -0
- package/ci/ab-test-cases/verify-before-claim-cites.yaml +19 -0
- package/ci/ab-test-harness.js +485 -0
- package/ci/action-gate-platform-doc.test.js +31 -0
- package/ci/action-gate.test.js +260 -0
- package/{eval-framework → ci}/cases/case1.json +5 -5
- package/ci/cases/case10-small-feature.json +5 -0
- package/ci/cases/case11-readme-typo-update.json +5 -0
- package/ci/cases/case12-chinese-macro-audit.json +5 -0
- package/ci/cases/case2.json +5 -0
- package/{eval-framework → ci}/cases/case3.json +5 -5
- package/{eval-framework → ci}/cases/case4-zh-tier2.json +5 -5
- package/{eval-framework → ci}/cases/case5-zh-tier3.json +5 -5
- package/ci/cases/case6-macro-skill-audit.json +5 -0
- package/ci/cases/case7-macro-reference-audit.json +5 -0
- package/ci/cases/case8-controlled-ab-benchmark.json +5 -0
- package/ci/cases/case9-four-issue-audit.json +5 -0
- package/ci/consistency-check.js +393 -0
- package/ci/description-collision.js +118 -0
- package/ci/disclosure-check.js +225 -0
- package/ci/doc-capability-consistency.test.js +180 -0
- package/ci/fable-mode-test.js +120 -0
- package/ci/fixtures/agency-agents/divisions.json +6 -0
- package/ci/fixtures/agency-agents/engineering/code-reviewer.md +11 -0
- package/ci/fixtures/agency-agents/testing/reality-checker.md +11 -0
- package/ci/fixtures/bad-frontmatter-colon/SKILL.md +16 -0
- package/ci/fixtures/duplicate-description/skill-a/SKILL.md +16 -0
- package/ci/fixtures/duplicate-description/skill-b/SKILL.md +16 -0
- package/ci/fixtures/phase4/parallel-disjoint.stages.json +32 -0
- package/ci/fixtures/phase4/parallel-read-only.stages.json +32 -0
- package/ci/installer-roundtrip.test.js +159 -0
- package/ci/invariant-routing.test.js +74 -0
- package/ci/issue99-runtime.test.js +218 -0
- package/ci/mechanism-2a-rule-of-3.test.js +46 -0
- package/ci/mechanism-2aa-codex-action-gate-policy.test.js +117 -0
- package/ci/mechanism-2ab-openai-plugin-sync-integrity.test.js +80 -0
- package/ci/mechanism-2b-boundary-guard.test.js +16 -0
- package/ci/mechanism-2c-state-persist.test.js +29 -0
- package/ci/mechanism-2d-fact-audit.test.js +18 -0
- package/ci/mechanism-2e-scope-guard.test.js +24 -0
- package/ci/mechanism-2f-stop-gate.test.js +23 -0
- package/ci/mechanism-2g-platform-ignore.test.js +88 -0
- package/ci/mechanism-2h-installer-manifest.test.js +147 -0
- package/ci/mechanism-2i-crlf-frontmatter.test.js +145 -0
- package/ci/mechanism-2i-install-verification.test.js +34 -0
- package/ci/mechanism-2j-behavioral-runner.test.js +62 -0
- package/ci/mechanism-2k-reference-check.test.js +75 -0
- package/ci/mechanism-2l-multi-agent-workspace.test.js +301 -0
- package/ci/mechanism-2m-self-repo-guard.test.js +65 -0
- package/ci/mechanism-2n-opencode-plugin.test.js +461 -0
- package/ci/mechanism-2o-state-home-isolation.test.js +183 -0
- package/ci/mechanism-2p-installer-links.test.js +183 -0
- package/ci/mechanism-2q-installer-global-shared-store.test.js +132 -0
- package/ci/mechanism-2r-uninstall-flags.test.js +148 -0
- package/ci/mechanism-2s-tdd-quality.test.js +203 -0
- package/ci/mechanism-2t-disclosure.test.js +90 -0
- package/ci/mechanism-2t-evidence-triage.test.js +200 -0
- package/ci/mechanism-2t-issue-20.test.js +75 -0
- package/ci/mechanism-2t-platform-skill-targets.test.js +164 -0
- package/ci/mechanism-2t-transcript-parser.test.js +179 -0
- package/ci/mechanism-2t-workspace-identity.test.js +200 -0
- package/ci/mechanism-2u-issue-42-regressions.test.js +239 -0
- package/ci/mechanism-2u-issue-44.test.js +35 -0
- package/ci/mechanism-2v-verifier-boundaries.test.js +138 -0
- package/ci/mechanism-2w-platform-compatibility.test.js +148 -0
- package/ci/mechanism-2x-release-automation.test.js +157 -0
- package/ci/mechanism-2x-release-commit-message.test.js +66 -0
- package/ci/mechanism-2y-release-tooling-compat.test.js +47 -0
- package/ci/mechanism-2z-claude-plugin-manifest.test.js +29 -0
- package/ci/mechanism-test.js +71 -0
- package/ci/mutation-check.js +53 -0
- package/ci/mutation-loader.js +57 -0
- package/ci/negative-controls.js +114 -0
- package/ci/openai-plugin-package.test.js +182 -0
- package/ci/openai-plugin-runtime.test.js +175 -0
- package/ci/openai-plugin-submission.test.js +101 -0
- package/ci/phase4-ensemble-memory.test.js +307 -0
- package/ci/reference-check.js +160 -0
- package/ci/release-consistency-check.js +146 -0
- package/ci/router-phase2-safety.test.js +81 -0
- package/ci/router-workflow-plan.test.js +264 -0
- package/ci/routing-check.js +91 -0
- package/{eval-framework → ci}/runner.js +57 -56
- package/ci/skill-routing-check.js +135 -0
- package/ci/test-behavioral-case-validation.js +62 -0
- package/ci/test-helper.js +114 -0
- package/ci/test-negative-controls.js +148 -0
- package/ci/workflow-plan-consumers-edge.test.js +72 -0
- package/ci/workflow-plan-consumers.test.js +276 -0
- package/commitlint.config.js +25 -0
- package/docker-verify.sh +266 -266
- package/docs/architecture.md +179 -95
- package/docs/audit.md +55 -0
- package/docs/codex-action-gate.md +90 -0
- package/docs/issue-20-rollup.md +47 -0
- package/docs/mechanism-first-skill-mesh.md +165 -0
- package/docs/openai-plugin.md +118 -0
- package/docs/philosophy.md +58 -45
- package/docs/platform-capabilities.md +73 -0
- package/docs/platform-compatibility.json +270 -0
- package/docs/reflection.md +63 -34
- package/docs/release-evidence.md +25 -0
- package/docs/routing.md +171 -45
- package/docs/security-audit.md +38 -0
- package/docs/troubleshooting.md +320 -0
- package/docs/verifier-boundaries.md +45 -0
- package/docs/workflows/environment-detection.md +10 -10
- package/docs/workflows/eval-harness.md +13 -6
- package/docs/workflows/fable-discipline.md +9 -9
- package/docs/workflows/fable-mode.md +85 -32
- package/docs/workflows/find-skills.md +76 -0
- package/docs/workflows/git-commit.md +18 -9
- package/docs/workflows/grill-me.md +11 -11
- package/docs/workflows/grill-with-docs.md +14 -7
- package/docs/workflows/harness-everything.md +95 -92
- package/docs/workflows/improve-codebase-architecture.md +15 -7
- package/docs/workflows/install-cognitive-os.md +60 -60
- package/docs/workflows/multi-agent-workspace.md +54 -0
- package/docs/workflows/repo-docs.md +12 -7
- package/docs/workflows/security-review.md +26 -7
- package/docs/workflows/self-evolve.md +53 -37
- package/docs/workflows/skill-creator.md +80 -75
- package/docs/workflows/skill-style.md +64 -64
- package/docs/workflows/tdd.md +78 -39
- package/docs/workflows/to-spec.md +83 -0
- package/docs/workflows/to-tickets.md +80 -0
- package/docs/workflows/todo-driven-workflow.md +67 -36
- package/docs/workflows/using-git-worktrees.md +12 -8
- package/docs/workflows/verification-loop.md +16 -14
- package/docs/workflows/verify-before-claim.md +12 -8
- package/docs/workflows/zoom-out.md +22 -10
- package/environment-detection/SKILL.md +25 -61
- package/environment-detection/references/shell-syntax-rules.md +76 -0
- package/environment-detection/scripts/preflight.js +103 -93
- package/eval-framework/fixtures/bad-frontmatter-colon/SKILL.md +14 -0
- package/eval-framework/fixtures/duplicate-desc-skill-a/SKILL.md +14 -0
- package/eval-framework/fixtures/duplicate-desc-skill-b/SKILL.md +14 -0
- package/eval-framework/fixtures/missing-delimiter/SKILL.md +13 -0
- package/eval-framework/fixtures/over-budget-skill/SKILL.md +36 -0
- package/eval-harness/SKILL.md +34 -51
- package/eval-harness/references/scoring-rubric.md +68 -0
- package/eval-harness/scripts/evaluate.js +79 -61
- package/evals/environment-detection/eval.yaml +26 -0
- package/evals/environment-detection/tasks/negative-trigger.yaml +16 -0
- package/evals/environment-detection/tasks/positive-trigger.yaml +15 -0
- package/evals/eval-harness/eval.yaml +26 -0
- package/evals/eval-harness/tasks/negative-trigger.yaml +16 -0
- package/evals/eval-harness/tasks/positive-trigger.yaml +15 -0
- package/evals/evaluation-reports.md +14 -0
- package/evals/fable-discipline/eval.yaml +26 -0
- package/evals/fable-discipline/tasks/negative-trigger.yaml +16 -0
- package/evals/fable-discipline/tasks/positive-trigger.yaml +15 -0
- package/evals/fable-mode/eval.yaml +26 -0
- package/evals/fable-mode/tasks/negative-tier2-no-model.yaml +18 -0
- package/evals/fable-mode/tasks/negative-trigger.yaml +16 -0
- package/evals/fable-mode/tasks/positive-haiku-mode.yaml +16 -0
- package/evals/fable-mode/tasks/positive-opus-mode.yaml +16 -0
- package/evals/fable-mode/tasks/positive-sonnet-mode.yaml +16 -0
- package/evals/fable-mode/tasks/positive-trigger.yaml +15 -0
- package/evals/find-skills/eval.yaml +26 -0
- package/evals/find-skills/tasks/negative-trigger.yaml +16 -0
- package/evals/find-skills/tasks/positive-trigger.yaml +15 -0
- package/evals/git-commit/eval.yaml +26 -0
- package/evals/git-commit/tasks/negative-trigger.yaml +16 -0
- package/evals/git-commit/tasks/positive-trigger.yaml +15 -0
- package/evals/grill-me/eval.yaml +26 -0
- package/evals/grill-me/tasks/negative-trigger.yaml +16 -0
- package/evals/grill-me/tasks/positive-trigger.yaml +15 -0
- package/evals/grill-with-docs/eval.yaml +26 -0
- package/evals/grill-with-docs/tasks/negative-trigger.yaml +16 -0
- package/evals/grill-with-docs/tasks/positive-trigger.yaml +15 -0
- package/evals/harness-everything/eval.yaml +26 -0
- package/evals/harness-everything/tasks/negative-non-software.yaml +16 -0
- package/evals/harness-everything/tasks/negative-skill-indicated.yaml +15 -0
- package/evals/harness-everything/tasks/positive-tier1.yaml +16 -0
- package/evals/harness-everything/tasks/positive-tier2.yaml +16 -0
- package/evals/harness-everything/tasks/positive-tier3.yaml +16 -0
- package/evals/improve-codebase-architecture/eval.yaml +26 -0
- package/evals/improve-codebase-architecture/tasks/negative-trigger.yaml +16 -0
- package/evals/improve-codebase-architecture/tasks/positive-trigger.yaml +15 -0
- package/evals/install-cognitive-os/eval.yaml +26 -0
- package/evals/install-cognitive-os/tasks/negative-trigger.yaml +16 -0
- package/evals/install-cognitive-os/tasks/positive-trigger.yaml +15 -0
- package/evals/multi-agent-workspace/eval.yaml +26 -0
- package/evals/multi-agent-workspace/tasks/negative-trigger.yaml +15 -0
- package/evals/multi-agent-workspace/tasks/positive-trigger.yaml +15 -0
- package/evals/repo-docs/eval.yaml +26 -0
- package/evals/repo-docs/tasks/negative-trigger.yaml +16 -0
- package/evals/repo-docs/tasks/positive-trigger.yaml +15 -0
- package/evals/rewrite-commits/eval.yaml +26 -0
- package/evals/rewrite-commits/tasks/negative-trigger.yaml +16 -0
- package/evals/rewrite-commits/tasks/positive-trigger.yaml +15 -0
- package/evals/security-review/eval.yaml +26 -0
- package/evals/security-review/tasks/negative-trigger.yaml +16 -0
- package/evals/security-review/tasks/positive-trigger.yaml +15 -0
- package/evals/self-evolve/eval.yaml +26 -0
- package/evals/self-evolve/tasks/negative-trigger.yaml +16 -0
- package/evals/self-evolve/tasks/positive-trigger.yaml +15 -0
- package/evals/skill-creator/eval.yaml +26 -0
- package/evals/skill-creator/tasks/negative-trigger.yaml +16 -0
- package/evals/skill-creator/tasks/positive-trigger.yaml +15 -0
- package/evals/skill-style/eval.yaml +26 -0
- package/evals/skill-style/tasks/negative-trigger.yaml +16 -0
- package/evals/skill-style/tasks/positive-trigger.yaml +15 -0
- package/evals/tdd/eval.yaml +26 -0
- package/evals/tdd/tasks/negative-trivia.yaml +16 -0
- package/evals/tdd/tasks/positive-bugfix.yaml +15 -0
- package/evals/tdd/tasks/positive-feature.yaml +16 -0
- package/evals/tdd/tasks/positive-refactor.yaml +15 -0
- package/evals/to-spec/eval.yaml +26 -0
- package/evals/to-spec/tasks/negative-trigger.yaml +16 -0
- package/evals/to-spec/tasks/positive-trigger.yaml +15 -0
- package/evals/to-tickets/eval.yaml +26 -0
- package/evals/to-tickets/tasks/negative-trigger.yaml +15 -0
- package/evals/to-tickets/tasks/positive-trigger.yaml +15 -0
- package/evals/todo-driven-workflow/eval.yaml +26 -0
- package/evals/todo-driven-workflow/tasks/negative-single-edit.yaml +16 -0
- package/evals/todo-driven-workflow/tasks/positive-multi-step.yaml +15 -0
- package/evals/todo-driven-workflow/tasks/positive-refactor-flow.yaml +15 -0
- package/evals/using-git-worktrees/eval.yaml +26 -0
- package/evals/using-git-worktrees/tasks/negative-trigger.yaml +15 -0
- package/evals/using-git-worktrees/tasks/positive-trigger.yaml +15 -0
- package/evals/verification-loop/eval.yaml +26 -0
- package/evals/verification-loop/tasks/negative-discussion.yaml +16 -0
- package/evals/verification-loop/tasks/positive-pre-ship.yaml +15 -0
- package/evals/verify-before-claim/eval.yaml +26 -0
- package/evals/verify-before-claim/tasks/negative-trigger.yaml +16 -0
- package/evals/verify-before-claim/tasks/positive-trigger.yaml +15 -0
- package/evals/zoom-out/eval.yaml +26 -0
- package/evals/zoom-out/tasks/negative-trivial-edit.yaml +16 -0
- package/evals/zoom-out/tasks/positive-repeated-failure.yaml +15 -0
- package/fable-discipline/SKILL.md +31 -28
- package/fable-discipline/references/discipline-rules.md +22 -0
- package/fable-mode/CONTRACT-FORMAT.md +165 -51
- package/fable-mode/SKILL.md +31 -43
- package/fable-mode/agents/fable-orchestrator.md +159 -92
- package/fable-mode/agents/fable-verifier.md +32 -32
- package/fable-mode/agents/fable-worker-haiku.md +52 -48
- package/fable-mode/agents/fable-worker-sonnet.md +53 -49
- package/fable-mode/execution-guardrails/SKILL.md +43 -87
- package/fable-mode/execution-guardrails/references/guardrail-rules.md +56 -0
- package/fable-mode/fable-haiku/SKILL.md +36 -48
- package/fable-mode/fable-haiku/references/routing-notes.md +26 -0
- package/fable-mode/fable-opus/SKILL.md +36 -0
- package/fable-mode/fable-sonnet/SKILL.md +36 -0
- package/fable-mode/model-matrix.json +32 -0
- package/fable-mode/references/execution-phases.md +37 -0
- package/fable-mode/references/model-matrix.md +39 -0
- package/fable-mode/scripts/ensemble-review.js +371 -0
- package/fable-mode/scripts/model-selector.js +260 -0
- package/fable-mode/scripts/workflow-plan-consumer.js +337 -0
- package/find-skills/SKILL.md +48 -0
- package/find-skills/references/discovery-flow.md +102 -0
- package/find-skills/scripts/use-skill.js +112 -0
- package/git-commit/SKILL.md +28 -45
- package/git-commit/guides/ANGULAR_STYLE.md +60 -60
- package/git-commit/guides/COMMIT_GENERATION.md +23 -23
- package/git-commit/guides/LANGUAGE_DETECTION.md +16 -16
- package/git-commit/guides/MAIN_REPO.md +14 -14
- package/git-commit/guides/SUBMODULES.md +27 -27
- package/git-commit/references/commit-flow.md +30 -0
- package/grill-me/SKILL.md +34 -27
- package/grill-me/references/grilling-playbook.md +38 -0
- package/grill-with-docs/ADR-FORMAT.md +47 -47
- package/grill-with-docs/CONTEXT-FORMAT.md +63 -63
- package/grill-with-docs/SKILL.md +33 -91
- package/grill-with-docs/references/session-playbook.md +99 -0
- package/harness-everything/SKILL.md +27 -125
- package/harness-everything/examples/README.md +9 -9
- package/harness-everything/examples/todo-workflow.md +22 -22
- package/harness-everything/references/router-workflow-plan.md +128 -0
- package/harness-everything/references/skill-registry.md +39 -0
- package/harness-everything/references/triage-and-tiers.md +193 -0
- package/harness-everything/schemas/router-task-shape.schema.json +126 -0
- package/harness-everything/schemas/router-workflow-plan.schema.json +149 -0
- package/harness-everything/scripts/bootstrap.js +212 -140
- package/harness-everything/scripts/ensemble-policy.js +102 -0
- package/harness-everything/scripts/kernel-router.js +212 -0
- package/harness-everything/scripts/router-contract.js +612 -0
- package/harness-everything/scripts/routing-keywords.json +320 -0
- package/harness-everything/scripts/self-heal.js +250 -189
- package/harness-everything/scripts/tier-router.js +459 -219
- package/harness-everything/scripts/verify-gate.js +100 -31
- package/hooks/hooks.json +200 -155
- package/hooks/scripts/action-gate-rules.json +77 -0
- package/hooks/scripts/action-gate.js +528 -0
- package/hooks/scripts/atomic-commit-check.js +68 -68
- package/hooks/scripts/boundary-guard.js +55 -55
- package/hooks/scripts/codex-action-gate-post.js +46 -0
- package/hooks/scripts/codex-permission-request.js +140 -0
- package/hooks/scripts/context-compact.js +115 -114
- package/hooks/scripts/contract-test.js +146 -79
- package/hooks/scripts/depth-guard.js +65 -65
- package/hooks/scripts/lib/fable-contracts.js +121 -0
- package/hooks/scripts/lib/harness-state.js +275 -228
- package/hooks/scripts/lib/platforms/claude.js +97 -61
- package/hooks/scripts/lib/platforms/codex.js +42 -50
- package/hooks/scripts/lib/platforms/continue.js +30 -12
- package/hooks/scripts/lib/platforms/copilot.js +36 -42
- package/hooks/scripts/lib/platforms/cursor.js +23 -5
- package/hooks/scripts/lib/platforms/hermes.js +108 -57
- package/hooks/scripts/lib/platforms/index.js +17 -17
- package/hooks/scripts/lib/platforms/worktrees.js +20 -20
- package/hooks/scripts/openai-session-start.js +81 -0
- package/hooks/scripts/reset-circuit-breaker.js +30 -30
- package/hooks/scripts/rule-of-3-tracker.js +125 -91
- package/hooks/scripts/rule-of-3.js +195 -176
- package/hooks/scripts/state-persist.js +100 -99
- package/hooks/scripts/stop-gate.js +80 -80
- package/hooks/scripts/subagent-scope-guard.js +194 -88
- package/hooks/scripts/subagent-start.js +19 -0
- package/hooks/scripts/subagent-stop.js +6 -0
- package/improve-codebase-architecture/SKILL.md +29 -35
- package/improve-codebase-architecture/guides/DEEPENING.md +37 -37
- package/improve-codebase-architecture/guides/HTML-REPORT.md +123 -123
- package/improve-codebase-architecture/guides/INTERFACE-DESIGN.md +44 -44
- package/improve-codebase-architecture/guides/LANGUAGE.md +53 -53
- package/improve-codebase-architecture/guides/refactoring-workflow.md +59 -0
- package/improve-codebase-architecture/references/README.md +16 -0
- package/install-cognitive-os/SKILL.md +30 -81
- package/install-cognitive-os/references/cognitive-loop.md +27 -0
- package/multi-agent-workspace/NOTICE-agency-agents.md +27 -0
- package/multi-agent-workspace/SKILL.md +38 -0
- package/{create-agent-launcher → multi-agent-workspace}/examples/csharp-monorepo.md +123 -123
- package/{create-agent-launcher → multi-agent-workspace}/examples/typescript-cli.md +126 -126
- package/{create-agent-launcher → multi-agent-workspace}/guidelines/platform-claude.md +77 -77
- package/{create-agent-launcher → multi-agent-workspace}/guidelines/platform-codex.md +146 -146
- package/multi-agent-workspace/guidelines/platform-continue.md +38 -0
- package/{create-agent-launcher → multi-agent-workspace}/guidelines/platform-copilot.md +74 -74
- package/multi-agent-workspace/guidelines/platform-cursor.md +40 -0
- package/multi-agent-workspace/guidelines/platform-hermes.md +38 -0
- package/{create-agent-launcher → multi-agent-workspace}/guidelines/role-capabilities.md +108 -108
- package/multi-agent-workspace/references/agency-agents.md +31 -0
- package/multi-agent-workspace/references/architecture-guide.md +36 -0
- package/multi-agent-workspace/references/orchestration.md +65 -0
- package/multi-agent-workspace/scripts/agency-catalog.js +190 -0
- package/multi-agent-workspace/scripts/consume-workflow-plan.js +129 -0
- package/multi-agent-workspace/scripts/index_memory.js +104 -0
- package/multi-agent-workspace/scripts/path-boundary.js +59 -0
- package/multi-agent-workspace/scripts/project-docs-resolver.js +357 -0
- package/multi-agent-workspace/scripts/scaffold.js +408 -0
- package/multi-agent-workspace/templates/AGENTS.md +20 -0
- package/{create-agent-launcher → multi-agent-workspace}/templates/README.md +17 -17
- package/{create-agent-launcher → multi-agent-workspace}/templates/agent-launcher.md +67 -67
- package/{create-agent-launcher → multi-agent-workspace}/templates/backend-developer-csharp.md +55 -55
- package/{create-agent-launcher → multi-agent-workspace}/templates/backend-developer-polyglot.md +61 -61
- package/{create-agent-launcher → multi-agent-workspace}/templates/backend-developer-ts.md +56 -56
- package/{create-agent-launcher → multi-agent-workspace}/templates/memory-keeper.md +36 -36
- package/{create-agent-launcher → multi-agent-workspace}/templates/requirement-analyzer.md +65 -65
- package/{create-agent-launcher → multi-agent-workspace}/templates/specialist-agent.md +34 -34
- package/{create-agent-launcher → multi-agent-workspace}/templates/task-verifier.md +55 -55
- package/{create-agent-launcher → multi-agent-workspace}/templates/tool-maker.md +32 -32
- package/multi-agent-workspace/workflows/01-init.md +22 -0
- package/multi-agent-workspace/workflows/02-analysis.md +23 -0
- package/multi-agent-workspace/workflows/03-generation.md +25 -0
- package/multi-agent-workspace/workflows/04-launcher.md +16 -0
- package/opencode-plugin/README.md +137 -0
- package/opencode-plugin/hooks/verify.js +190 -0
- package/opencode-plugin/index.mjs +547 -0
- package/opencode-plugin/plugin.json +9 -0
- package/package.json +63 -22
- package/plugins/harness-everything/.codex-plugin/plugin.json +39 -0
- package/plugins/harness-everything/hooks/hooks.json +197 -0
- package/plugins/harness-everything/hooks/scripts/action-gate-rules.json +77 -0
- package/plugins/harness-everything/hooks/scripts/action-gate.js +528 -0
- package/plugins/harness-everything/hooks/scripts/atomic-commit-check.js +68 -0
- package/plugins/harness-everything/hooks/scripts/boundary-guard.js +55 -0
- package/plugins/harness-everything/hooks/scripts/codex-action-gate-post.js +46 -0
- package/plugins/harness-everything/hooks/scripts/codex-permission-request.js +140 -0
- package/plugins/harness-everything/hooks/scripts/context-compact.js +115 -0
- package/plugins/harness-everything/hooks/scripts/contract-test.js +146 -0
- package/plugins/harness-everything/hooks/scripts/depth-guard.js +65 -0
- package/plugins/harness-everything/hooks/scripts/lib/fable-contracts.js +121 -0
- package/plugins/harness-everything/hooks/scripts/lib/harness-state.js +275 -0
- package/plugins/harness-everything/hooks/scripts/lib/platforms/claude.js +168 -0
- package/plugins/harness-everything/hooks/scripts/lib/platforms/codex.js +88 -0
- package/plugins/harness-everything/hooks/scripts/lib/platforms/continue.js +91 -0
- package/plugins/harness-everything/hooks/scripts/lib/platforms/copilot.js +85 -0
- package/plugins/harness-everything/hooks/scripts/lib/platforms/cursor.js +89 -0
- package/plugins/harness-everything/hooks/scripts/lib/platforms/hermes.js +108 -0
- package/plugins/harness-everything/hooks/scripts/lib/platforms/index.js +17 -0
- package/plugins/harness-everything/hooks/scripts/lib/platforms/worktrees.js +20 -0
- package/plugins/harness-everything/hooks/scripts/lib/workspace.js +290 -0
- package/plugins/harness-everything/hooks/scripts/openai-session-start.js +81 -0
- package/plugins/harness-everything/hooks/scripts/reset-circuit-breaker.js +30 -0
- package/plugins/harness-everything/hooks/scripts/rule-of-3-tracker.js +125 -0
- package/plugins/harness-everything/hooks/scripts/rule-of-3.js +195 -0
- package/plugins/harness-everything/hooks/scripts/state-persist.js +100 -0
- package/plugins/harness-everything/hooks/scripts/stop-gate.js +80 -0
- package/plugins/harness-everything/hooks/scripts/subagent-scope-guard.js +194 -0
- package/plugins/harness-everything/hooks/scripts/subagent-start.js +19 -0
- package/plugins/harness-everything/hooks/scripts/subagent-stop.js +6 -0
- package/plugins/harness-everything/hooks/session-start.js +81 -0
- package/plugins/harness-everything/plugin.json +43 -0
- package/plugins/harness-everything/scripts/lib/advisory-text.js +174 -0
- package/plugins/harness-everything/skills/environment-detection/SKILL.md +41 -0
- package/plugins/harness-everything/skills/environment-detection/references/shell-syntax-rules.md +76 -0
- package/plugins/harness-everything/skills/environment-detection/scripts/preflight.js +103 -0
- package/plugins/harness-everything/skills/eval-harness/SKILL.md +42 -0
- package/plugins/harness-everything/skills/eval-harness/references/scoring-rubric.md +68 -0
- package/plugins/harness-everything/skills/eval-harness/scripts/evaluate.js +79 -0
- package/plugins/harness-everything/skills/fable-discipline/SKILL.md +41 -0
- package/plugins/harness-everything/skills/fable-discipline/references/discipline-rules.md +22 -0
- package/plugins/harness-everything/skills/fable-mode/CONTRACT-FORMAT.md +165 -0
- package/plugins/harness-everything/skills/fable-mode/SKILL.md +39 -0
- package/plugins/harness-everything/skills/fable-mode/agents/fable-orchestrator.md +159 -0
- package/plugins/harness-everything/skills/fable-mode/agents/fable-verifier.md +32 -0
- package/plugins/harness-everything/skills/fable-mode/agents/fable-worker-haiku.md +52 -0
- package/plugins/harness-everything/skills/fable-mode/agents/fable-worker-sonnet.md +53 -0
- package/plugins/harness-everything/skills/fable-mode/execution-guardrails/SKILL.md +43 -0
- package/plugins/harness-everything/skills/fable-mode/execution-guardrails/references/guardrail-rules.md +56 -0
- package/plugins/harness-everything/skills/fable-mode/fable-haiku/SKILL.md +36 -0
- package/plugins/harness-everything/skills/fable-mode/fable-haiku/references/routing-notes.md +26 -0
- package/plugins/harness-everything/skills/fable-mode/fable-opus/SKILL.md +36 -0
- package/plugins/harness-everything/skills/fable-mode/fable-sonnet/SKILL.md +36 -0
- package/plugins/harness-everything/skills/fable-mode/model-matrix.json +32 -0
- package/plugins/harness-everything/skills/fable-mode/references/execution-phases.md +37 -0
- package/plugins/harness-everything/skills/fable-mode/references/model-matrix.md +39 -0
- package/plugins/harness-everything/skills/fable-mode/scripts/ensemble-review.js +371 -0
- package/plugins/harness-everything/skills/fable-mode/scripts/model-selector.js +260 -0
- package/plugins/harness-everything/skills/fable-mode/scripts/workflow-plan-consumer.js +337 -0
- package/plugins/harness-everything/skills/find-skills/SKILL.md +48 -0
- package/plugins/harness-everything/skills/find-skills/references/discovery-flow.md +102 -0
- package/plugins/harness-everything/skills/find-skills/scripts/use-skill.js +112 -0
- package/plugins/harness-everything/skills/git-commit/SKILL.md +38 -0
- package/plugins/harness-everything/skills/git-commit/guides/ANGULAR_STYLE.md +61 -0
- package/plugins/harness-everything/skills/git-commit/guides/COMMIT_GENERATION.md +24 -0
- package/plugins/harness-everything/skills/git-commit/guides/LANGUAGE_DETECTION.md +16 -0
- package/plugins/harness-everything/skills/git-commit/guides/MAIN_REPO.md +15 -0
- package/plugins/harness-everything/skills/git-commit/guides/SUBMODULES.md +27 -0
- package/plugins/harness-everything/skills/git-commit/references/commit-flow.md +30 -0
- package/plugins/harness-everything/skills/grill-me/SKILL.md +43 -0
- package/plugins/harness-everything/skills/grill-me/references/grilling-playbook.md +38 -0
- package/plugins/harness-everything/skills/grill-with-docs/ADR-FORMAT.md +47 -0
- package/plugins/harness-everything/skills/grill-with-docs/CONTEXT-FORMAT.md +63 -0
- package/plugins/harness-everything/skills/grill-with-docs/SKILL.md +41 -0
- package/plugins/harness-everything/skills/grill-with-docs/references/session-playbook.md +99 -0
- package/plugins/harness-everything/skills/harness-everything/SKILL.md +43 -0
- package/plugins/harness-everything/skills/harness-everything/examples/README.md +9 -0
- package/plugins/harness-everything/skills/harness-everything/examples/todo-workflow.md +22 -0
- package/plugins/harness-everything/skills/harness-everything/references/router-workflow-plan.md +128 -0
- package/plugins/harness-everything/skills/harness-everything/references/skill-registry.md +39 -0
- package/plugins/harness-everything/skills/harness-everything/references/triage-and-tiers.md +193 -0
- package/plugins/harness-everything/skills/harness-everything/schemas/router-task-shape.schema.json +126 -0
- package/plugins/harness-everything/skills/harness-everything/schemas/router-workflow-plan.schema.json +149 -0
- package/plugins/harness-everything/skills/harness-everything/scripts/bootstrap.js +212 -0
- package/plugins/harness-everything/skills/harness-everything/scripts/ensemble-policy.js +102 -0
- package/plugins/harness-everything/skills/harness-everything/scripts/kernel-router.js +212 -0
- package/plugins/harness-everything/skills/harness-everything/scripts/router-contract.js +612 -0
- package/plugins/harness-everything/skills/harness-everything/scripts/routing-keywords.json +320 -0
- package/plugins/harness-everything/skills/harness-everything/scripts/self-heal.js +250 -0
- package/plugins/harness-everything/skills/harness-everything/scripts/tier-router.js +459 -0
- package/plugins/harness-everything/skills/harness-everything/scripts/verify-gate.js +100 -0
- package/plugins/harness-everything/skills/improve-codebase-architecture/SKILL.md +38 -0
- package/plugins/harness-everything/skills/improve-codebase-architecture/guides/DEEPENING.md +37 -0
- package/plugins/harness-everything/skills/improve-codebase-architecture/guides/HTML-REPORT.md +123 -0
- package/plugins/harness-everything/skills/improve-codebase-architecture/guides/INTERFACE-DESIGN.md +44 -0
- package/plugins/harness-everything/skills/improve-codebase-architecture/guides/LANGUAGE.md +53 -0
- package/plugins/harness-everything/skills/improve-codebase-architecture/guides/refactoring-workflow.md +59 -0
- package/plugins/harness-everything/skills/improve-codebase-architecture/references/README.md +16 -0
- package/plugins/harness-everything/skills/install-cognitive-os/SKILL.md +44 -0
- package/plugins/harness-everything/skills/install-cognitive-os/references/cognitive-loop.md +27 -0
- package/plugins/harness-everything/skills/multi-agent-workspace/NOTICE-agency-agents.md +27 -0
- package/plugins/harness-everything/skills/multi-agent-workspace/SKILL.md +38 -0
- package/plugins/harness-everything/skills/multi-agent-workspace/examples/csharp-monorepo.md +123 -0
- package/plugins/harness-everything/skills/multi-agent-workspace/examples/typescript-cli.md +126 -0
- package/plugins/harness-everything/skills/multi-agent-workspace/guidelines/platform-claude.md +77 -0
- package/plugins/harness-everything/skills/multi-agent-workspace/guidelines/platform-codex.md +146 -0
- package/plugins/harness-everything/skills/multi-agent-workspace/guidelines/platform-continue.md +38 -0
- package/plugins/harness-everything/skills/multi-agent-workspace/guidelines/platform-copilot.md +74 -0
- package/plugins/harness-everything/skills/multi-agent-workspace/guidelines/platform-cursor.md +40 -0
- package/plugins/harness-everything/skills/multi-agent-workspace/guidelines/platform-hermes.md +38 -0
- package/plugins/harness-everything/skills/multi-agent-workspace/guidelines/role-capabilities.md +108 -0
- package/plugins/harness-everything/skills/multi-agent-workspace/references/agency-agents.md +31 -0
- package/plugins/harness-everything/skills/multi-agent-workspace/references/architecture-guide.md +36 -0
- package/plugins/harness-everything/skills/multi-agent-workspace/references/orchestration.md +65 -0
- package/plugins/harness-everything/skills/multi-agent-workspace/scripts/agency-catalog.js +190 -0
- package/plugins/harness-everything/skills/multi-agent-workspace/scripts/consume-workflow-plan.js +129 -0
- package/plugins/harness-everything/skills/multi-agent-workspace/scripts/index_memory.js +104 -0
- package/plugins/harness-everything/skills/multi-agent-workspace/scripts/path-boundary.js +59 -0
- package/plugins/harness-everything/skills/multi-agent-workspace/scripts/project-docs-resolver.js +357 -0
- package/plugins/harness-everything/skills/multi-agent-workspace/scripts/scaffold.js +408 -0
- package/plugins/harness-everything/skills/multi-agent-workspace/templates/AGENTS.md +20 -0
- package/plugins/harness-everything/skills/multi-agent-workspace/templates/README.md +17 -0
- package/plugins/harness-everything/skills/multi-agent-workspace/templates/agent-launcher.md +68 -0
- package/plugins/harness-everything/skills/multi-agent-workspace/templates/backend-developer-csharp.md +56 -0
- package/plugins/harness-everything/skills/multi-agent-workspace/templates/backend-developer-polyglot.md +62 -0
- package/plugins/harness-everything/skills/multi-agent-workspace/templates/backend-developer-ts.md +57 -0
- package/plugins/harness-everything/skills/multi-agent-workspace/templates/memory-keeper.md +37 -0
- package/plugins/harness-everything/skills/multi-agent-workspace/templates/requirement-analyzer.md +66 -0
- package/plugins/harness-everything/skills/multi-agent-workspace/templates/specialist-agent.md +34 -0
- package/plugins/harness-everything/skills/multi-agent-workspace/templates/task-verifier.md +56 -0
- package/plugins/harness-everything/skills/multi-agent-workspace/templates/tool-maker.md +33 -0
- package/plugins/harness-everything/skills/multi-agent-workspace/workflows/01-init.md +22 -0
- package/plugins/harness-everything/skills/multi-agent-workspace/workflows/02-analysis.md +23 -0
- package/plugins/harness-everything/skills/multi-agent-workspace/workflows/03-generation.md +25 -0
- package/plugins/harness-everything/skills/multi-agent-workspace/workflows/04-launcher.md +16 -0
- package/plugins/harness-everything/skills/repo-docs/SKILL.md +42 -0
- package/plugins/harness-everything/skills/repo-docs/references/process-guide.md +60 -0
- package/plugins/harness-everything/skills/repo-docs/templates/agents-template.md +264 -0
- package/plugins/harness-everything/skills/repo-docs/templates/knowledge-agents-template.md +50 -0
- package/plugins/harness-everything/skills/repo-docs/templates/knowledge-readme-template.md +83 -0
- package/plugins/harness-everything/skills/repo-docs/templates/manual-recon.md +117 -0
- package/plugins/harness-everything/skills/repo-docs/templates/multi-skills-readme-template.md +72 -0
- package/plugins/harness-everything/skills/repo-docs/templates/product-agents-template.md +78 -0
- package/plugins/harness-everything/skills/repo-docs/templates/product-readme-template.md +95 -0
- package/plugins/harness-everything/skills/repo-docs/templates/readme-template.md +136 -0
- package/plugins/harness-everything/skills/repo-docs/templates/single-skill-readme-template.md +70 -0
- package/plugins/harness-everything/skills/repo-docs/templates/skills-agents-template.md +60 -0
- package/plugins/harness-everything/skills/rewrite-commits/SKILL.md +37 -0
- package/plugins/harness-everything/skills/rewrite-commits/references/workflow.md +18 -0
- package/plugins/harness-everything/skills/security-review/SKILL.md +47 -0
- package/plugins/harness-everything/skills/security-review/guides/OWASP-PATTERNS.md +31 -0
- package/plugins/harness-everything/skills/security-review/guides/STRIDE-THREAT-MODEL.md +21 -0
- package/plugins/harness-everything/skills/security-review/references/security-checklist.md +478 -0
- package/plugins/harness-everything/skills/security-review/scripts/audit-secrets.js +85 -0
- package/plugins/harness-everything/skills/self-evolve/SKILL.md +45 -0
- package/plugins/harness-everything/skills/self-evolve/references/memory-resolution.md +61 -0
- package/plugins/harness-everything/skills/self-evolve/scripts/persist-memory.js +272 -0
- package/plugins/harness-everything/skills/self-evolve/scripts/register-dynamic-skill.js +253 -0
- package/plugins/harness-everything/skills/self-evolve/scripts/self-regression.js +125 -0
- package/plugins/harness-everything/skills/skill-creator/SKILL.md +47 -0
- package/plugins/harness-everything/skills/skill-creator/references/authoring-workflow.md +14 -0
- package/plugins/harness-everything/skills/skill-creator/references/dynamic-generation-contract.md +21 -0
- package/plugins/harness-everything/skills/skill-creator/references/quality-checklist.md +16 -0
- package/plugins/harness-everything/skills/skill-creator/references/quality-principles.md +81 -0
- package/plugins/harness-everything/skills/skill-creator/references/testing-workflow.md +27 -0
- package/plugins/harness-everything/skills/skill-style/SKILL.md +42 -0
- package/plugins/harness-everything/skills/skill-style/references/style-guide.md +62 -0
- package/plugins/harness-everything/skills/tdd/SKILL.md +38 -0
- package/plugins/harness-everything/skills/tdd/fixtures/pass-evidence.json +100 -0
- package/plugins/harness-everything/skills/tdd/fixtures/quality-scenarios.json +20 -0
- package/plugins/harness-everything/skills/tdd/guides/deep-modules.md +33 -0
- package/plugins/harness-everything/skills/tdd/guides/interface-design.md +31 -0
- package/plugins/harness-everything/skills/tdd/guides/mocking.md +59 -0
- package/plugins/harness-everything/skills/tdd/guides/refactoring.md +10 -0
- package/plugins/harness-everything/skills/tdd/guides/tests.md +61 -0
- package/plugins/harness-everything/skills/tdd/references/core-discipline.md +51 -0
- package/plugins/harness-everything/skills/tdd/references/integration-testing.md +13 -0
- package/plugins/harness-everything/skills/tdd/references/quality-model.md +68 -0
- package/plugins/harness-everything/skills/tdd/references/unit-testing.md +15 -0
- package/plugins/harness-everything/skills/tdd/schemas/quality-evidence.schema.json +211 -0
- package/plugins/harness-everything/skills/tdd/schemas/quality-report.schema.json +110 -0
- package/plugins/harness-everything/skills/tdd/scripts/quality-gate.js +475 -0
- package/plugins/harness-everything/skills/to-spec/SKILL.md +40 -0
- package/plugins/harness-everything/skills/to-spec/references/process.md +77 -0
- package/plugins/harness-everything/skills/to-spec/scripts/check-project-docs.js +239 -0
- package/plugins/harness-everything/skills/to-spec/scripts/project-docs-resolver.js +111 -0
- package/plugins/harness-everything/skills/to-spec/templates/cli-reference.md +31 -0
- package/plugins/harness-everything/skills/to-spec/templates/dev-doc.md +27 -0
- package/plugins/harness-everything/skills/to-spec/templates/feature-spec.md +59 -0
- package/plugins/harness-everything/skills/to-spec/templates/schema-doc.md +31 -0
- package/plugins/harness-everything/skills/to-tickets/SKILL.md +42 -0
- package/plugins/harness-everything/skills/to-tickets/references/publishing-and-templates.md +97 -0
- package/plugins/harness-everything/skills/todo-driven-workflow/SKILL.md +42 -0
- package/plugins/harness-everything/skills/todo-driven-workflow/references/execution-guide.md +27 -0
- package/plugins/harness-everything/skills/using-git-worktrees/SKILL.md +49 -0
- package/plugins/harness-everything/skills/using-git-worktrees/references/workflow-details.md +108 -0
- package/plugins/harness-everything/skills/verification-loop/SKILL.md +46 -0
- package/plugins/harness-everything/skills/verification-loop/references/verification-phases.md +116 -0
- package/plugins/harness-everything/skills/verification-loop/templates/verification-report.template.md +15 -0
- package/plugins/harness-everything/skills/verify-before-claim/SKILL.md +43 -0
- package/plugins/harness-everything/skills/verify-before-claim/references/verification-guide.md +73 -0
- package/plugins/harness-everything/skills/zoom-out/SKILL.md +42 -0
- package/plugins/harness-everything/skills/zoom-out/references/circuit-breaker.md +66 -0
- package/plugins/harness-everything/skills/zoom-out/templates/zoom-out-report.template.md +20 -0
- package/references/definition-of-done.md +67 -0
- package/references/performance-checklist.md +153 -0
- package/references/platform-advisory-template.md +58 -0
- package/references/security-checklist.md +205 -0
- package/repo-docs/SKILL.md +34 -37
- package/repo-docs/references/process-guide.md +60 -0
- package/repo-docs/templates/agents-template.md +264 -264
- package/repo-docs/templates/knowledge-agents-template.md +50 -50
- package/repo-docs/templates/knowledge-readme-template.md +83 -83
- package/repo-docs/templates/manual-recon.md +117 -117
- package/repo-docs/templates/multi-skills-readme-template.md +72 -72
- package/repo-docs/templates/product-agents-template.md +78 -78
- package/repo-docs/templates/product-readme-template.md +95 -95
- package/repo-docs/templates/readme-template.md +136 -136
- package/repo-docs/templates/single-skill-readme-template.md +70 -70
- package/repo-docs/templates/skills-agents-template.md +60 -60
- package/rewrite-commits/SKILL.md +22 -17
- package/rewrite-commits/references/workflow.md +18 -0
- package/scripts/build-openai-submission.js +158 -0
- package/scripts/installer.js +621 -506
- package/scripts/lib/advisory-text.js +33 -22
- package/scripts/lib/claude-hooks.js +18 -5
- package/scripts/lib/execution-contract.js +70 -0
- package/scripts/lib/gitignore.js +199 -80
- package/scripts/lib/manifest.js +83 -23
- package/scripts/lib/path-boundary.js +90 -0
- package/scripts/lib/prompts.js +1 -1
- package/scripts/lib/skills.js +191 -51
- package/scripts/lib/workspace.js +269 -7
- package/scripts/sync-openai-plugin.js +91 -0
- package/scripts/sync-release-version.js +212 -0
- package/scripts/verify-install.js +147 -0
- package/security-review/SKILL.md +33 -497
- package/security-review/guides/OWASP-PATTERNS.md +31 -0
- package/security-review/guides/STRIDE-THREAT-MODEL.md +21 -0
- package/security-review/references/security-checklist.md +478 -0
- package/security-review/scripts/audit-secrets.js +85 -0
- package/self-evolve/SKILL.md +36 -45
- package/self-evolve/references/memory-resolution.md +61 -0
- package/self-evolve/scripts/persist-memory.js +272 -40
- package/self-evolve/scripts/register-dynamic-skill.js +253 -0
- package/self-evolve/scripts/self-regression.js +125 -113
- package/skill-creator/SKILL.md +33 -77
- package/skill-creator/references/authoring-workflow.md +14 -0
- package/skill-creator/references/dynamic-generation-contract.md +21 -0
- package/skill-creator/references/quality-checklist.md +16 -0
- package/skill-creator/references/quality-principles.md +81 -81
- package/skill-creator/references/testing-workflow.md +27 -32
- package/skill-style/SKILL.md +33 -39
- package/skill-style/references/style-guide.md +62 -0
- package/submission/openai/README.md +67 -0
- package/submission/openai/listing.json +31 -0
- package/submission/openai/test-cases.json +61 -0
- package/tdd/SKILL.md +30 -45
- package/tdd/fixtures/pass-evidence.json +100 -0
- package/tdd/fixtures/quality-scenarios.json +20 -0
- package/tdd/guides/deep-modules.md +33 -33
- package/tdd/guides/interface-design.md +31 -31
- package/tdd/guides/mocking.md +59 -59
- package/tdd/guides/refactoring.md +10 -10
- package/tdd/guides/tests.md +61 -61
- package/tdd/references/core-discipline.md +51 -0
- package/tdd/references/integration-testing.md +13 -0
- package/tdd/references/quality-model.md +68 -0
- package/tdd/references/unit-testing.md +15 -0
- package/tdd/schemas/quality-evidence.schema.json +211 -0
- package/tdd/schemas/quality-report.schema.json +110 -0
- package/tdd/scripts/quality-gate.js +475 -0
- package/to-spec/SKILL.md +40 -0
- package/to-spec/references/process.md +77 -0
- package/to-spec/scripts/check-project-docs.js +239 -0
- package/to-spec/scripts/project-docs-resolver.js +111 -0
- package/to-spec/templates/cli-reference.md +31 -0
- package/to-spec/templates/dev-doc.md +27 -0
- package/to-spec/templates/feature-spec.md +59 -0
- package/to-spec/templates/schema-doc.md +31 -0
- package/to-tickets/SKILL.md +42 -0
- package/to-tickets/references/publishing-and-templates.md +97 -0
- package/todo-driven-workflow/SKILL.md +28 -40
- package/todo-driven-workflow/references/execution-guide.md +27 -0
- package/using-git-worktrees/SKILL.md +29 -141
- package/using-git-worktrees/references/workflow-details.md +108 -0
- package/verification-loop/SKILL.md +29 -121
- package/verification-loop/references/verification-phases.md +116 -0
- package/verification-loop/templates/verification-report.template.md +15 -0
- package/verify-before-claim/SKILL.md +27 -70
- package/verify-before-claim/references/verification-guide.md +73 -0
- package/zoom-out/SKILL.md +28 -76
- package/zoom-out/references/circuit-breaker.md +66 -0
- package/zoom-out/templates/zoom-out-report.template.md +20 -0
- package/build-multi-agent-system/AGENTS.md +0 -50
- package/build-multi-agent-system/README.md +0 -63
- package/build-multi-agent-system/SKILL.md +0 -68
- package/create-agent-launcher/SKILL.md +0 -48
- package/create-agent-launcher/guidelines/platform-gemini.md +0 -101
- package/create-agent-launcher/workflows/01-init.md +0 -22
- package/create-agent-launcher/workflows/02-analysis.md +0 -26
- package/create-agent-launcher/workflows/03-generation.md +0 -29
- package/create-agent-launcher/workflows/04-launcher.md +0 -20
- package/docs/workflows/build-multi-agent-system.md +0 -62
- package/docs/workflows/create-agent-launcher.md +0 -62
- package/eval-framework/behavioral-test.js +0 -105
- package/eval-framework/cases/case2.json +0 -5
- package/eval-framework/mechanism-test.js +0 -251
- package/harness-everything/scripts/todo-cli.js +0 -157
- package/skill-style/STYLE.md +0 -37
|
@@ -0,0 +1,80 @@
|
|
|
1
|
+
#!/usr/bin/env node
|
|
2
|
+
'use strict';
|
|
3
|
+
|
|
4
|
+
const assert = require('assert');
|
|
5
|
+
const crypto = require('crypto');
|
|
6
|
+
const { spawnSync } = require('child_process');
|
|
7
|
+
const fs = require('fs');
|
|
8
|
+
const os = require('os');
|
|
9
|
+
const path = require('path');
|
|
10
|
+
|
|
11
|
+
const ROOT = path.resolve(__dirname, '..');
|
|
12
|
+
|
|
13
|
+
function treeDigest(root) {
|
|
14
|
+
const hash = crypto.createHash('sha256');
|
|
15
|
+
function visit(dir, relative = '') {
|
|
16
|
+
const entries = fs.readdirSync(dir, { withFileTypes: true }).sort((a, b) => a.name.localeCompare(b.name));
|
|
17
|
+
for (const entry of entries) {
|
|
18
|
+
const rel = relative ? `${relative}/${entry.name}` : entry.name;
|
|
19
|
+
const full = path.join(dir, entry.name);
|
|
20
|
+
hash.update(`${entry.isDirectory() ? 'D' : 'F'}:${rel}\0`);
|
|
21
|
+
if (entry.isDirectory()) visit(full, rel);
|
|
22
|
+
else if (entry.isFile()) hash.update(fs.readFileSync(full));
|
|
23
|
+
}
|
|
24
|
+
}
|
|
25
|
+
visit(root);
|
|
26
|
+
return hash.digest('hex');
|
|
27
|
+
}
|
|
28
|
+
|
|
29
|
+
const tempBase = fs.mkdtempSync(path.join(os.tmpdir(), 'harness-plugin-sync-'));
|
|
30
|
+
const copyRoot = path.join(tempBase, 'repo');
|
|
31
|
+
|
|
32
|
+
try {
|
|
33
|
+
fs.cpSync(ROOT, copyRoot, {
|
|
34
|
+
recursive: true,
|
|
35
|
+
filter(source) {
|
|
36
|
+
const rel = path.relative(ROOT, source).replace(/\\/g, '/');
|
|
37
|
+
if (!rel) return true;
|
|
38
|
+
return !(
|
|
39
|
+
rel === '.git' || rel.startsWith('.git/') ||
|
|
40
|
+
rel === 'node_modules' || rel.startsWith('node_modules/') ||
|
|
41
|
+
rel === '.harness' || rel.startsWith('.harness/')
|
|
42
|
+
);
|
|
43
|
+
},
|
|
44
|
+
});
|
|
45
|
+
|
|
46
|
+
const pluginRoot = path.join(copyRoot, 'plugins', 'harness-everything');
|
|
47
|
+
const before = treeDigest(pluginRoot);
|
|
48
|
+
const run = spawnSync(process.execPath, [path.join(copyRoot, 'scripts', 'sync-openai-plugin.js')], {
|
|
49
|
+
cwd: copyRoot,
|
|
50
|
+
encoding: 'utf8',
|
|
51
|
+
});
|
|
52
|
+
assert.strictEqual(run.status, 0, `plugin:sync equivalent failed:\n${run.stdout}\n${run.stderr}`);
|
|
53
|
+
const after = treeDigest(pluginRoot);
|
|
54
|
+
assert.strictEqual(after, before, 'sync-openai-plugin.js must be idempotent against the committed plugin package');
|
|
55
|
+
|
|
56
|
+
const hooks = JSON.parse(fs.readFileSync(path.join(pluginRoot, 'hooks', 'hooks.json'), 'utf8'));
|
|
57
|
+
const referenced = new Set();
|
|
58
|
+
for (const groups of Object.values(hooks.hooks || {})) {
|
|
59
|
+
for (const group of groups || []) {
|
|
60
|
+
for (const hook of group.hooks || []) {
|
|
61
|
+
for (const field of ['command', 'commandWindows']) {
|
|
62
|
+
const command = hook[field];
|
|
63
|
+
if (typeof command !== 'string') continue;
|
|
64
|
+
const match = command.match(/(?:\$PLUGIN_ROOT|%PLUGIN_ROOT%)[\\/]([^"']+)/);
|
|
65
|
+
if (match) referenced.add(match[1].replace(/\\/g, '/'));
|
|
66
|
+
}
|
|
67
|
+
}
|
|
68
|
+
}
|
|
69
|
+
}
|
|
70
|
+
for (const relative of referenced) {
|
|
71
|
+
assert.ok(fs.existsSync(path.join(pluginRoot, relative)), `sync must preserve every hook target: ${relative}`);
|
|
72
|
+
}
|
|
73
|
+
|
|
74
|
+
assert.ok(fs.existsSync(path.join(pluginRoot, 'hooks', 'scripts', 'codex-action-gate-post.js')), 'sync must preserve the Codex PostToolUse adapter');
|
|
75
|
+
assert.ok(fs.existsSync(path.join(pluginRoot, 'hooks', 'scripts', 'codex-permission-request.js')), 'sync must preserve the Codex PermissionRequest adapter');
|
|
76
|
+
|
|
77
|
+
console.log('PASS: OpenAI/Codex plugin sync is idempotent and preserves every referenced hook runtime file.');
|
|
78
|
+
} finally {
|
|
79
|
+
fs.rmSync(tempBase, { recursive: true, force: true });
|
|
80
|
+
}
|
|
@@ -0,0 +1,16 @@
|
|
|
1
|
+
const fs = require('fs');
|
|
2
|
+
const helper = require('./test-helper');
|
|
3
|
+
|
|
4
|
+
console.log('\n[2b] Boundary guard...');
|
|
5
|
+
|
|
6
|
+
const bigFile = helper.tempFile('.mechanism-test-big.tmp');
|
|
7
|
+
fs.writeFileSync(bigFile, 'x'.repeat(600 * 1024));
|
|
8
|
+
const boundaryResult = helper.runHook('boundary-guard.js', { tool_name: 'Read', tool_input: { file_path: bigFile } });
|
|
9
|
+
helper.check(
|
|
10
|
+
'2b. Blocks a 600KB Read without limit/offset (exit=2, BLOCKED)',
|
|
11
|
+
boundaryResult.code === 2 && boundaryResult.stderr.includes('[Boundary Guard] BLOCKED'),
|
|
12
|
+
`Got exit=${boundaryResult.code}, stderr="${boundaryResult.stderr.slice(0, 200)}"`
|
|
13
|
+
);
|
|
14
|
+
fs.unlinkSync(bigFile);
|
|
15
|
+
|
|
16
|
+
helper.finish();
|
|
@@ -0,0 +1,29 @@
|
|
|
1
|
+
const helper = require('./test-helper');
|
|
2
|
+
|
|
3
|
+
console.log('\n[2c] State persistence (WAL)...');
|
|
4
|
+
|
|
5
|
+
helper.runHook('state-persist.js', {
|
|
6
|
+
session_id: helper.SESSION_ID,
|
|
7
|
+
tool_name: 'Bash',
|
|
8
|
+
tool_response: { stdout: '', stderr: 'npm ERR! mech-test failure' },
|
|
9
|
+
});
|
|
10
|
+
const failedHandoff = helper.readState('handoff-state.json');
|
|
11
|
+
helper.check(
|
|
12
|
+
'2c. Failing Bash call is recorded as status=failed',
|
|
13
|
+
failedHandoff.status === 'failed' && failedHandoff.errorSummary.includes('mech-test failure'),
|
|
14
|
+
`Got ${JSON.stringify(failedHandoff)}`
|
|
15
|
+
);
|
|
16
|
+
|
|
17
|
+
helper.runHook('state-persist.js', {
|
|
18
|
+
session_id: helper.SESSION_ID,
|
|
19
|
+
tool_name: 'Bash',
|
|
20
|
+
tool_response: { stdout: 'ok', exitCode: 0 },
|
|
21
|
+
});
|
|
22
|
+
const resolvedHandoff = helper.readState('handoff-state.json');
|
|
23
|
+
helper.check(
|
|
24
|
+
'2c. A subsequent successful command clears status back to idle',
|
|
25
|
+
resolvedHandoff.status === 'idle' && !!resolvedHandoff.lastResolved,
|
|
26
|
+
`Got ${JSON.stringify(resolvedHandoff)}`
|
|
27
|
+
);
|
|
28
|
+
|
|
29
|
+
helper.finish();
|
|
@@ -0,0 +1,18 @@
|
|
|
1
|
+
const { execSync } = require('child_process');
|
|
2
|
+
const path = require('path');
|
|
3
|
+
const helper = require('./test-helper');
|
|
4
|
+
|
|
5
|
+
console.log('\n[2d] Fact-audit reminder...');
|
|
6
|
+
|
|
7
|
+
const tierRouterPath = path.join(helper.root, 'harness-everything', 'scripts', 'tier-router.js');
|
|
8
|
+
const factAuditOut = execSync(`node "${tierRouterPath}"`, {
|
|
9
|
+
encoding: 'utf8',
|
|
10
|
+
input: JSON.stringify({ prompt: 'what exit code does this hook use by default and is it documented' }),
|
|
11
|
+
});
|
|
12
|
+
helper.check(
|
|
13
|
+
'2d. tier-router.js reads {"prompt":...} from stdin and emits FACT-AUDIT REMINDER',
|
|
14
|
+
factAuditOut.includes('FACT-AUDIT REMINDER'),
|
|
15
|
+
`Output was:\n${factAuditOut.slice(0, 300)}`
|
|
16
|
+
);
|
|
17
|
+
|
|
18
|
+
helper.finish();
|
|
@@ -0,0 +1,24 @@
|
|
|
1
|
+
const fs = require('fs');
|
|
2
|
+
const path = require('path');
|
|
3
|
+
const helper = require('./test-helper');
|
|
4
|
+
|
|
5
|
+
console.log('\n[2e] Subagent scope guard...');
|
|
6
|
+
|
|
7
|
+
helper.runHook('subagent-scope-guard.js', {
|
|
8
|
+
session_id: helper.SESSION_ID, tool_name: 'Task', hook_event_name: 'PreToolUse', tool_input: {},
|
|
9
|
+
});
|
|
10
|
+
const scopeFile = helper.tempFile('.mechanism-test-scope.tmp');
|
|
11
|
+
fs.writeFileSync(scopeFile, 'unexpected change');
|
|
12
|
+
const scopeResult = helper.runHook('subagent-scope-guard.js', {
|
|
13
|
+
session_id: helper.SESSION_ID, tool_name: 'Task', hook_event_name: 'PostToolUse', tool_input: {},
|
|
14
|
+
});
|
|
15
|
+
helper.check(
|
|
16
|
+
'2e. Flags the newly-created file after a Task burst (exit=2, lists filename)',
|
|
17
|
+
scopeResult.code === 2 &&
|
|
18
|
+
scopeResult.stderr.includes('Subagent Scope Guard') &&
|
|
19
|
+
scopeResult.stderr.includes(path.basename(scopeFile)),
|
|
20
|
+
`Got exit=${scopeResult.code}, stderr="${scopeResult.stderr.slice(0, 300)}"`
|
|
21
|
+
);
|
|
22
|
+
fs.unlinkSync(scopeFile);
|
|
23
|
+
|
|
24
|
+
helper.finish();
|
|
@@ -0,0 +1,23 @@
|
|
|
1
|
+
const fs = require('fs');
|
|
2
|
+
const helper = require('./test-helper');
|
|
3
|
+
|
|
4
|
+
console.log('\n[2f] Stop gate...');
|
|
5
|
+
|
|
6
|
+
helper.writeState('handoff-state.json', { status: 'idle', lastEditAt: Date.now(), lastVerifyAt: 0 });
|
|
7
|
+
const dirtyFile = helper.tempFile('.mechanism-test-dirty.tmp');
|
|
8
|
+
fs.writeFileSync(dirtyFile, 'dirty');
|
|
9
|
+
const firstStop = helper.runHook('stop-gate.js', { session_id: helper.SESSION_ID });
|
|
10
|
+
helper.check(
|
|
11
|
+
'2f. First stop after an unverified edit bounces (exit=2)',
|
|
12
|
+
firstStop.code === 2 && firstStop.stderr.includes('[Stop Gate]'),
|
|
13
|
+
`Got exit=${firstStop.code}, stderr="${firstStop.stderr.slice(0, 200)}"`
|
|
14
|
+
);
|
|
15
|
+
const secondStop = helper.runHook('stop-gate.js', { session_id: helper.SESSION_ID });
|
|
16
|
+
helper.check(
|
|
17
|
+
'2f. Same edit batch does not bounce twice (exit=0)',
|
|
18
|
+
secondStop.code === 0,
|
|
19
|
+
`Got exit=${secondStop.code}`
|
|
20
|
+
);
|
|
21
|
+
fs.unlinkSync(dirtyFile);
|
|
22
|
+
|
|
23
|
+
helper.finish();
|
|
@@ -0,0 +1,88 @@
|
|
|
1
|
+
const fs = require('fs');
|
|
2
|
+
const path = require('path');
|
|
3
|
+
const helper = require('./test-helper');
|
|
4
|
+
|
|
5
|
+
console.log('\n[2g] Platform ignore patterns / Gitignore matching...');
|
|
6
|
+
|
|
7
|
+
const mockWs = helper.tempDir('.mechanism-test-mock-ws');
|
|
8
|
+
fs.mkdirSync(mockWs, { recursive: true });
|
|
9
|
+
|
|
10
|
+
const platforms = require('../hooks/scripts/lib/platforms');
|
|
11
|
+
const claudePlatform = platforms.find(p => p.name === 'claude');
|
|
12
|
+
const cursorPlatform = platforms.find(p => p.name === 'cursor');
|
|
13
|
+
|
|
14
|
+
// 1. Claude platform ignore pattern with mixed skills
|
|
15
|
+
const claudeSkillsDir = path.join(mockWs, '.claude', 'skills');
|
|
16
|
+
fs.mkdirSync(path.join(claudeSkillsDir, 'harness-skill'), { recursive: true });
|
|
17
|
+
fs.mkdirSync(path.join(claudeSkillsDir, 'custom-skill'), { recursive: true });
|
|
18
|
+
|
|
19
|
+
fs.writeFileSync(
|
|
20
|
+
path.join(claudeSkillsDir, 'harness-skill', 'SKILL.md'),
|
|
21
|
+
'---\nauthor: Miya Daniel\n---\n',
|
|
22
|
+
'utf8'
|
|
23
|
+
);
|
|
24
|
+
fs.writeFileSync(
|
|
25
|
+
path.join(claudeSkillsDir, 'custom-skill', 'SKILL.md'),
|
|
26
|
+
'---\nauthor: Someone Else\n---\n',
|
|
27
|
+
'utf8'
|
|
28
|
+
);
|
|
29
|
+
|
|
30
|
+
const claudePatterns = claudePlatform.getIgnorePatterns(mockWs);
|
|
31
|
+
helper.check(
|
|
32
|
+
'2g. getIgnorePatterns returns ONLY Harness skills',
|
|
33
|
+
claudePatterns.includes('.claude/harness-everything/') &&
|
|
34
|
+
claudePatterns.includes('.claude/skills/harness-skill/') &&
|
|
35
|
+
!claudePatterns.includes('.claude/skills/custom-skill/'),
|
|
36
|
+
`Got patterns: ${JSON.stringify(claudePatterns)}`
|
|
37
|
+
);
|
|
38
|
+
|
|
39
|
+
// 2. test isMatch for claude parent matching to avoid double appending
|
|
40
|
+
helper.check(
|
|
41
|
+
'2g. isMatch returns true for parent folder match (Claude)',
|
|
42
|
+
claudePlatform.isMatch('.claude/harness-everything/', '.claude') === true &&
|
|
43
|
+
claudePlatform.isMatch('.claude/harness-everything/', '.claude/') === true,
|
|
44
|
+
`isMatch for .claude failed`
|
|
45
|
+
);
|
|
46
|
+
|
|
47
|
+
helper.check(
|
|
48
|
+
'2g. isMatch returns true for exact match of harness skill (Claude)',
|
|
49
|
+
claudePlatform.isMatch('.claude/skills/harness-skill/', '.claude/skills/harness-skill/') === true &&
|
|
50
|
+
claudePlatform.isMatch('.claude/skills/harness-skill/', '.claude/skills/harness-skill') === true,
|
|
51
|
+
`isMatch for exact skill path failed`
|
|
52
|
+
);
|
|
53
|
+
|
|
54
|
+
// 3. Cursor platform ignore pattern with mixed skills
|
|
55
|
+
const cursorSkillsDir = path.join(mockWs, '.cursor', 'skills');
|
|
56
|
+
fs.mkdirSync(path.join(cursorSkillsDir, 'harness-skill'), { recursive: true });
|
|
57
|
+
fs.mkdirSync(path.join(cursorSkillsDir, 'custom-skill'), { recursive: true });
|
|
58
|
+
|
|
59
|
+
fs.writeFileSync(
|
|
60
|
+
path.join(cursorSkillsDir, 'harness-skill', 'SKILL.md'),
|
|
61
|
+
'---\nauthor: Miya Daniel\n---\n',
|
|
62
|
+
'utf8'
|
|
63
|
+
);
|
|
64
|
+
fs.writeFileSync(
|
|
65
|
+
path.join(cursorSkillsDir, 'custom-skill', 'SKILL.md'),
|
|
66
|
+
'---\nauthor: Someone Else\n---\n',
|
|
67
|
+
'utf8'
|
|
68
|
+
);
|
|
69
|
+
|
|
70
|
+
const cursorPatterns = cursorPlatform.getIgnorePatterns(mockWs);
|
|
71
|
+
helper.check(
|
|
72
|
+
'2g. getIgnorePatterns returns ONLY Harness skills (Cursor)',
|
|
73
|
+
cursorPatterns.includes('.cursor/harness-everything/') &&
|
|
74
|
+
cursorPatterns.includes('.cursor/skills/harness-skill/') &&
|
|
75
|
+
!cursorPatterns.includes('.cursor/skills/custom-skill/'),
|
|
76
|
+
`Got patterns: ${JSON.stringify(cursorPatterns)}`
|
|
77
|
+
);
|
|
78
|
+
|
|
79
|
+
helper.check(
|
|
80
|
+
'2g. isMatch returns true for parent folder match (Cursor)',
|
|
81
|
+
cursorPlatform.isMatch('.cursor/harness-everything/', '.cursor') === true &&
|
|
82
|
+
cursorPlatform.isMatch('.cursor/harness-everything/', '.cursor/') === true,
|
|
83
|
+
`isMatch for .cursor failed`
|
|
84
|
+
);
|
|
85
|
+
|
|
86
|
+
fs.rmSync(mockWs, { recursive: true, force: true });
|
|
87
|
+
|
|
88
|
+
helper.finish();
|
|
@@ -0,0 +1,147 @@
|
|
|
1
|
+
const fs = require('fs');
|
|
2
|
+
const path = require('path');
|
|
3
|
+
const helper = require('./test-helper');
|
|
4
|
+
|
|
5
|
+
console.log('\n[2h] Installer manifest and skill bookkeeping...');
|
|
6
|
+
|
|
7
|
+
const manifestMockWs = helper.tempDir('.mechanism-test-manifest-ws');
|
|
8
|
+
fs.mkdirSync(manifestMockWs, { recursive: true });
|
|
9
|
+
|
|
10
|
+
const manifestHelper = require('../scripts/lib/manifest');
|
|
11
|
+
const skillsHelper = require('../scripts/lib/skills');
|
|
12
|
+
const testManifestPath = path.join(manifestMockWs, 'manifest.json');
|
|
13
|
+
|
|
14
|
+
// 1. Verify readManifest of non-existent path
|
|
15
|
+
const initialManifest = manifestHelper.readManifest(testManifestPath);
|
|
16
|
+
helper.check(
|
|
17
|
+
'2h. readManifest returns fallback structure when file does not exist',
|
|
18
|
+
initialManifest.package === 'harness-everything' && Array.isArray(initialManifest.skills) && initialManifest.skills.length === 0,
|
|
19
|
+
`Got: ${JSON.stringify(initialManifest)}`
|
|
20
|
+
);
|
|
21
|
+
|
|
22
|
+
// 2. Verify recordSkillInstall writes and serializes correctly
|
|
23
|
+
manifestHelper.recordSkillInstall(testManifestPath, '1.0.0', 'test-skill', '/path/to/test-skill');
|
|
24
|
+
const writtenManifest = manifestHelper.readManifest(testManifestPath);
|
|
25
|
+
helper.check(
|
|
26
|
+
'2h. recordSkillInstall successfully records skill information',
|
|
27
|
+
writtenManifest.version === '1.0.0' &&
|
|
28
|
+
writtenManifest.skills.length === 1 &&
|
|
29
|
+
writtenManifest.skills[0].id === 'test-skill' &&
|
|
30
|
+
writtenManifest.skills[0].dirPath === '/path/to/test-skill',
|
|
31
|
+
`Got: ${JSON.stringify(writtenManifest)}`
|
|
32
|
+
);
|
|
33
|
+
|
|
34
|
+
// 3. Verify recordGeneratedSkill and removeGeneratedSkill
|
|
35
|
+
manifestHelper.recordGeneratedSkill(testManifestPath, 'gen-skill', '/path/to/gen-skill', 'desc', ['trigger']);
|
|
36
|
+
const withGenerated = manifestHelper.readManifest(testManifestPath);
|
|
37
|
+
helper.check(
|
|
38
|
+
'2h. recordGeneratedSkill successfully records generated skill with triggers',
|
|
39
|
+
withGenerated.generated &&
|
|
40
|
+
withGenerated.generated.length === 1 &&
|
|
41
|
+
withGenerated.generated[0].id === 'gen-skill' &&
|
|
42
|
+
withGenerated.generated[0].triggers.includes('trigger'),
|
|
43
|
+
`Got: ${JSON.stringify(withGenerated)}`
|
|
44
|
+
);
|
|
45
|
+
|
|
46
|
+
manifestHelper.removeGeneratedSkill(testManifestPath, '/path/to/gen-skill');
|
|
47
|
+
const afterRemoveGen = manifestHelper.readManifest(testManifestPath);
|
|
48
|
+
helper.check(
|
|
49
|
+
'2h. removeGeneratedSkill removes generated skill correctly',
|
|
50
|
+
afterRemoveGen.generated && afterRemoveGen.generated.length === 0,
|
|
51
|
+
`Got: ${JSON.stringify(afterRemoveGen)}`
|
|
52
|
+
);
|
|
53
|
+
|
|
54
|
+
// 4. Verify isHarnessSkillDir matches correct author
|
|
55
|
+
const mockSkillPath = path.join(manifestMockWs, 'mock-skill-folder');
|
|
56
|
+
fs.mkdirSync(mockSkillPath, { recursive: true });
|
|
57
|
+
const mockSkillMdPath = path.join(mockSkillPath, 'SKILL.md');
|
|
58
|
+
|
|
59
|
+
fs.writeFileSync(mockSkillMdPath, '---\nauthor: Miya Daniel\n---\n', 'utf8');
|
|
60
|
+
const isHarness = skillsHelper.isHarnessSkillDir(mockSkillPath);
|
|
61
|
+
helper.check(
|
|
62
|
+
'2h. isHarnessSkillDir detects Harness author correctly',
|
|
63
|
+
isHarness === true,
|
|
64
|
+
`Expected true but got false`
|
|
65
|
+
);
|
|
66
|
+
|
|
67
|
+
// Regression: spec-compliant frontmatter (author nested under metadata:) must
|
|
68
|
+
// also be recognized. Guards the getMeta parser in scripts/lib/skills.js.
|
|
69
|
+
fs.writeFileSync(mockSkillMdPath, '---\nname: spec-skill\ndescription: x\nlicense: Apache-2.0\nmetadata:\n author: Miya Daniel\n version: 0.3.3\n---\n', 'utf8');
|
|
70
|
+
const isHarnessSpec = skillsHelper.isHarnessSkillDir(mockSkillPath);
|
|
71
|
+
helper.check(
|
|
72
|
+
'2h. isHarnessSkillDir detects author under metadata: (spec format)',
|
|
73
|
+
isHarnessSpec === true,
|
|
74
|
+
`Expected true but got false`
|
|
75
|
+
);
|
|
76
|
+
|
|
77
|
+
fs.writeFileSync(mockSkillMdPath, '---\nauthor: Outsider\n---\n', 'utf8');
|
|
78
|
+
const isNotHarness = skillsHelper.isHarnessSkillDir(mockSkillPath);
|
|
79
|
+
helper.check(
|
|
80
|
+
'2h. isHarnessSkillDir returns false for foreign authors',
|
|
81
|
+
isNotHarness === false,
|
|
82
|
+
`Expected false but got true`
|
|
83
|
+
);
|
|
84
|
+
|
|
85
|
+
// 5. Verify removeSkillFromManifest removes skill and deletes empty manifest
|
|
86
|
+
manifestHelper.removeSkillFromManifest(testManifestPath, '/path/to/test-skill');
|
|
87
|
+
helper.check(
|
|
88
|
+
'2h. removeSkillFromManifest cleans up or unlinks empty manifest',
|
|
89
|
+
!fs.existsSync(testManifestPath),
|
|
90
|
+
`Manifest file still exists after removing all entries!`
|
|
91
|
+
);
|
|
92
|
+
|
|
93
|
+
// 6. Verify Fable agent definitions install into the native Claude agent
|
|
94
|
+
// directory, preserve a conflicting user file, and are manifest tracked.
|
|
95
|
+
const agentTargetDir = path.join(manifestMockWs, 'agents');
|
|
96
|
+
const agentManifestPath = path.join(manifestMockWs, 'agent-manifest.json');
|
|
97
|
+
fs.mkdirSync(agentTargetDir, { recursive: true });
|
|
98
|
+
const conflictPath = path.join(agentTargetDir, 'fable-orchestrator.md');
|
|
99
|
+
fs.writeFileSync(conflictPath, '---\nname: fable-orchestrator\n---\nuser-owned\n', 'utf8');
|
|
100
|
+
skillsHelper.installAgentsToTarget({
|
|
101
|
+
target: { path: agentTargetDir, label: 'test-agents/', manifestPath: agentManifestPath },
|
|
102
|
+
harnessSourceDir: path.join(__dirname, '..'),
|
|
103
|
+
packageVersion: '1.0.0'
|
|
104
|
+
});
|
|
105
|
+
const installedAgents = manifestHelper.readManifest(agentManifestPath).agents;
|
|
106
|
+
helper.check(
|
|
107
|
+
'2h. Fable agents are manifest tracked without overwriting a conflicting file',
|
|
108
|
+
installedAgents.length === 3 &&
|
|
109
|
+
fs.readFileSync(conflictPath, 'utf8').includes('user-owned') &&
|
|
110
|
+
fs.existsSync(path.join(agentTargetDir, 'fable-verifier.md')),
|
|
111
|
+
`Got: ${JSON.stringify(installedAgents)}`
|
|
112
|
+
);
|
|
113
|
+
|
|
114
|
+
for (const agent of [...installedAgents]) {
|
|
115
|
+
skillsHelper.removeAgent({ ...agent, manifestPath: agentManifestPath });
|
|
116
|
+
}
|
|
117
|
+
helper.check(
|
|
118
|
+
'2h. Removing Fable agents removes only tracked files and manifest state',
|
|
119
|
+
fs.existsSync(conflictPath) && !fs.existsSync(agentManifestPath),
|
|
120
|
+
`Agent conflict or manifest was removed unexpectedly`
|
|
121
|
+
);
|
|
122
|
+
|
|
123
|
+
const identicalPath = path.join(agentTargetDir, 'fable-verifier.md');
|
|
124
|
+
fs.copyFileSync(path.join(__dirname, '..', 'fable-mode', 'agents', 'fable-verifier.md'), identicalPath);
|
|
125
|
+
skillsHelper.installAgentsToTarget({
|
|
126
|
+
target: { path: agentTargetDir, label: 'test-agents/', manifestPath: agentManifestPath },
|
|
127
|
+
harnessSourceDir: path.join(__dirname, '..'),
|
|
128
|
+
packageVersion: '1.0.0'
|
|
129
|
+
});
|
|
130
|
+
const secondInstall = manifestHelper.readManifest(agentManifestPath).agents;
|
|
131
|
+
helper.check(
|
|
132
|
+
'2h. Pre-existing identical agent is not claimed as Harness-owned',
|
|
133
|
+
secondInstall.length === 2,
|
|
134
|
+
`Got: ${JSON.stringify(secondInstall)}`
|
|
135
|
+
);
|
|
136
|
+
for (const agent of [...secondInstall]) {
|
|
137
|
+
skillsHelper.removeAgent({ ...agent, manifestPath: agentManifestPath });
|
|
138
|
+
}
|
|
139
|
+
helper.check(
|
|
140
|
+
'2h. Pre-existing identical agent survives tracked-agent removal',
|
|
141
|
+
fs.existsSync(conflictPath) && fs.existsSync(identicalPath),
|
|
142
|
+
`A pre-existing agent was removed unexpectedly`
|
|
143
|
+
);
|
|
144
|
+
|
|
145
|
+
fs.rmSync(manifestMockWs, { recursive: true, force: true });
|
|
146
|
+
|
|
147
|
+
helper.finish();
|
|
@@ -0,0 +1,145 @@
|
|
|
1
|
+
#!/usr/bin/env node
|
|
2
|
+
/**
|
|
3
|
+
* [2i] CRLF frontmatter immunity for quality gates.
|
|
4
|
+
*
|
|
5
|
+
* Regression guard for issue #5: every frontmatter parser in eval-framework
|
|
6
|
+
* must tolerate CRLF line endings. A Windows clone (core.autocrlf=true) used
|
|
7
|
+
* to produce 83 spurious consistency failures AND — worse — made the
|
|
8
|
+
* collision check silently compare 27 empty descriptions and pass as a no-op.
|
|
9
|
+
*
|
|
10
|
+
* Method: copy the repo to a temp dir, convert every SKILL.md to CRLF, then:
|
|
11
|
+
* (a) consistency-check must PASS on the converted tree;
|
|
12
|
+
* (b) collision check must PASS unmodified;
|
|
13
|
+
* (c) negative control: plant a duplicated description pair -> collision
|
|
14
|
+
* check must FAIL. If parsing ever regresses to empty strings again,
|
|
15
|
+
* jaccard(NaN) trips nothing and this control goes red.
|
|
16
|
+
*/
|
|
17
|
+
|
|
18
|
+
const { spawnSync } = require('child_process');
|
|
19
|
+
const fs = require('fs');
|
|
20
|
+
const os = require('os');
|
|
21
|
+
const path = require('path');
|
|
22
|
+
|
|
23
|
+
const projectRoot = path.resolve(__dirname, '..');
|
|
24
|
+
let tmpRoot = null;
|
|
25
|
+
|
|
26
|
+
function check(name, condition, detail) {
|
|
27
|
+
if (condition) {
|
|
28
|
+
console.log(`✅ ${name}`);
|
|
29
|
+
return true;
|
|
30
|
+
}
|
|
31
|
+
console.error(`❌ ${name}`);
|
|
32
|
+
if (detail) console.error(` ${detail}`);
|
|
33
|
+
return false;
|
|
34
|
+
}
|
|
35
|
+
|
|
36
|
+
function listSkillMdFiles(dir) {
|
|
37
|
+
const out = [];
|
|
38
|
+
for (const entry of fs.readdirSync(dir, { withFileTypes: true })) {
|
|
39
|
+
if (entry.name.startsWith('.') || !entry.isDirectory()) continue;
|
|
40
|
+
const skillMd = path.join(dir, entry.name, 'SKILL.md');
|
|
41
|
+
if (fs.existsSync(skillMd)) out.push(skillMd);
|
|
42
|
+
const subDir = path.join(dir, entry.name);
|
|
43
|
+
for (const sub of fs.readdirSync(subDir, { withFileTypes: true })) {
|
|
44
|
+
if (!sub.isDirectory()) continue;
|
|
45
|
+
const nested = path.join(subDir, sub.name, 'SKILL.md');
|
|
46
|
+
if (fs.existsSync(nested)) out.push(nested);
|
|
47
|
+
}
|
|
48
|
+
}
|
|
49
|
+
return out;
|
|
50
|
+
}
|
|
51
|
+
|
|
52
|
+
function toCrlf(file) {
|
|
53
|
+
const raw = fs.readFileSync(file, 'utf8');
|
|
54
|
+
fs.writeFileSync(file, raw.replace(/\r?\n/g, '\r\n'), 'utf8');
|
|
55
|
+
}
|
|
56
|
+
|
|
57
|
+
function setDescription(file, newDesc) {
|
|
58
|
+
const lines = fs.readFileSync(file, 'utf8').split('\n');
|
|
59
|
+
const idx = lines.findIndex(l => l.replace(/\r$/, '').startsWith('description:'));
|
|
60
|
+
if (idx === -1) throw new Error(`no description line in ${file}`);
|
|
61
|
+
lines[idx] = `description: ${newDesc}`;
|
|
62
|
+
fs.writeFileSync(file, lines.join('\n'), 'utf8');
|
|
63
|
+
}
|
|
64
|
+
|
|
65
|
+
function getDescription(file) {
|
|
66
|
+
const raw = fs.readFileSync(file, 'utf8');
|
|
67
|
+
const fm = (raw.match(/^---\r?\n([\s\S]*?)\r?\n---/) || [])[1] || '';
|
|
68
|
+
return ((fm.match(/^description:\s*(.+)$/m) || [])[1] || '').trim();
|
|
69
|
+
}
|
|
70
|
+
|
|
71
|
+
function runGate(cwd, script) {
|
|
72
|
+
const r = spawnSync(process.execPath, [path.join('ci', script)], {
|
|
73
|
+
cwd,
|
|
74
|
+
encoding: 'utf8',
|
|
75
|
+
});
|
|
76
|
+
return { code: r.status, stdout: r.stdout || '', stderr: r.stderr || '' };
|
|
77
|
+
}
|
|
78
|
+
|
|
79
|
+
let failures = 0;
|
|
80
|
+
|
|
81
|
+
try {
|
|
82
|
+
// Hermetic workspace: full repo copy minus VCS.
|
|
83
|
+
tmpRoot = fs.mkdtempSync(path.join(os.tmpdir(), 'harness-crlf-test-'));
|
|
84
|
+
fs.cpSync(projectRoot, tmpRoot, {
|
|
85
|
+
recursive: true,
|
|
86
|
+
filter: (src) => {
|
|
87
|
+
const rel = path.relative(projectRoot, src);
|
|
88
|
+
const first = rel.split(path.sep)[0];
|
|
89
|
+
return first !== '.git' && first !== 'node_modules';
|
|
90
|
+
},
|
|
91
|
+
});
|
|
92
|
+
|
|
93
|
+
// Copy node_modules separately to preserve symlinks
|
|
94
|
+
fs.cpSync(path.join(projectRoot, 'node_modules'), path.join(tmpRoot, 'node_modules'), {
|
|
95
|
+
recursive: true,
|
|
96
|
+
});
|
|
97
|
+
|
|
98
|
+
const skillMdFiles = listSkillMdFiles(tmpRoot);
|
|
99
|
+
if (!check(
|
|
100
|
+
'2i-pre. Found enough SKILL.md fixtures in copied tree',
|
|
101
|
+
skillMdFiles.length >= 20,
|
|
102
|
+
`Got ${skillMdFiles.length}`
|
|
103
|
+
)) failures++;
|
|
104
|
+
skillMdFiles.forEach(toCrlf);
|
|
105
|
+
|
|
106
|
+
// (a) Consistency gate must survive a fully-CRLF working tree.
|
|
107
|
+
const consistency = runGate(tmpRoot, 'consistency-check.js');
|
|
108
|
+
if (!check(
|
|
109
|
+
'2i-a. consistency-check passes on all-CRLF tree',
|
|
110
|
+
consistency.code === 0 && consistency.stdout.includes('ALL CONSISTENCY CHECKS PASSED'),
|
|
111
|
+
`exit=${consistency.code}, tail="${(consistency.stdout + consistency.stderr).trim().slice(-200)}"`
|
|
112
|
+
)) failures++;
|
|
113
|
+
|
|
114
|
+
// (b) Collision gate must pass unmodified...
|
|
115
|
+
const collisionClean = runGate(tmpRoot, 'description-collision.js');
|
|
116
|
+
if (!check(
|
|
117
|
+
'2i-b. description-collision passes on all-CRLF tree',
|
|
118
|
+
collisionClean.code === 0,
|
|
119
|
+
`exit=${collisionClean.code}, tail="${(collisionClean.stdout + collisionClean.stderr).trim().slice(-200)}"`
|
|
120
|
+
)) failures++;
|
|
121
|
+
|
|
122
|
+
// (c) ...and must still be comparing REAL descriptions. Plant a duplicate:
|
|
123
|
+
// if parsing regressed to empty strings, similarity would be NaN/0 and this
|
|
124
|
+
// would silently pass — exactly the pre-fix failure mode.
|
|
125
|
+
const donor = path.join(tmpRoot, 'tdd', 'SKILL.md');
|
|
126
|
+
const target = path.join(tmpRoot, 'zoom-out', 'SKILL.md');
|
|
127
|
+
setDescription(target, getDescription(donor));
|
|
128
|
+
const collisionDup = runGate(tmpRoot, 'description-collision.js');
|
|
129
|
+
if (!check(
|
|
130
|
+
'2i-c. Negative control: planted CRLF duplicate FAILS collision check',
|
|
131
|
+
collisionDup.code !== 0,
|
|
132
|
+
`exit=${collisionDup.code} (a silent no-op parses empty descriptions and passes here), tail="${(collisionDup.stdout + collisionDup.stderr).trim().slice(-200)}"`
|
|
133
|
+
)) failures++;
|
|
134
|
+
} catch (err) {
|
|
135
|
+
failures++;
|
|
136
|
+
console.error(`❌ 2i-setup. Test harness error: ${err.message}`);
|
|
137
|
+
} finally {
|
|
138
|
+
if (tmpRoot) {
|
|
139
|
+
try { fs.rmSync(tmpRoot, { recursive: true, force: true }); } catch (e) { /* best-effort */ }
|
|
140
|
+
}
|
|
141
|
+
}
|
|
142
|
+
|
|
143
|
+
if (failures > 0) process.exit(1);
|
|
144
|
+
console.log('\n[2i] CRLF frontmatter immunity: all checks passed.');
|
|
145
|
+
process.exit(0);
|
|
@@ -0,0 +1,34 @@
|
|
|
1
|
+
const fs = require('fs');
|
|
2
|
+
const path = require('path');
|
|
3
|
+
const helper = require('./test-helper');
|
|
4
|
+
const manifest = require('../scripts/lib/manifest');
|
|
5
|
+
const { verifyManifest } = require('../scripts/verify-install');
|
|
6
|
+
|
|
7
|
+
console.log('\n[2i] Installed version and skill-tree verification...');
|
|
8
|
+
const root = helper.tempDir('.mechanism-test-install-verification');
|
|
9
|
+
const installedRoot = path.join(root, 'skills');
|
|
10
|
+
const installedSkill = path.join(installedRoot, 'tdd');
|
|
11
|
+
const manifestPath = path.join(root, 'harness-everything', 'manifest.json');
|
|
12
|
+
fs.mkdirSync(installedRoot, { recursive: true });
|
|
13
|
+
fs.cpSync(path.join(__dirname, '..', 'tdd'), installedSkill, { recursive: true });
|
|
14
|
+
manifest.recordSkillInstall(manifestPath, require('../package.json').version, 'tdd', installedSkill);
|
|
15
|
+
|
|
16
|
+
let result = verifyManifest(manifestPath, installedRoot);
|
|
17
|
+
helper.check('2i. matching version and tree pass', result.failures.length === 0, result.failures.join('; '));
|
|
18
|
+
|
|
19
|
+
fs.appendFileSync(path.join(installedSkill, 'SKILL.md'), '\nlocal drift\n');
|
|
20
|
+
result = verifyManifest(manifestPath, installedRoot);
|
|
21
|
+
helper.check('2i. modified installed skill is reported stale', result.failures.some(f => f.includes('STALE tree: tdd')), JSON.stringify(result.failures));
|
|
22
|
+
|
|
23
|
+
const data = manifest.readManifest(manifestPath);
|
|
24
|
+
data.version = '0.3.3';
|
|
25
|
+
manifest.writeManifest(manifestPath, data);
|
|
26
|
+
result = verifyManifest(manifestPath, installedRoot);
|
|
27
|
+
helper.check('2i. old package version is reported stale', result.failures.some(f => f.includes('STALE version')), JSON.stringify(result.failures));
|
|
28
|
+
|
|
29
|
+
fs.rmSync(manifestPath, { force: true });
|
|
30
|
+
result = verifyManifest(manifestPath, installedRoot);
|
|
31
|
+
helper.check('2i. legacy install without a manifest is reported stale', result.failures.some(f => f.includes('STALE untracked installation')), JSON.stringify(result.failures));
|
|
32
|
+
|
|
33
|
+
fs.rmSync(root, { recursive: true, force: true });
|
|
34
|
+
helper.finish();
|