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 +1,62 @@
|
|
|
1
|
+
const helper = require('./test-helper');
|
|
2
|
+
const fs = require('fs');
|
|
3
|
+
const os = require('os');
|
|
4
|
+
const path = require('path');
|
|
5
|
+
const { buildEngineInvocation, buildWorkspace } = require('../behavioral-evals/run');
|
|
6
|
+
|
|
7
|
+
console.log('\n[2j] Behavioral runner argv integrity...');
|
|
8
|
+
const prompt = "Add punctuation stripping to slug.js, then run npm test. Do not truncate this request.";
|
|
9
|
+
for (const engine of ['opencode', 'claude']) {
|
|
10
|
+
const invocation = buildEngineInvocation(engine, prompt, 'C:/fixture with spaces', 16);
|
|
11
|
+
helper.check(
|
|
12
|
+
`2j. ${engine} receives the complete prompt as one argv value`,
|
|
13
|
+
invocation.args.includes(prompt),
|
|
14
|
+
JSON.stringify(invocation.args)
|
|
15
|
+
);
|
|
16
|
+
}
|
|
17
|
+
|
|
18
|
+
const baseline = buildEngineInvocation('claude', prompt, 'C:/fixture with spaces', 16, 'baseline');
|
|
19
|
+
helper.check('2j. Claude baseline excludes user customizations', baseline.args.includes('--safe-mode') && baseline.args.includes('--setting-sources') && baseline.args.includes('project,local'), JSON.stringify(baseline.args));
|
|
20
|
+
const treatment = buildEngineInvocation('claude', prompt, 'C:/fixture with spaces', 16, 'treatment');
|
|
21
|
+
helper.check('2j. Claude treatment keeps only project/local customizations', treatment.args.includes('--setting-sources') && treatment.args.includes('project,local') && !treatment.args.includes('--safe-mode'), JSON.stringify(treatment.args));
|
|
22
|
+
helper.check('2j. Claude emits parseable verbose stream JSON', treatment.args.includes('--output-format') && treatment.args.includes('stream-json') && treatment.args.includes('--verbose') && !treatment.args.includes('json'), JSON.stringify(treatment.args));
|
|
23
|
+
|
|
24
|
+
let validWorkspace = null;
|
|
25
|
+
try {
|
|
26
|
+
validWorkspace = buildWorkspace({
|
|
27
|
+
id: 'fixture-valid',
|
|
28
|
+
fixture: { files: [{ path: 'nested/fixture.txt', content: 'fixture smoke proof' }] },
|
|
29
|
+
});
|
|
30
|
+
helper.check(
|
|
31
|
+
'2j. behavioral workspace builder accepts a valid nested fixture through the public execution contract',
|
|
32
|
+
fs.readFileSync(path.join(validWorkspace, 'nested', 'fixture.txt'), 'utf8') === 'fixture smoke proof\n',
|
|
33
|
+
validWorkspace
|
|
34
|
+
);
|
|
35
|
+
} catch (error) {
|
|
36
|
+
helper.check(
|
|
37
|
+
'2j. behavioral workspace builder accepts a valid nested fixture through the public execution contract',
|
|
38
|
+
false,
|
|
39
|
+
error && error.stack ? error.stack : String(error)
|
|
40
|
+
);
|
|
41
|
+
} finally {
|
|
42
|
+
if (validWorkspace) fs.rmSync(validWorkspace, { recursive: true, force: true });
|
|
43
|
+
}
|
|
44
|
+
|
|
45
|
+
const traversalTarget = path.join(os.tmpdir(), `harness-behavioral-traversal-${process.pid}-${Date.now()}.txt`);
|
|
46
|
+
let traversalError = null;
|
|
47
|
+
try {
|
|
48
|
+
buildWorkspace({ id: 'fixture-traversal', fixture: { files: [{ path: `../${path.basename(traversalTarget)}`, content: 'path traversal proof' }] } });
|
|
49
|
+
} catch (error) {
|
|
50
|
+
traversalError = error;
|
|
51
|
+
}
|
|
52
|
+
helper.check(
|
|
53
|
+
'2j. behavioral fixtures reject traversal for the expected boundary reason',
|
|
54
|
+
Boolean(
|
|
55
|
+
traversalError &&
|
|
56
|
+
/path must stay inside the generated workspace/.test(traversalError.message) &&
|
|
57
|
+
!fs.existsSync(traversalTarget)
|
|
58
|
+
),
|
|
59
|
+
traversalError && traversalError.stack ? traversalError.stack : traversalTarget
|
|
60
|
+
);
|
|
61
|
+
|
|
62
|
+
helper.finish();
|
|
@@ -0,0 +1,75 @@
|
|
|
1
|
+
const fs = require('fs');
|
|
2
|
+
const path = require('path');
|
|
3
|
+
const helper = require('./test-helper');
|
|
4
|
+
const { checkSkillReferences } = require('./reference-check');
|
|
5
|
+
|
|
6
|
+
console.log('\n[2k] Skill reference checker negative control...');
|
|
7
|
+
const root = helper.tempDir('.mechanism-test-reference-check');
|
|
8
|
+
const skill = path.join(root, 'fixture-skill');
|
|
9
|
+
fs.mkdirSync(skill, { recursive: true });
|
|
10
|
+
fs.writeFileSync(path.join(skill, 'SKILL.md'), 'Run `<this-skill-dir>/scripts/missing.js` before editing.\n');
|
|
11
|
+
let result = checkSkillReferences(root, ['fixture-skill']);
|
|
12
|
+
helper.check('2k. dangling executable reference fails', result.failures.length === 1 && result.failures[0].reference === '<this-skill-dir>/scripts/missing.js', JSON.stringify(result.failures));
|
|
13
|
+
|
|
14
|
+
fs.mkdirSync(path.join(skill, 'scripts'), { recursive: true });
|
|
15
|
+
fs.writeFileSync(path.join(skill, 'scripts', 'missing.js'), '');
|
|
16
|
+
result = checkSkillReferences(root, ['fixture-skill']);
|
|
17
|
+
helper.check('2k. existing executable reference passes', result.failures.length === 0, JSON.stringify(result.failures));
|
|
18
|
+
|
|
19
|
+
// Regression: skill-local paths must say which base they use. A bare path
|
|
20
|
+
// used to pass because the checker guessed the current skill directory.
|
|
21
|
+
fs.mkdirSync(path.join(skill, 'references'), { recursive: true });
|
|
22
|
+
fs.writeFileSync(path.join(skill, 'references', 'guide.md'), '');
|
|
23
|
+
fs.writeFileSync(path.join(skill, 'SKILL.md'), 'Deep dive: `references/guide.md`\n');
|
|
24
|
+
result = checkSkillReferences(root, ['fixture-skill']);
|
|
25
|
+
helper.check(
|
|
26
|
+
'2k. bare skill-local paths fail the explicit marker rule',
|
|
27
|
+
result.failures.length === 1 && result.failures[0].reference === 'references/guide.md' && result.failures[0].target.includes('bare path'),
|
|
28
|
+
JSON.stringify(result.failures)
|
|
29
|
+
);
|
|
30
|
+
fs.writeFileSync(path.join(skill, 'SKILL.md'), 'Deep dive: `<this-skill-dir>/references/guide.md`\n');
|
|
31
|
+
result = checkSkillReferences(root, ['fixture-skill']);
|
|
32
|
+
helper.check('2k. explicit skill-local paths pass', result.failures.length === 0, JSON.stringify(result.failures));
|
|
33
|
+
|
|
34
|
+
// --- placeholder heads are authoritative and unknown ones fail ------------
|
|
35
|
+
// Regression: `<this-skill-dir>/hooks/x.js` used to jump to the repo root
|
|
36
|
+
// because a hardcoded ROOT_PREFIXES list outranked the explicit placeholder.
|
|
37
|
+
fs.mkdirSync(path.join(root, 'hooks'), { recursive: true });
|
|
38
|
+
fs.writeFileSync(path.join(root, 'hooks', 'shared.js'), '');
|
|
39
|
+
fs.writeFileSync(path.join(skill, 'SKILL.md'), 'Run `<this-skill-dir>/hooks/shared.js` first.\n');
|
|
40
|
+
result = checkSkillReferences(root, ['fixture-skill']);
|
|
41
|
+
helper.check(
|
|
42
|
+
'2k. <this-skill-dir>/ is authoritative and never falls back to the repo root',
|
|
43
|
+
result.failures.length === 1 && result.failures[0].target.split(path.sep).join('/') === 'fixture-skill/hooks/shared.js',
|
|
44
|
+
JSON.stringify(result.failures)
|
|
45
|
+
);
|
|
46
|
+
|
|
47
|
+
fs.mkdirSync(path.join(skill, 'hooks'), { recursive: true });
|
|
48
|
+
fs.writeFileSync(path.join(skill, 'hooks', 'shared.js'), '');
|
|
49
|
+
result = checkSkillReferences(root, ['fixture-skill']);
|
|
50
|
+
helper.check('2k. <this-skill-dir>/ resolves under the skill itself', result.failures.length === 0, JSON.stringify(result.failures));
|
|
51
|
+
|
|
52
|
+
// `../` reaches a sibling of the skill directory.
|
|
53
|
+
fs.writeFileSync(path.join(root, 'shared-note.md'), '');
|
|
54
|
+
fs.writeFileSync(path.join(skill, 'SKILL.md'), 'Deep dive: `<this-skill-dir>/../shared-note.md`\n');
|
|
55
|
+
result = checkSkillReferences(root, ['fixture-skill']);
|
|
56
|
+
helper.check('2k. <this-skill-dir>/../ reaches a sibling', result.failures.length === 0, JSON.stringify(result.failures));
|
|
57
|
+
|
|
58
|
+
// Regression: a typo'd placeholder used to be skipped silently, which made
|
|
59
|
+
// every path in the file invisible to this gate instead of failing it.
|
|
60
|
+
fs.writeFileSync(path.join(skill, 'SKILL.md'), 'Deep dive: `<this_folder>/../shared-note.md`\n');
|
|
61
|
+
result = checkSkillReferences(root, ['fixture-skill']);
|
|
62
|
+
helper.check(
|
|
63
|
+
'2k. unknown placeholder fails instead of being skipped',
|
|
64
|
+
result.failures.length === 1 && result.failures[0].target.indexOf('unknown placeholder <this_folder>') === 0,
|
|
65
|
+
JSON.stringify(result.failures)
|
|
66
|
+
);
|
|
67
|
+
|
|
68
|
+
// `<workspace>/` names a path in the USER's project, so it is deliberately
|
|
69
|
+
// recorded as unchecked rather than resolved against this repo.
|
|
70
|
+
fs.writeFileSync(path.join(skill, 'SKILL.md'), 'Writes `<workspace>/tasks/todo.md` as it goes.\n');
|
|
71
|
+
result = checkSkillReferences(root, ['fixture-skill']);
|
|
72
|
+
helper.check('2k. <workspace>/ paths are skipped, not resolved locally', result.failures.length === 0, JSON.stringify(result.failures));
|
|
73
|
+
|
|
74
|
+
fs.rmSync(root, { recursive: true, force: true });
|
|
75
|
+
helper.finish();
|
|
@@ -0,0 +1,301 @@
|
|
|
1
|
+
const fs = require('fs');
|
|
2
|
+
const path = require('path');
|
|
3
|
+
const { spawnSync } = require('child_process');
|
|
4
|
+
const helper = require('./test-helper');
|
|
5
|
+
|
|
6
|
+
console.log('\n[2l] Multi-agent workspace integration...');
|
|
7
|
+
|
|
8
|
+
const root = path.resolve(__dirname, '..');
|
|
9
|
+
const scaffold = path.join(root, 'multi-agent-workspace', 'scripts', 'scaffold.js');
|
|
10
|
+
const indexer = path.join(root, 'multi-agent-workspace', 'scripts', 'index_memory.js');
|
|
11
|
+
const { getRuntimeRoot, resolveProjectDocs } = require('../multi-agent-workspace/scripts/project-docs-resolver');
|
|
12
|
+
const fixture = path.join(__dirname, 'fixtures', 'agency-agents');
|
|
13
|
+
const stateHome = helper.tempDir('.mechanism-test-multi-agent-state');
|
|
14
|
+
fs.mkdirSync(stateHome, { recursive: true });
|
|
15
|
+
process.env.HARNESS_STATE_HOME = stateHome;
|
|
16
|
+
|
|
17
|
+
const workspaceSkill = fs.readFileSync(path.join(root, 'multi-agent-workspace', 'SKILL.md'), 'utf8');
|
|
18
|
+
const grillPlaybook = fs.readFileSync(path.join(root, 'grill-with-docs', 'references', 'session-playbook.md'), 'utf8');
|
|
19
|
+
helper.check(
|
|
20
|
+
'2l. skill contract names global runtime and resolved document writes',
|
|
21
|
+
workspaceSkill.includes('| **State Mutations** |') && workspaceSkill.includes('global workspace-keyed Harness state home') &&
|
|
22
|
+
workspaceSkill.includes('resolver-selected committable repository paths'),
|
|
23
|
+
workspaceSkill
|
|
24
|
+
);
|
|
25
|
+
helper.check(
|
|
26
|
+
'2l. grill fallback names committable document paths',
|
|
27
|
+
grillPlaybook.includes('**Committable fallback**') && grillPlaybook.includes('`docs/adr/`') &&
|
|
28
|
+
!grillPlaybook.includes('.github/harness-everything/adr/'),
|
|
29
|
+
grillPlaybook
|
|
30
|
+
);
|
|
31
|
+
|
|
32
|
+
function run(target, args) {
|
|
33
|
+
return spawnSync(process.execPath, [scaffold, '--workspace', target, ...args], {
|
|
34
|
+
cwd: root,
|
|
35
|
+
env: { ...process.env, HARNESS_STATE_HOME: stateHome },
|
|
36
|
+
encoding: 'utf8'
|
|
37
|
+
});
|
|
38
|
+
}
|
|
39
|
+
|
|
40
|
+
function manifestAt(target) {
|
|
41
|
+
return JSON.parse(fs.readFileSync(path.join(getRuntimeRoot(target), 'manifest.json'), 'utf8'));
|
|
42
|
+
}
|
|
43
|
+
|
|
44
|
+
const fallback = helper.tempDir('.mechanism-test-multi-agent-fallback');
|
|
45
|
+
fs.mkdirSync(fallback, { recursive: true });
|
|
46
|
+
const fallbackRun = run(fallback, []);
|
|
47
|
+
const fallbackManifest = manifestAt(fallback);
|
|
48
|
+
helper.check(
|
|
49
|
+
'2l. missing agency source is explicit and still scaffolds core workspace',
|
|
50
|
+
fallbackRun.status === 0 && fallbackManifest.agency.status === 'unavailable' && fallbackManifest.zones.length === 6 &&
|
|
51
|
+
fallbackManifest.documentationZones.length === 3 &&
|
|
52
|
+
fallbackManifest.paths.decision.source === 'fallback' && fallbackManifest.paths.decision.relativePath === 'docs/adr' &&
|
|
53
|
+
fallbackManifest.paths.domain.source === 'fallback' && fallbackManifest.paths.domain.relativePath === 'docs/domain' &&
|
|
54
|
+
fallbackManifest.paths.architecture.source === 'fallback' && fallbackManifest.paths.architecture.relativePath === 'docs/architecture' &&
|
|
55
|
+
fs.existsSync(path.join(getRuntimeRoot(fallback), 'memory-index.md')) &&
|
|
56
|
+
!fs.existsSync(path.join(fallback, '.harness')),
|
|
57
|
+
fallbackRun.stderr || fallbackRun.stdout
|
|
58
|
+
);
|
|
59
|
+
|
|
60
|
+
const mapped = helper.tempDir('.mechanism-test-multi-agent-context-map');
|
|
61
|
+
fs.mkdirSync(path.join(mapped, 'src', 'ordering'), { recursive: true });
|
|
62
|
+
fs.mkdirSync(path.join(mapped, 'src', 'billing'), { recursive: true });
|
|
63
|
+
fs.writeFileSync(path.join(mapped, 'src', 'ordering', 'CONTEXT.md'), '# Ordering\n');
|
|
64
|
+
fs.writeFileSync(path.join(mapped, 'src', 'billing', 'CONTEXT.md'), '# Billing\n');
|
|
65
|
+
fs.writeFileSync(path.join(mapped, 'CONTEXT-MAP.md'), [
|
|
66
|
+
'# Context Map', '',
|
|
67
|
+
'- [Ordering](./src/ordering/CONTEXT.md)',
|
|
68
|
+
'- [Billing](./src/billing/CONTEXT.md)'
|
|
69
|
+
].join('\n'));
|
|
70
|
+
fs.mkdirSync(path.join(mapped, '.claude', 'harness-everything'), { recursive: true });
|
|
71
|
+
fs.writeFileSync(path.join(mapped, '.claude', 'harness-everything', 'manifest.json'), JSON.stringify({ projectDocs: {
|
|
72
|
+
decisionLocation: 'records/decisions', domainLocation: 'records/domain', architectureLocation: 'records/architecture'
|
|
73
|
+
} }));
|
|
74
|
+
const mappedResolution = resolveProjectDocs(mapped);
|
|
75
|
+
helper.check(
|
|
76
|
+
'2l. root CONTEXT-MAP takes precedence for every document zone',
|
|
77
|
+
mappedResolution.paths.decision.resolution === 'explicit' && mappedResolution.paths.decision.source === 'context-map' &&
|
|
78
|
+
mappedResolution.paths.domain.resolution === 'explicit' && mappedResolution.paths.domain.source === 'context-map' &&
|
|
79
|
+
mappedResolution.paths.architecture.resolution === 'explicit' && mappedResolution.paths.architecture.source === 'context-map' &&
|
|
80
|
+
mappedResolution.paths.decision.path === path.join(mapped, 'src') &&
|
|
81
|
+
mappedResolution.paths.domain.path === path.join(mapped, 'src') &&
|
|
82
|
+
mappedResolution.paths.architecture.path === path.join(mapped, 'src'),
|
|
83
|
+
JSON.stringify(mappedResolution.paths)
|
|
84
|
+
);
|
|
85
|
+
|
|
86
|
+
if (process.platform === 'win32') {
|
|
87
|
+
const caseMapped = helper.tempDir('.mechanism-test-multi-agent-context-map-case');
|
|
88
|
+
fs.mkdirSync(path.join(caseMapped, 'src', 'ordering'), { recursive: true });
|
|
89
|
+
fs.mkdirSync(path.join(caseMapped, 'src', 'billing'), { recursive: true });
|
|
90
|
+
fs.writeFileSync(path.join(caseMapped, 'src', 'ordering', 'CONTEXT.md'), '# Ordering\n');
|
|
91
|
+
fs.writeFileSync(path.join(caseMapped, 'src', 'billing', 'CONTEXT.md'), '# Billing\n');
|
|
92
|
+
fs.writeFileSync(path.join(caseMapped, 'CONTEXT-MAP.md'), [
|
|
93
|
+
'- [Ordering](./src/ordering/CONTEXT.md)',
|
|
94
|
+
'- [Billing](./SRC/billing/CONTEXT.md)'
|
|
95
|
+
].join('\n'));
|
|
96
|
+
const caseResolution = resolveProjectDocs(caseMapped);
|
|
97
|
+
helper.check(
|
|
98
|
+
'2l. CONTEXT-MAP common paths are case-insensitive on Windows',
|
|
99
|
+
caseResolution.paths.domain.path === path.join(caseMapped, 'src'),
|
|
100
|
+
JSON.stringify(caseResolution.paths)
|
|
101
|
+
);
|
|
102
|
+
}
|
|
103
|
+
|
|
104
|
+
const configured = helper.tempDir('.mechanism-test-multi-agent-project-docs');
|
|
105
|
+
fs.mkdirSync(path.join(configured, '.claude', 'harness-everything'), { recursive: true });
|
|
106
|
+
fs.writeFileSync(path.join(configured, '.claude', 'harness-everything', 'manifest.json'), JSON.stringify({ projectDocs: {
|
|
107
|
+
decisionLocation: 'records/decisions', domainLocation: 'records/domain', architectureLocation: 'records/architecture'
|
|
108
|
+
} }));
|
|
109
|
+
const configuredResolution = resolveProjectDocs(configured);
|
|
110
|
+
helper.check(
|
|
111
|
+
'2l. projectDocs paths are explicit and do not become inferred paths',
|
|
112
|
+
configuredResolution.paths.decision.resolution === 'explicit' && configuredResolution.paths.domain.resolution === 'explicit' &&
|
|
113
|
+
configuredResolution.paths.architecture.resolution === 'explicit',
|
|
114
|
+
JSON.stringify(configuredResolution.paths)
|
|
115
|
+
);
|
|
116
|
+
|
|
117
|
+
const inferred = helper.tempDir('.mechanism-test-multi-agent-inference');
|
|
118
|
+
for (const zone of ['adr', 'domain', 'architecture']) fs.mkdirSync(path.join(inferred, 'docs', zone), { recursive: true });
|
|
119
|
+
const inferredResolution = resolveProjectDocs(inferred);
|
|
120
|
+
helper.check(
|
|
121
|
+
'2l. existing repository document folders are inferred',
|
|
122
|
+
inferredResolution.paths.decision.resolution === 'inferred' && inferredResolution.paths.domain.resolution === 'inferred' &&
|
|
123
|
+
inferredResolution.paths.architecture.resolution === 'inferred',
|
|
124
|
+
JSON.stringify(inferredResolution.paths)
|
|
125
|
+
);
|
|
126
|
+
|
|
127
|
+
const migrated = helper.tempDir('.mechanism-test-multi-agent-migration');
|
|
128
|
+
const migratedLegacy = path.join(migrated, '.harness', 'multi-agent', 'decisions');
|
|
129
|
+
fs.mkdirSync(migratedLegacy, { recursive: true });
|
|
130
|
+
fs.writeFileSync(path.join(migratedLegacy, 'authored.md'), 'authored decision\n');
|
|
131
|
+
fs.writeFileSync(path.join(migrated, '.harness', 'multi-agent', 'manifest.json'), '{}');
|
|
132
|
+
const migratedRun = run(migrated, []);
|
|
133
|
+
helper.check(
|
|
134
|
+
'2l. legacy authored records migrate to the resolved repository path',
|
|
135
|
+
migratedRun.status === 0 && fs.readFileSync(path.join(migrated, 'docs', 'adr', 'authored.md'), 'utf8') === 'authored decision\n' &&
|
|
136
|
+
!fs.existsSync(path.join(migrated, '.harness')),
|
|
137
|
+
migratedRun.stderr || migratedRun.stdout
|
|
138
|
+
);
|
|
139
|
+
|
|
140
|
+
const conflictTarget = helper.tempDir('.mechanism-test-multi-agent-migration-conflict');
|
|
141
|
+
const conflictLegacy = path.join(conflictTarget, '.harness', 'multi-agent', 'decisions');
|
|
142
|
+
fs.mkdirSync(conflictLegacy, { recursive: true });
|
|
143
|
+
fs.writeFileSync(path.join(conflictLegacy, 'authored.md'), 'old\n');
|
|
144
|
+
fs.mkdirSync(path.join(conflictTarget, 'docs', 'adr'), { recursive: true });
|
|
145
|
+
fs.writeFileSync(path.join(conflictTarget, 'docs', 'adr', 'authored.md'), 'new\n');
|
|
146
|
+
const conflictRun = run(conflictTarget, []);
|
|
147
|
+
helper.check(
|
|
148
|
+
'2l. migration conflicts preserve both existing files and legacy source',
|
|
149
|
+
conflictRun.status !== 0 && (conflictRun.stderr + conflictRun.stdout).includes('migration conflict') &&
|
|
150
|
+
fs.readFileSync(path.join(conflictTarget, 'docs', 'adr', 'authored.md'), 'utf8') === 'new\n' &&
|
|
151
|
+
fs.existsSync(path.join(conflictTarget, '.harness', 'multi-agent', 'decisions', 'authored.md')),
|
|
152
|
+
conflictRun.stderr || conflictRun.stdout
|
|
153
|
+
);
|
|
154
|
+
|
|
155
|
+
const collidingTarget = helper.tempDir('.mechanism-test-multi-agent-migration-collision');
|
|
156
|
+
const collidingLegacy = path.join(collidingTarget, '.harness', 'multi-agent');
|
|
157
|
+
for (const zone of ['decisions', 'domain', 'architecture']) {
|
|
158
|
+
fs.mkdirSync(path.join(collidingLegacy, zone), { recursive: true });
|
|
159
|
+
fs.writeFileSync(path.join(collidingLegacy, zone, 'same.md'), `${zone}\n`);
|
|
160
|
+
}
|
|
161
|
+
fs.mkdirSync(path.join(collidingTarget, '.claude', 'harness-everything'), { recursive: true });
|
|
162
|
+
fs.writeFileSync(path.join(collidingTarget, '.claude', 'harness-everything', 'manifest.json'), JSON.stringify({ projectDocs: {
|
|
163
|
+
decisionLocation: 'records', domainLocation: 'records', architectureLocation: 'records'
|
|
164
|
+
} }));
|
|
165
|
+
const collidingRun = run(collidingTarget, []);
|
|
166
|
+
helper.check(
|
|
167
|
+
'2l. migration rejects duplicate destinations before copying or deleting records',
|
|
168
|
+
collidingRun.status !== 0 && (collidingRun.stderr + collidingRun.stdout).includes('migration conflict') &&
|
|
169
|
+
!fs.existsSync(path.join(collidingTarget, 'records', 'same.md')) &&
|
|
170
|
+
fs.readFileSync(path.join(collidingLegacy, 'decisions', 'same.md'), 'utf8') === 'decisions\n' &&
|
|
171
|
+
fs.readFileSync(path.join(collidingLegacy, 'domain', 'same.md'), 'utf8') === 'domain\n' &&
|
|
172
|
+
fs.readFileSync(path.join(collidingLegacy, 'architecture', 'same.md'), 'utf8') === 'architecture\n',
|
|
173
|
+
collidingRun.stderr || collidingRun.stdout
|
|
174
|
+
);
|
|
175
|
+
|
|
176
|
+
const junctionTarget = helper.tempDir('.mechanism-test-multi-agent-junction');
|
|
177
|
+
const junctionOutside = helper.tempDir('.mechanism-test-multi-agent-junction-outside');
|
|
178
|
+
fs.mkdirSync(junctionOutside, { recursive: true });
|
|
179
|
+
fs.mkdirSync(path.join(junctionTarget, 'linked-docs'), { recursive: true });
|
|
180
|
+
fs.rmdirSync(path.join(junctionTarget, 'linked-docs'));
|
|
181
|
+
fs.symlinkSync(junctionOutside, path.join(junctionTarget, 'linked-docs'), 'junction');
|
|
182
|
+
fs.mkdirSync(path.join(junctionTarget, '.claude', 'harness-everything'), { recursive: true });
|
|
183
|
+
fs.writeFileSync(path.join(junctionTarget, '.claude', 'harness-everything', 'manifest.json'), JSON.stringify({ projectDocs: {
|
|
184
|
+
architectureLocation: 'linked-docs/architecture'
|
|
185
|
+
} }));
|
|
186
|
+
const junctionRun = run(junctionTarget, []);
|
|
187
|
+
helper.check(
|
|
188
|
+
'2l. document paths reject junction escapes from the workspace',
|
|
189
|
+
junctionRun.status !== 0 && (junctionRun.stderr + junctionRun.stdout).includes('must stay inside workspace') &&
|
|
190
|
+
!fs.existsSync(path.join(junctionOutside, 'architecture')),
|
|
191
|
+
junctionRun.stderr || junctionRun.stdout
|
|
192
|
+
);
|
|
193
|
+
|
|
194
|
+
const linkedLegacyTarget = helper.tempDir('.mechanism-test-multi-agent-linked-legacy');
|
|
195
|
+
const linkedLegacyOutside = helper.tempDir('.mechanism-test-multi-agent-linked-legacy-outside');
|
|
196
|
+
const linkedLegacyRoot = path.join(linkedLegacyTarget, '.harness', 'multi-agent');
|
|
197
|
+
fs.mkdirSync(path.join(linkedLegacyOutside, 'decisions'), { recursive: true });
|
|
198
|
+
fs.writeFileSync(path.join(linkedLegacyOutside, 'decisions', 'outside-authored.md'), 'outside authored\n');
|
|
199
|
+
fs.mkdirSync(path.dirname(linkedLegacyRoot), { recursive: true });
|
|
200
|
+
fs.symlinkSync(linkedLegacyOutside, linkedLegacyRoot, 'junction');
|
|
201
|
+
const linkedLegacyRun = run(linkedLegacyTarget, []);
|
|
202
|
+
helper.check(
|
|
203
|
+
'2l. legacy migration rejects a linked source before copying or deleting outside files',
|
|
204
|
+
linkedLegacyRun.status !== 0 &&
|
|
205
|
+
fs.existsSync(path.join(linkedLegacyOutside, 'decisions', 'outside-authored.md')) &&
|
|
206
|
+
!fs.existsSync(path.join(linkedLegacyTarget, 'docs', 'adr', 'outside-authored.md')),
|
|
207
|
+
linkedLegacyRun.stderr || linkedLegacyRun.stdout
|
|
208
|
+
);
|
|
209
|
+
|
|
210
|
+
const routerPath = path.join(root, 'multi-agent-workspace', 'templates', 'AGENTS.md');
|
|
211
|
+
const runtimeRouter = path.join(getRuntimeRoot(fallback), 'AGENTS.md');
|
|
212
|
+
const router = fs.existsSync(runtimeRouter) ? fs.readFileSync(runtimeRouter, 'utf8') : '';
|
|
213
|
+
helper.check(
|
|
214
|
+
'2l. installed router template resolves paths into global runtime',
|
|
215
|
+
fs.existsSync(routerPath) && fs.existsSync(runtimeRouter) &&
|
|
216
|
+
router.includes(`Write working state to \`${fallbackManifest.runtime.state}\``) &&
|
|
217
|
+
router.includes(`decision records at \`${fallbackManifest.paths.decision.path}\``) &&
|
|
218
|
+
router.includes(`architecture records at \`${fallbackManifest.paths.architecture.path}\``) &&
|
|
219
|
+
!router.includes('.harness/multi-agent'),
|
|
220
|
+
router
|
|
221
|
+
);
|
|
222
|
+
|
|
223
|
+
const missingSourceTarget = helper.tempDir('.mechanism-test-multi-agent-missing-source');
|
|
224
|
+
fs.mkdirSync(missingSourceTarget, { recursive: true });
|
|
225
|
+
const missingSource = run(missingSourceTarget, ['--agency-source', path.join(missingSourceTarget, 'does-not-exist')]);
|
|
226
|
+
helper.check('2l. an explicitly missing agency source fails clearly', missingSource.status !== 0 && (missingSource.stderr + missingSource.stdout).includes('agency source not found'), missingSource.stderr + missingSource.stdout);
|
|
227
|
+
|
|
228
|
+
const sourced = helper.tempDir('.mechanism-test-multi-agent-source');
|
|
229
|
+
fs.mkdirSync(sourced, { recursive: true });
|
|
230
|
+
const sourcedRun = run(sourced, ['--agency-source', fixture, '--division', 'engineering', '--platform', 'codex']);
|
|
231
|
+
const sourcedManifest = manifestAt(sourced);
|
|
232
|
+
const sourcedRoot = getRuntimeRoot(sourced);
|
|
233
|
+
const catalog = JSON.parse(fs.readFileSync(path.join(sourcedRoot, 'roles', 'agency-catalog.json'), 'utf8'));
|
|
234
|
+
const handoff = JSON.parse(fs.readFileSync(path.join(sourcedRoot, 'handoff.json'), 'utf8'));
|
|
235
|
+
helper.check(
|
|
236
|
+
'2l. fixture source selects metadata and records a verified handoff',
|
|
237
|
+
sourcedRun.status === 0 && sourcedManifest.agency.status === 'available' && sourcedManifest.agency.agentCount === 2 &&
|
|
238
|
+
sourcedManifest.selectedAgents.length === 1 && catalog.source.agentCount === 2 && handoff.verification.indexer === 'passed',
|
|
239
|
+
sourcedRun.stderr || sourcedRun.stdout
|
|
240
|
+
);
|
|
241
|
+
|
|
242
|
+
const secondRun = run(sourced, ['--agency-source', fixture, '--division', 'engineering', '--platform', 'codex']);
|
|
243
|
+
helper.check('2l. repeated scaffold is idempotent', secondRun.status === 0, secondRun.stderr || secondRun.stdout);
|
|
244
|
+
|
|
245
|
+
const generatedIndexer = spawnSync(process.execPath, [indexer, '--workspace', sourced, '--manifest', path.join(sourcedRoot, 'manifest.json'), '--output', path.join(sourcedRoot, 'memory-index.md')], { encoding: 'utf8', env: { ...process.env, HARNESS_STATE_HOME: stateHome } });
|
|
246
|
+
const index = fs.readFileSync(path.join(sourcedRoot, 'memory-index.md'), 'utf8');
|
|
247
|
+
helper.check('2l. installed indexer runs independently and lists selected specialist', generatedIndexer.status === 0 && index.includes('Code Reviewer') && !fs.existsSync(path.join(sourcedRoot, 'index_memory.js')) && !fs.existsSync(path.join(sourced, 'AGENTS.md')), generatedIndexer.stderr || generatedIndexer.stdout);
|
|
248
|
+
|
|
249
|
+
const deepOutput = path.join(sourced, 'scratch', 'deep', 'memory.md');
|
|
250
|
+
const deepIndexer = spawnSync(process.execPath, [indexer, '--workspace', sourced, '--manifest', path.join(sourcedRoot, 'manifest.json'), '--output', deepOutput], { encoding: 'utf8', env: { ...process.env, HARNESS_STATE_HOME: stateHome } });
|
|
251
|
+
helper.check('2l. indexer uses explicit manifest/output paths without depth inference', deepIndexer.status === 0 && fs.readFileSync(deepOutput, 'utf8').includes('Code Reviewer'), deepIndexer.stderr || deepIndexer.stdout);
|
|
252
|
+
|
|
253
|
+
const missingManifestIndexer = spawnSync(process.execPath, [indexer, '--workspace', sourced, '--manifest', path.join(sourced, 'does-not-exist.json'), '--output', path.join(sourced, 'missing.md')], { encoding: 'utf8', env: { ...process.env, HARNESS_STATE_HOME: stateHome } });
|
|
254
|
+
helper.check('2l. indexer fails clearly when its manifest is missing', missingManifestIndexer.status !== 0 && (missingManifestIndexer.stderr + missingManifestIndexer.stdout).includes('manifest not found'), missingManifestIndexer.stderr || missingManifestIndexer.stdout);
|
|
255
|
+
|
|
256
|
+
const unsupported = run(helper.tempDir('.mechanism-test-multi-agent-unsupported'), ['--platform', 'unknown']);
|
|
257
|
+
helper.check('2l. unsupported platform fails explicitly', unsupported.status !== 0 && (unsupported.stderr + unsupported.stdout).includes('unsupported platform'), unsupported.stderr + unsupported.stdout);
|
|
258
|
+
|
|
259
|
+
const duplicateSource = helper.tempDir('.mechanism-test-multi-agent-duplicate');
|
|
260
|
+
fs.cpSync(fixture, duplicateSource, { recursive: true });
|
|
261
|
+
const duplicateFile = path.join(duplicateSource, 'testing', 'reality-checker.md');
|
|
262
|
+
fs.writeFileSync(duplicateFile, fs.readFileSync(duplicateFile, 'utf8').replace('name: Reality Checker', 'name: Code Reviewer'));
|
|
263
|
+
const duplicateTarget = helper.tempDir('.mechanism-test-multi-agent-duplicate-target');
|
|
264
|
+
fs.mkdirSync(duplicateTarget, { recursive: true });
|
|
265
|
+
const duplicate = run(duplicateTarget, ['--agency-source', duplicateSource]);
|
|
266
|
+
helper.check('2l. duplicate agent names fail before writing output', duplicate.status !== 0 && (duplicate.stderr + duplicate.stdout).includes('duplicate agent name'), duplicate.stderr + duplicate.stdout);
|
|
267
|
+
|
|
268
|
+
const emptySource = helper.tempDir('.mechanism-test-multi-agent-empty');
|
|
269
|
+
fs.cpSync(fixture, emptySource, { recursive: true });
|
|
270
|
+
const emptyManifestPath = path.join(emptySource, 'divisions.json');
|
|
271
|
+
const emptyManifest = JSON.parse(fs.readFileSync(emptyManifestPath, 'utf8'));
|
|
272
|
+
emptyManifest.divisions.empty = { label: 'Empty', icon: 'Box', color: '#000000' };
|
|
273
|
+
fs.writeFileSync(emptyManifestPath, JSON.stringify(emptyManifest, null, 2));
|
|
274
|
+
fs.mkdirSync(path.join(emptySource, 'empty'));
|
|
275
|
+
const emptyTarget = helper.tempDir('.mechanism-test-multi-agent-empty-target');
|
|
276
|
+
fs.mkdirSync(emptyTarget, { recursive: true });
|
|
277
|
+
const empty = run(emptyTarget, ['--agency-source', emptySource]);
|
|
278
|
+
helper.check('2l. empty division fails before writing output', empty.status !== 0 && (empty.stderr + empty.stdout).includes('division is empty'), empty.stderr + empty.stdout);
|
|
279
|
+
|
|
280
|
+
const unsafeSource = helper.tempDir('.mechanism-test-multi-agent-unsafe-id');
|
|
281
|
+
fs.cpSync(fixture, unsafeSource, { recursive: true });
|
|
282
|
+
const unsafeManifest = JSON.parse(fs.readFileSync(path.join(unsafeSource, 'divisions.json'), 'utf8'));
|
|
283
|
+
unsafeManifest.divisions['../outside'] = { label: 'Outside', icon: 'Alert', color: '#000000' };
|
|
284
|
+
fs.writeFileSync(path.join(unsafeSource, 'divisions.json'), JSON.stringify(unsafeManifest, null, 2));
|
|
285
|
+
const unsafeTarget = helper.tempDir('.mechanism-test-multi-agent-unsafe-target');
|
|
286
|
+
fs.mkdirSync(unsafeTarget, { recursive: true });
|
|
287
|
+
const unsafe = run(unsafeTarget, ['--agency-source', unsafeSource]);
|
|
288
|
+
helper.check('2l. unsafe division path is rejected before reading outside source', unsafe.status !== 0 && (unsafe.stderr + unsafe.stdout).includes('safe path segment'), unsafe.stderr + unsafe.stdout);
|
|
289
|
+
|
|
290
|
+
const driftTarget = helper.tempDir('.mechanism-test-multi-agent-drift');
|
|
291
|
+
fs.mkdirSync(driftTarget, { recursive: true });
|
|
292
|
+
const initial = run(driftTarget, ['--agency-source', fixture]);
|
|
293
|
+
const driftManifestPath = path.join(getRuntimeRoot(driftTarget), 'manifest.json');
|
|
294
|
+
const driftManifest = JSON.parse(fs.readFileSync(driftManifestPath, 'utf8'));
|
|
295
|
+
driftManifest.agency.sourceRevision = 'previous-source-revision';
|
|
296
|
+
fs.writeFileSync(driftManifestPath, JSON.stringify(driftManifest, null, 2));
|
|
297
|
+
const drift = run(driftTarget, ['--agency-source', fixture]);
|
|
298
|
+
const allowedDrift = run(driftTarget, ['--agency-source', fixture, '--allow-source-drift']);
|
|
299
|
+
helper.check('2l. source revision drift requires an explicit override', initial.status === 0 && drift.status !== 0 && allowedDrift.status === 0, drift.stderr + drift.stdout);
|
|
300
|
+
|
|
301
|
+
helper.finish();
|
|
@@ -0,0 +1,65 @@
|
|
|
1
|
+
const fs = require('fs');
|
|
2
|
+
const path = require('path');
|
|
3
|
+
const { execFileSync } = require('child_process');
|
|
4
|
+
const helper = require('./test-helper');
|
|
5
|
+
const { isHarnessRepo } = require('../scripts/lib/workspace');
|
|
6
|
+
|
|
7
|
+
console.log('\n[2m] Self-repo guard identifies the harness repo by identity, not path...');
|
|
8
|
+
|
|
9
|
+
const root = helper.root;
|
|
10
|
+
const bootstrap = path.join(root, 'harness-everything', 'scripts', 'bootstrap.js');
|
|
11
|
+
const selfHeal = path.join(root, 'harness-everything', 'scripts', 'self-heal.js');
|
|
12
|
+
const sandbox = helper.tempDir('.mechanism-test-self-repo-guard');
|
|
13
|
+
|
|
14
|
+
// Two fixture checkouts, both OUTSIDE the scripts' own directory tree. This is
|
|
15
|
+
// the npx/global-install shape: harnessSourceDir !== workspaceRoot, so a
|
|
16
|
+
// path-comparison guard fails open on both (issue #40).
|
|
17
|
+
function fixture(name, pkgName) {
|
|
18
|
+
const dir = path.join(sandbox, name);
|
|
19
|
+
fs.mkdirSync(path.join(dir, '.git'), { recursive: true });
|
|
20
|
+
fs.writeFileSync(path.join(dir, 'package.json'), JSON.stringify({ name: pkgName, version: '0.0.0' }), 'utf8');
|
|
21
|
+
// AGENTS.md exists but carries no Harness advisory marker — exactly the
|
|
22
|
+
// state the real repo is in.
|
|
23
|
+
fs.writeFileSync(path.join(dir, 'AGENTS.md'), '---\ndescription: hand written\n---\n\n# AGENTS.md\n', 'utf8');
|
|
24
|
+
return dir;
|
|
25
|
+
}
|
|
26
|
+
|
|
27
|
+
const selfRepo = fixture('harness-checkout', 'harness-everything');
|
|
28
|
+
const otherRepo = fixture('unrelated-project', 'some-consumer-app');
|
|
29
|
+
|
|
30
|
+
helper.check('2m. isHarnessRepo() recognises a harness checkout', isHarnessRepo(selfRepo) === true);
|
|
31
|
+
helper.check('2m. isHarnessRepo() rejects an unrelated project', isHarnessRepo(otherRepo) === false);
|
|
32
|
+
helper.check('2m. isHarnessRepo() is false when package.json is absent', isHarnessRepo(path.join(sandbox, 'nope')) === false);
|
|
33
|
+
|
|
34
|
+
function run(script, cwd, args) {
|
|
35
|
+
try {
|
|
36
|
+
return execFileSync(process.execPath, [script, ...(args || [])], { cwd, encoding: 'utf8' });
|
|
37
|
+
} catch (err) {
|
|
38
|
+
return (err.stdout || '') + (err.stderr || '');
|
|
39
|
+
}
|
|
40
|
+
}
|
|
41
|
+
|
|
42
|
+
const ADVICE = '[Self-Heal] Missing integration touchpoints';
|
|
43
|
+
|
|
44
|
+
const selfOut = run(bootstrap, selfRepo);
|
|
45
|
+
helper.check(
|
|
46
|
+
'2m. bootstrap run out-of-tree against a harness checkout advises no repair',
|
|
47
|
+
!selfOut.includes(ADVICE),
|
|
48
|
+
selfOut
|
|
49
|
+
);
|
|
50
|
+
|
|
51
|
+
const otherOut = run(bootstrap, otherRepo);
|
|
52
|
+
helper.check(
|
|
53
|
+
'2m. bootstrap still advises repair for an unrelated project (guard is not blanket suppression)',
|
|
54
|
+
otherOut.includes(ADVICE),
|
|
55
|
+
otherOut
|
|
56
|
+
);
|
|
57
|
+
|
|
58
|
+
const healOut = run(selfHeal, selfRepo, ['--check']);
|
|
59
|
+
helper.check(
|
|
60
|
+
'2m. self-heal reports an unmarked AGENTS.md as present-without-block, not missing',
|
|
61
|
+
healOut.includes('present, no Harness advisory block') && !healOut.includes('AGENTS.md) - not created'),
|
|
62
|
+
healOut
|
|
63
|
+
);
|
|
64
|
+
|
|
65
|
+
helper.finish();
|