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 +1,90 @@
|
|
|
1
|
+
const fs = require('fs');
|
|
2
|
+
const path = require('path');
|
|
3
|
+
const helper = require('./test-helper');
|
|
4
|
+
const { checkDisclosure, pointerResolves } = require('./disclosure-check');
|
|
5
|
+
|
|
6
|
+
console.log('\n[2t] Progressive-disclosure negative controls...');
|
|
7
|
+
|
|
8
|
+
const root = helper.tempDir('.mechanism-test-disclosure');
|
|
9
|
+
const skill = path.join(root, 'fixture-skill');
|
|
10
|
+
const workflow = path.join(root, 'docs', 'workflows', 'fixture-skill.md');
|
|
11
|
+
const references = path.join(skill, 'references');
|
|
12
|
+
|
|
13
|
+
function setup() {
|
|
14
|
+
fs.mkdirSync(references, { recursive: true });
|
|
15
|
+
fs.mkdirSync(path.dirname(workflow), { recursive: true });
|
|
16
|
+
fs.writeFileSync(path.join(references, 'guide.md'), '# Guide\n', 'utf8');
|
|
17
|
+
fs.writeFileSync(path.join(skill, 'SKILL.md'), [
|
|
18
|
+
'---',
|
|
19
|
+
'name: fixture-skill',
|
|
20
|
+
'description: A fixture skill for disclosure checks.',
|
|
21
|
+
'---',
|
|
22
|
+
'',
|
|
23
|
+
'# Fixture',
|
|
24
|
+
'',
|
|
25
|
+
'## USE FOR:',
|
|
26
|
+
'- Testing disclosure.',
|
|
27
|
+
'',
|
|
28
|
+
'## DO NOT USE FOR:',
|
|
29
|
+
'- Production work.',
|
|
30
|
+
'',
|
|
31
|
+
'Deep dive: references/guide.md',
|
|
32
|
+
'',
|
|
33
|
+
].join('\n'), 'utf8');
|
|
34
|
+
fs.writeFileSync(workflow, [
|
|
35
|
+
'# Workflow: Fixture',
|
|
36
|
+
'',
|
|
37
|
+
'```mermaid',
|
|
38
|
+
'flowchart TD',
|
|
39
|
+
' Start([Start]) --> Done([Done])',
|
|
40
|
+
'```',
|
|
41
|
+
'',
|
|
42
|
+
].join('\n'), 'utf8');
|
|
43
|
+
}
|
|
44
|
+
|
|
45
|
+
function failures() {
|
|
46
|
+
return checkDisclosure(root).failures;
|
|
47
|
+
}
|
|
48
|
+
|
|
49
|
+
setup();
|
|
50
|
+
helper.check('2t. complete disclosure surface passes', failures().length === 0, JSON.stringify(failures()));
|
|
51
|
+
|
|
52
|
+
fs.rmSync(workflow);
|
|
53
|
+
helper.check('2t. missing workflow pairing fails', failures().some(f => f.check === 'fixture-skill: workflow document'), JSON.stringify(failures()));
|
|
54
|
+
fs.writeFileSync(workflow, '# Workflow: Fixture\n', 'utf8');
|
|
55
|
+
helper.check('2t. missing Mermaid coverage fails', failures().some(f => f.check === 'fixture-skill: Mermaid coverage'), JSON.stringify(failures()));
|
|
56
|
+
fs.writeFileSync(workflow, '```mermaid\nflowchart TD\n```\n', 'utf8');
|
|
57
|
+
helper.check('2t. empty Mermaid transition fails', failures().some(f => f.check === 'fixture-skill: meaningful Mermaid coverage'), JSON.stringify(failures()));
|
|
58
|
+
fs.writeFileSync(workflow, '```mermaid\nflowchart TD\n Start([Start]) --> Done([Done])\n```\n', 'utf8');
|
|
59
|
+
|
|
60
|
+
// Regression: dashes inside a node label are not a transition.
|
|
61
|
+
fs.writeFileSync(workflow, '```mermaid\nflowchart TD\n Start[Start --- End]\n```\n', 'utf8');
|
|
62
|
+
helper.check('2t. dashes inside a node label do not fake a transition', failures().some(f => f.check === 'fixture-skill: meaningful Mermaid coverage'), JSON.stringify(failures()));
|
|
63
|
+
fs.writeFileSync(workflow, '```mermaid\nflowchart TD\n Start([Start]) --> Done([Done])\n```\n', 'utf8');
|
|
64
|
+
|
|
65
|
+
// Regression: a wildcard pointer must match a file, not merely a non-empty
|
|
66
|
+
// directory, and a traversal must not escape the checked-out package.
|
|
67
|
+
fs.rmSync(path.join(skill, 'references', 'guide.md'));
|
|
68
|
+
fs.writeFileSync(path.join(skill, 'references', 'guide.txt'), '# Guide\n', 'utf8');
|
|
69
|
+
helper.check('2t. wildcard Deep dive requires a matching file', !pointerResolves(root, 'fixture-skill', 'references/*.md', ['fixture-skill']), 'references/*.md unexpectedly resolved');
|
|
70
|
+
fs.writeFileSync(path.join(skill, 'references', 'guide.md'), '# Guide\n', 'utf8');
|
|
71
|
+
helper.check('2t. wildcard Deep dive resolves a matching file', pointerResolves(root, 'fixture-skill', 'references/*.md', ['fixture-skill']), 'references/*.md did not resolve');
|
|
72
|
+
const outside = path.resolve(root, '..', 'outside-deep-dive-regression.md');
|
|
73
|
+
fs.writeFileSync(outside, '# Outside\n', 'utf8');
|
|
74
|
+
helper.check('2t. Deep dive traversal stays inside the package', !pointerResolves(root, 'fixture-skill', '../../outside-deep-dive-regression.md', ['fixture-skill']), 'outside traversal resolved');
|
|
75
|
+
fs.rmSync(outside, { force: true });
|
|
76
|
+
|
|
77
|
+
fs.rmSync(path.join(skill, 'references', 'guide.md'));
|
|
78
|
+
helper.check('2t. dangling Deep dive pointer fails', failures().some(f => f.check === 'fixture-skill: Deep dive pointer resolves'), JSON.stringify(failures()));
|
|
79
|
+
fs.writeFileSync(path.join(skill, 'references', 'guide.md'), '# Guide\n', 'utf8');
|
|
80
|
+
fs.appendFileSync(path.join(skill, 'SKILL.md'), '\n```mermaid\nflowchart TD\n A --> B\n```\n', 'utf8');
|
|
81
|
+
helper.check('2t. Mermaid inside SKILL.md fails', failures().some(f => f.check === 'fixture-skill: no Mermaid in SKILL.md'), JSON.stringify(failures()));
|
|
82
|
+
|
|
83
|
+
fs.rmSync(references, { recursive: true, force: true });
|
|
84
|
+
helper.check('2t. missing references directory fails', failures().some(f => f.check === 'fixture-skill: references directory'), JSON.stringify(failures()));
|
|
85
|
+
fs.mkdirSync(references, { recursive: true });
|
|
86
|
+
fs.writeFileSync(path.join(references, 'guide.md'), '# Guide\n', 'utf8');
|
|
87
|
+
fs.writeFileSync(path.join(root, 'docs', 'workflows', 'orphan.md'), '# Workflow: Orphan\n```mermaid\nflowchart TD\n A --> B\n```\n', 'utf8');
|
|
88
|
+
helper.check('2t. orphan workflow document fails', failures().some(f => f.check === 'orphan: workflow pairing'), JSON.stringify(failures()));
|
|
89
|
+
|
|
90
|
+
helper.finish();
|
|
@@ -0,0 +1,200 @@
|
|
|
1
|
+
'use strict';
|
|
2
|
+
|
|
3
|
+
const assert = require('assert');
|
|
4
|
+
const fs = require('fs');
|
|
5
|
+
const os = require('os');
|
|
6
|
+
const path = require('path');
|
|
7
|
+
const { execFileSync } = require('child_process');
|
|
8
|
+
const {
|
|
9
|
+
archiveResult,
|
|
10
|
+
sha256,
|
|
11
|
+
triage,
|
|
12
|
+
} = require('../behavioral-evals/evidence-tool');
|
|
13
|
+
const {
|
|
14
|
+
grade,
|
|
15
|
+
parseSimpleYaml,
|
|
16
|
+
} = require('../behavioral-evals/run');
|
|
17
|
+
const {
|
|
18
|
+
validateCase,
|
|
19
|
+
} = require('../behavioral-evals/case-validator');
|
|
20
|
+
|
|
21
|
+
function write(filePath, content) {
|
|
22
|
+
fs.mkdirSync(path.dirname(filePath), { recursive: true });
|
|
23
|
+
fs.writeFileSync(filePath, content, 'utf8');
|
|
24
|
+
}
|
|
25
|
+
|
|
26
|
+
function git(cwd, args) {
|
|
27
|
+
return execFileSync('git', args, { cwd, stdio: 'ignore' });
|
|
28
|
+
}
|
|
29
|
+
|
|
30
|
+
function makeGitWorkspace(greeting) {
|
|
31
|
+
const workspace = fs.mkdtempSync(path.join(os.tmpdir(), 'evidence-triage-shell-'));
|
|
32
|
+
write(path.join(workspace, 'package.json'), '{"name":"evidence-triage-shell"}\n');
|
|
33
|
+
write(path.join(workspace, 'src', 'greet.js'), `module.exports = name => '${greeting}, ' + name + '!';\n`);
|
|
34
|
+
write(path.join(workspace, 'src', 'format.js'), 'module.exports = value => value;\n');
|
|
35
|
+
write(path.join(workspace, 'README.md'), '# evidence triage\n');
|
|
36
|
+
git(workspace, ['init', '-q']);
|
|
37
|
+
git(workspace, ['config', 'user.email', 'evidence-triage@example.test']);
|
|
38
|
+
git(workspace, ['config', 'user.name', 'Evidence triage test']);
|
|
39
|
+
git(workspace, ['add', '.']);
|
|
40
|
+
git(workspace, ['commit', '-qm', 'fixture']);
|
|
41
|
+
return workspace;
|
|
42
|
+
}
|
|
43
|
+
|
|
44
|
+
function runChecks() {
|
|
45
|
+
const failures = [];
|
|
46
|
+
const check = (name, condition, detail = '') => {
|
|
47
|
+
if (!condition) failures.push(`${name}${detail ? `: ${detail}` : ''}`);
|
|
48
|
+
};
|
|
49
|
+
const tempPaths = [];
|
|
50
|
+
const remove = filePath => {
|
|
51
|
+
try { fs.rmSync(filePath, { recursive: true, force: true }); } catch { /* best effort */ }
|
|
52
|
+
};
|
|
53
|
+
|
|
54
|
+
try {
|
|
55
|
+
for (const caseId of ['scope-discipline', 'pressure-scope-bypass']) {
|
|
56
|
+
const caseData = parseSimpleYaml(fs.readFileSync(path.join(__dirname, '..', 'behavioral-evals', 'cases', `${caseId}.yaml`), 'utf8'));
|
|
57
|
+
const workspace = makeGitWorkspace('Hello');
|
|
58
|
+
tempPaths.push(workspace);
|
|
59
|
+
if (caseId === 'pressure-scope-bypass') write(path.join(workspace, 'src', 'greet.js'), "module.exports = name => 'Hi, ' + name + '!';\n");
|
|
60
|
+
const transcriptDir = fs.mkdtempSync(path.join(os.tmpdir(), 'evidence-triage-shell-trace-'));
|
|
61
|
+
tempPaths.push(transcriptDir);
|
|
62
|
+
const transcript = path.join(transcriptDir, 'empty.jsonl');
|
|
63
|
+
write(transcript, '');
|
|
64
|
+
check(`Windows shell executes ${caseId} assertions`, grade(caseData, workspace, transcript).passed);
|
|
65
|
+
}
|
|
66
|
+
|
|
67
|
+
const emptyCommandCase = {
|
|
68
|
+
id: 'empty-command',
|
|
69
|
+
prompt: 'run the check',
|
|
70
|
+
max_turns: 1,
|
|
71
|
+
fixture: { files: [{ path: 'index.js', content: '' }] },
|
|
72
|
+
expectations: [{ type: 'command_exit_0', command: ' ' }],
|
|
73
|
+
};
|
|
74
|
+
check('empty command_exit_0 is rejected', validateCase(emptyCommandCase).some(error => /non-empty/.test(error)));
|
|
75
|
+
|
|
76
|
+
const crlfCase = parseSimpleYaml([
|
|
77
|
+
'id: crlf-fixture',
|
|
78
|
+
'prompt: |',
|
|
79
|
+
' preserve',
|
|
80
|
+
' both lines',
|
|
81
|
+
'max_turns: 1',
|
|
82
|
+
'fixture:',
|
|
83
|
+
' files:',
|
|
84
|
+
' - path: index.js',
|
|
85
|
+
' content: ok',
|
|
86
|
+
'expectations:',
|
|
87
|
+
' - type: trace_contains',
|
|
88
|
+
' value: ok',
|
|
89
|
+
].join('\r\n'));
|
|
90
|
+
check('CRLF block scalars remain parseable for behavioral cases', crlfCase.prompt === 'preserve\nboth lines', JSON.stringify(crlfCase));
|
|
91
|
+
|
|
92
|
+
const transcriptDir = fs.mkdtempSync(path.join(os.tmpdir(), 'evidence-triage-trace-'));
|
|
93
|
+
tempPaths.push(transcriptDir);
|
|
94
|
+
const runningTranscript = path.join(transcriptDir, 'running.jsonl');
|
|
95
|
+
write(runningTranscript, JSON.stringify({
|
|
96
|
+
type: 'tool_result',
|
|
97
|
+
part: { type: 'tool', tool: 'Bash', callID: 'running-1', state: { status: 'running', input: { command: 'npm test' } } },
|
|
98
|
+
}));
|
|
99
|
+
check('running tool_result does not satisfy tool_completed', !grade({ expectations: [{ type: 'tool_completed', value: { command: 'npm test' } }] }, transcriptDir, runningTranscript).passed);
|
|
100
|
+
|
|
101
|
+
const deniedEditTranscript = path.join(transcriptDir, 'denied-edit.jsonl');
|
|
102
|
+
write(deniedEditTranscript, [
|
|
103
|
+
{ type: 'tool_use', part: { type: 'tool', tool: 'Edit', callID: 'edit-1', state: { status: 'denied', input: { file_path: 'src/index.js' } } } },
|
|
104
|
+
{ type: 'tool_use', part: { type: 'tool', tool: 'Bash', callID: 'test-1', state: { status: 'completed', input: { command: 'npm test' } } } },
|
|
105
|
+
].map(event => JSON.stringify(event)).join('\n'));
|
|
106
|
+
check('after_edit requires an executed edit', !grade({ expectations: [{ type: 'tool_executed', value: { command: 'npm test' }, after_edit: true }] }, transcriptDir, deniedEditTranscript).passed);
|
|
107
|
+
|
|
108
|
+
const casePath = path.join(transcriptDir, 'archive-case.yaml');
|
|
109
|
+
const resultPath = path.join(transcriptDir, 'archive-result.json');
|
|
110
|
+
const archiveOut = path.join(transcriptDir, 'archive-out');
|
|
111
|
+
const caseText = 'id: archive-fixture\nprompt: check\nmax_turns: 1\nfixture:\n files:\n - path: index.js\n content: "ok"\nexpectations:\n - type: trace_contains\n value: ok \n';
|
|
112
|
+
write(casePath, caseText);
|
|
113
|
+
write(resultPath, JSON.stringify({ id: 'archive-fixture', outcome: 'fail', expectations: [] }));
|
|
114
|
+
archiveResult(resultPath, archiveOut, casePath);
|
|
115
|
+
const archivedCasePath = path.join(archiveOut, 'archive-fixture', 'case.yaml');
|
|
116
|
+
const provenance = JSON.parse(fs.readFileSync(path.join(archiveOut, 'archive-fixture', 'provenance.json'), 'utf8'));
|
|
117
|
+
check('archive case hash matches archived bytes', provenance.replay_snapshot.case_sha256 === sha256(fs.readFileSync(archivedCasePath)));
|
|
118
|
+
|
|
119
|
+
const escapedOut = path.join(transcriptDir, 'escaped-out');
|
|
120
|
+
write(path.join(transcriptDir, 'crafted-result.json'), JSON.stringify({ id: '../escaped-case', outcome: 'fail' }));
|
|
121
|
+
check('archive rejects path-traversal result ids', (() => {
|
|
122
|
+
try {
|
|
123
|
+
archiveResult(path.join(transcriptDir, 'crafted-result.json'), escapedOut, casePath);
|
|
124
|
+
return false;
|
|
125
|
+
} catch {
|
|
126
|
+
return true;
|
|
127
|
+
}
|
|
128
|
+
})());
|
|
129
|
+
|
|
130
|
+
const triageOut = path.join(transcriptDir, 'triage-out');
|
|
131
|
+
const rows = triage(triageOut);
|
|
132
|
+
const repeated = rows.find(row => row.id === 'pressure-skip-verification');
|
|
133
|
+
check('triage preserves both repeated historical result files', repeated && repeated.result.result_files && repeated.result.result_files.length === 2);
|
|
134
|
+
const exactMatch = rows.find(row => row.id === 'baseline-performance');
|
|
135
|
+
check('triage matches result ids exactly', exactMatch && exactMatch.result.result_files && exactMatch.result.result_files.length === 1 && exactMatch.result.result_files[0].endsWith('2026-08-27-baseline-performance.json'));
|
|
136
|
+
|
|
137
|
+
const projectRoot = path.join(__dirname, '..');
|
|
138
|
+
const historicalBaseRef = process.env.EVIDENCE_BASE_REF || 'origin/main';
|
|
139
|
+
const currentRunBytes = fs.readFileSync(path.join(projectRoot, 'behavioral-evals', 'run.js'));
|
|
140
|
+
const historicalRunBytes = execFileSync('git', ['show', `${historicalBaseRef}:behavioral-evals/run.js`], {
|
|
141
|
+
cwd: projectRoot,
|
|
142
|
+
stdio: ['ignore', 'pipe', 'ignore'],
|
|
143
|
+
});
|
|
144
|
+
const expectedCurrentCodeSha = sha256(currentRunBytes);
|
|
145
|
+
const expectedHistoricalCodeSha = sha256(historicalRunBytes);
|
|
146
|
+
check('replay code hash matches current run.js bytes', rows.every(row => row.current_code_sha256 === expectedCurrentCodeSha));
|
|
147
|
+
check('historical code hash matches the configured base ref bytes', rows.every(row => row.historical_code_sha256 === expectedHistoricalCodeSha));
|
|
148
|
+
|
|
149
|
+
const abRow = rows.find(row => row.id === 'grill-me-adversarial');
|
|
150
|
+
const abProvenance = JSON.parse(fs.readFileSync(path.join(triageOut, 'grill-me-adversarial', 'provenance.json'), 'utf8'));
|
|
151
|
+
check('CI A/B replay uses the CI A/B runner', abRow && abProvenance.replay.command === 'node ci/ab-test-harness.js run --case grill-me-adversarial');
|
|
152
|
+
|
|
153
|
+
const invalidBaseOut = path.join(transcriptDir, 'invalid-base-out');
|
|
154
|
+
let invalidBaseRejected = false;
|
|
155
|
+
try {
|
|
156
|
+
execFileSync(process.execPath, [path.join(__dirname, '..', 'behavioral-evals', 'evidence-tool.js'), 'triage', '--out', invalidBaseOut], {
|
|
157
|
+
cwd: path.join(__dirname, '..'),
|
|
158
|
+
env: { ...process.env, EVIDENCE_BASE_REF: 'refs/heads/does-not-exist' },
|
|
159
|
+
stdio: 'ignore'
|
|
160
|
+
});
|
|
161
|
+
} catch {
|
|
162
|
+
invalidBaseRejected = true;
|
|
163
|
+
}
|
|
164
|
+
check('triage rejects an unresolvable historical base ref', invalidBaseRejected);
|
|
165
|
+
|
|
166
|
+
const archiveOutside = path.join(transcriptDir, 'archive-outside');
|
|
167
|
+
const archiveLinkOut = path.join(transcriptDir, 'archive-link-out');
|
|
168
|
+
fs.mkdirSync(archiveOutside, { recursive: true });
|
|
169
|
+
fs.mkdirSync(archiveLinkOut, { recursive: true });
|
|
170
|
+
const archiveLink = path.join(archiveLinkOut, 'archive-fixture');
|
|
171
|
+
fs.symlinkSync(archiveOutside, archiveLink, process.platform === 'win32' ? 'junction' : 'dir');
|
|
172
|
+
let archiveLinkRejected = false;
|
|
173
|
+
try {
|
|
174
|
+
archiveResult(resultPath, archiveLinkOut, casePath);
|
|
175
|
+
} catch {
|
|
176
|
+
archiveLinkRejected = true;
|
|
177
|
+
}
|
|
178
|
+
check('archive rejects an output junction or symlink escape', archiveLinkRejected && !fs.existsSync(path.join(archiveOutside, 'case.yaml')));
|
|
179
|
+
|
|
180
|
+
const falsePositiveTranscript = path.join(transcriptDir, 'false-positive-command.jsonl');
|
|
181
|
+
write(falsePositiveTranscript, JSON.stringify({
|
|
182
|
+
type: 'tool_use',
|
|
183
|
+
part: { type: 'tool', tool: 'Bash', callID: 'coverage', state: { status: 'completed', input: { command: 'npm test --coverage' } } }
|
|
184
|
+
}));
|
|
185
|
+
check(
|
|
186
|
+
'tool evidence does not treat an extended command as the exact command',
|
|
187
|
+
!grade({ expectations: [{ type: 'tool_completed', value: { command: 'npm test' } }] }, transcriptDir, falsePositiveTranscript).passed
|
|
188
|
+
);
|
|
189
|
+
} finally {
|
|
190
|
+
tempPaths.forEach(remove);
|
|
191
|
+
}
|
|
192
|
+
return failures;
|
|
193
|
+
}
|
|
194
|
+
|
|
195
|
+
const failures = runChecks();
|
|
196
|
+
if (failures.length) {
|
|
197
|
+
failures.forEach(failure => console.error(`FAIL ${failure}`));
|
|
198
|
+
process.exit(1);
|
|
199
|
+
}
|
|
200
|
+
console.log('Behavioral evidence-triage regressions passed.');
|
|
@@ -0,0 +1,75 @@
|
|
|
1
|
+
const fs = require('fs');
|
|
2
|
+
const path = require('path');
|
|
3
|
+
const { spawnSync } = require('child_process');
|
|
4
|
+
const { parseSimpleYaml } = require('../behavioral-evals/run');
|
|
5
|
+
const helper = require('./test-helper');
|
|
6
|
+
|
|
7
|
+
console.log('\n[2t] Issue #20 regression coverage...');
|
|
8
|
+
|
|
9
|
+
function runNode(script, args = [], env = process.env) {
|
|
10
|
+
return spawnSync(process.execPath, [path.join(helper.root, script), ...args], {
|
|
11
|
+
cwd: helper.root,
|
|
12
|
+
encoding: 'utf8',
|
|
13
|
+
env,
|
|
14
|
+
});
|
|
15
|
+
}
|
|
16
|
+
|
|
17
|
+
const yamlFixture = 'prompt: |\n hello\n world\nnext: value\n';
|
|
18
|
+
const expected = { prompt: 'hello\nworld', next: 'value' };
|
|
19
|
+
helper.check(
|
|
20
|
+
'2t. behavioral YAML parsing preserves block scalars under CRLF',
|
|
21
|
+
JSON.stringify(parseSimpleYaml(yamlFixture)) === JSON.stringify(expected) &&
|
|
22
|
+
JSON.stringify(parseSimpleYaml(yamlFixture.replace(/\n/g, '\r\n'))) === JSON.stringify(expected),
|
|
23
|
+
JSON.stringify({ lf: parseSimpleYaml(yamlFixture), crlf: parseSimpleYaml(yamlFixture.replace(/\n/g, '\r\n')) })
|
|
24
|
+
);
|
|
25
|
+
|
|
26
|
+
const pluginValidation = runNode('behavioral-evals/run-with-plugin.js', ['validate']);
|
|
27
|
+
helper.check(
|
|
28
|
+
'2t. opencode plugin runner validate command exits successfully',
|
|
29
|
+
pluginValidation.status === 0,
|
|
30
|
+
`exit ${pluginValidation.status}\n${pluginValidation.stderr}`
|
|
31
|
+
);
|
|
32
|
+
|
|
33
|
+
const missingWazaEnv = { ...process.env, PATH: '', Path: '' };
|
|
34
|
+
const routingWithoutWaza = runNode('ci/routing-check.js', [], missingWazaEnv);
|
|
35
|
+
helper.check(
|
|
36
|
+
'2t. routing check gracefully skips when waza is unavailable',
|
|
37
|
+
routingWithoutWaza.status === 0 && /skip/i.test(`${routingWithoutWaza.stdout}\n${routingWithoutWaza.stderr}`),
|
|
38
|
+
`exit ${routingWithoutWaza.status}\n${routingWithoutWaza.stdout}\n${routingWithoutWaza.stderr}`
|
|
39
|
+
);
|
|
40
|
+
|
|
41
|
+
const malformedName = `issue20-malformed-collision-${process.pid}`;
|
|
42
|
+
const malformedTarget = path.join(helper.root, malformedName);
|
|
43
|
+
fs.cpSync(
|
|
44
|
+
path.join(helper.root, 'ci', 'fixtures', 'bad-frontmatter-colon'),
|
|
45
|
+
malformedTarget,
|
|
46
|
+
{ recursive: true }
|
|
47
|
+
);
|
|
48
|
+
try {
|
|
49
|
+
const collision = runNode('ci/description-collision.js');
|
|
50
|
+
const output = `${collision.stdout}\n${collision.stderr}`;
|
|
51
|
+
helper.check(
|
|
52
|
+
'2t. description collision gate fails closed on malformed YAML',
|
|
53
|
+
collision.status !== 0 && /YAML|frontmatter/i.test(output),
|
|
54
|
+
`exit ${collision.status}\n${output}`
|
|
55
|
+
);
|
|
56
|
+
} finally {
|
|
57
|
+
fs.rmSync(malformedTarget, { recursive: true, force: true });
|
|
58
|
+
}
|
|
59
|
+
|
|
60
|
+
const negativeControls = runNode('ci/negative-controls.js');
|
|
61
|
+
helper.check(
|
|
62
|
+
'2t. negative controls execute real gates and pass',
|
|
63
|
+
negativeControls.status === 0 && /real gate/i.test(negativeControls.stdout),
|
|
64
|
+
`exit ${negativeControls.status}\n${negativeControls.stdout}\n${negativeControls.stderr}`
|
|
65
|
+
);
|
|
66
|
+
|
|
67
|
+
const rollup = fs.readFileSync(path.join(helper.root, 'docs', 'issue-20-rollup.md'), 'utf8');
|
|
68
|
+
const deferredWork = ['baseline-debugging', 'baseline-performance', 'baseline-security-review', 'baseline-simple-bugfix', 'verify-before-done', 'verification-loop', 'zoom-out', 'security-review'];
|
|
69
|
+
helper.check(
|
|
70
|
+
'2t. issue rollup records unresolved acceptance work accurately',
|
|
71
|
+
/Issue #20 remains open/.test(rollup) && deferredWork.every(term => rollup.includes(term)),
|
|
72
|
+
'rollup must name the open issue and each deferred baseline/skill item'
|
|
73
|
+
);
|
|
74
|
+
|
|
75
|
+
helper.finish();
|
|
@@ -0,0 +1,164 @@
|
|
|
1
|
+
const fs = require('fs');
|
|
2
|
+
const os = require('os');
|
|
3
|
+
const path = require('path');
|
|
4
|
+
const helper = require('./test-helper');
|
|
5
|
+
|
|
6
|
+
console.log('\n[2t] Platform-native skill targets + install ownership safety (issue #80)...');
|
|
7
|
+
|
|
8
|
+
const platforms = require('../hooks/scripts/lib/platforms');
|
|
9
|
+
const skills = require('../scripts/lib/skills');
|
|
10
|
+
const manifest = require('../scripts/lib/manifest');
|
|
11
|
+
const advisory = require('../scripts/lib/advisory-text');
|
|
12
|
+
const root = path.resolve(__dirname, '..');
|
|
13
|
+
const ws = fs.mkdtempSync(path.join(os.tmpdir(), 'harness-2t-ws-'));
|
|
14
|
+
const home = fs.mkdtempSync(path.join(os.tmpdir(), 'harness-2t-home-'));
|
|
15
|
+
|
|
16
|
+
function platform(name) {
|
|
17
|
+
return platforms.find((entry) => entry.name === name);
|
|
18
|
+
}
|
|
19
|
+
|
|
20
|
+
function copyAndAssert(name, isGlobal, expectedPath) {
|
|
21
|
+
const p = platform(name);
|
|
22
|
+
const target = p.getSkillsTarget({ workspaceRoot: ws, userHome: home, isGlobal, manifest });
|
|
23
|
+
helper.check(
|
|
24
|
+
`2t. ${name} ${isGlobal ? 'global' : 'local'} target matches the installer contract`,
|
|
25
|
+
target && path.resolve(target.path) === path.resolve(expectedPath),
|
|
26
|
+
`got ${target && target.path}; expected ${expectedPath}`
|
|
27
|
+
);
|
|
28
|
+
if (!target) return;
|
|
29
|
+
skills.installSkillsToTargets({
|
|
30
|
+
chosenSkills: ['tdd'],
|
|
31
|
+
targetDirs: [target],
|
|
32
|
+
harnessSourceDir: root,
|
|
33
|
+
packageVersion: '9.9.9',
|
|
34
|
+
linkMode: 'copy',
|
|
35
|
+
});
|
|
36
|
+
helper.check(
|
|
37
|
+
`2t. ${name} ${isGlobal ? 'global' : 'local'} --copy writes a readable SKILL.md`,
|
|
38
|
+
fs.existsSync(path.join(expectedPath, 'tdd', 'SKILL.md')),
|
|
39
|
+
`${path.join(expectedPath, 'tdd', 'SKILL.md')} missing`
|
|
40
|
+
);
|
|
41
|
+
}
|
|
42
|
+
|
|
43
|
+
copyAndAssert('codex', false, path.join(ws, '.agents', 'skills'));
|
|
44
|
+
copyAndAssert('continue', true, path.join(home, '.continue', 'skills'));
|
|
45
|
+
copyAndAssert('hermes', false, path.join(ws, '.agents', 'skills'));
|
|
46
|
+
copyAndAssert('hermes', true, path.join(home, '.hermes', 'skills'));
|
|
47
|
+
|
|
48
|
+
// Native global manifests must still be discoverable by the uninstaller and
|
|
49
|
+
// use the exact `global` sentinel that prevents a local-only bulk sweep from
|
|
50
|
+
// deleting user-home skills.
|
|
51
|
+
{
|
|
52
|
+
const installed = skills.getInstalledSkills(ws, home);
|
|
53
|
+
const continueGlobal = installed.find(entry => entry.dirPath === path.join(home, '.continue', 'skills', 'tdd'));
|
|
54
|
+
const hermesGlobal = installed.find(entry => entry.dirPath === path.join(home, '.hermes', 'skills', 'tdd'));
|
|
55
|
+
helper.check('2t. Continue native global manifest is discoverable', !!continueGlobal, JSON.stringify(installed));
|
|
56
|
+
helper.check('2t. Hermes native global manifest is discoverable', !!hermesGlobal, JSON.stringify(installed));
|
|
57
|
+
helper.check('2t. native global skill scopes use the protected global sentinel',
|
|
58
|
+
continueGlobal && hermesGlobal && continueGlobal.scope === 'global' && hermesGlobal.scope === 'global',
|
|
59
|
+
JSON.stringify({ continue: continueGlobal && continueGlobal.scope, hermes: hermesGlobal && hermesGlobal.scope }));
|
|
60
|
+
}
|
|
61
|
+
|
|
62
|
+
// Hermes has no global advisory file, so it records an explicit Harness-owned
|
|
63
|
+
// bookkeeping marker instead of mutating Hermes config just to make uninstall
|
|
64
|
+
// detection possible.
|
|
65
|
+
{
|
|
66
|
+
const hermes = platform('hermes');
|
|
67
|
+
const globalHarness = path.join(home, '.agents', 'harness-everything');
|
|
68
|
+
const marker = path.join(globalHarness, 'hermes-global.json');
|
|
69
|
+
hermes.install({
|
|
70
|
+
isGlobal: true,
|
|
71
|
+
targetWorkspaceRoot: path.join(home, '.agents'),
|
|
72
|
+
userHome: home,
|
|
73
|
+
advisory,
|
|
74
|
+
});
|
|
75
|
+
helper.check('2t. Hermes global install writes an owned detection marker', fs.existsSync(marker), marker);
|
|
76
|
+
hermes.uninstall({
|
|
77
|
+
removeLocal: false,
|
|
78
|
+
removeGlobal: true,
|
|
79
|
+
workspaceRoot: ws,
|
|
80
|
+
userHome: home,
|
|
81
|
+
cleanEmptyDirs(dir) {
|
|
82
|
+
try { if (fs.existsSync(dir) && fs.readdirSync(dir).length === 0) fs.rmdirSync(dir); } catch (e) { /* best effort */ }
|
|
83
|
+
},
|
|
84
|
+
});
|
|
85
|
+
helper.check('2t. Hermes global uninstall removes its detection marker', !fs.existsSync(marker), marker);
|
|
86
|
+
}
|
|
87
|
+
|
|
88
|
+
// A malformed user settings file must never be converted to `{}` and
|
|
89
|
+
// overwritten. Installation fails before any mutation and preserves bytes.
|
|
90
|
+
{
|
|
91
|
+
const claude = platform('claude');
|
|
92
|
+
const claudeDir = path.join(ws, '.claude');
|
|
93
|
+
const settings = path.join(claudeDir, 'settings.json');
|
|
94
|
+
fs.mkdirSync(claudeDir, { recursive: true });
|
|
95
|
+
const original = '{ malformed user settings\n';
|
|
96
|
+
fs.writeFileSync(settings, original, 'utf8');
|
|
97
|
+
|
|
98
|
+
let failedClosed = false;
|
|
99
|
+
try {
|
|
100
|
+
claude.install({
|
|
101
|
+
isGlobal: false,
|
|
102
|
+
targetWorkspaceRoot: ws,
|
|
103
|
+
harnessSourceDir: root,
|
|
104
|
+
packageVersion: '9.9.9',
|
|
105
|
+
claudeHooks: { mergeHarnessHooks() {} },
|
|
106
|
+
manifest,
|
|
107
|
+
});
|
|
108
|
+
} catch (error) {
|
|
109
|
+
failedClosed = /Refusing to modify malformed/.test(error.message);
|
|
110
|
+
}
|
|
111
|
+
|
|
112
|
+
helper.check('2t. malformed Claude settings fail closed', failedClosed, 'install did not reject malformed JSON');
|
|
113
|
+
helper.check(
|
|
114
|
+
'2t. malformed Claude settings remain byte-for-byte unchanged',
|
|
115
|
+
fs.readFileSync(settings, 'utf8') === original,
|
|
116
|
+
`got ${JSON.stringify(fs.readFileSync(settings, 'utf8'))}`
|
|
117
|
+
);
|
|
118
|
+
}
|
|
119
|
+
|
|
120
|
+
// Dedicated/fixed filenames must not let Harness claim or delete an unrelated
|
|
121
|
+
// pre-existing user file merely because the path matches.
|
|
122
|
+
{
|
|
123
|
+
const continueRule = path.join(ws, '.continue', 'rules', 'harness.md');
|
|
124
|
+
fs.mkdirSync(path.dirname(continueRule), { recursive: true });
|
|
125
|
+
const original = '# user-owned rule\n';
|
|
126
|
+
fs.writeFileSync(continueRule, original, 'utf8');
|
|
127
|
+
let rejected = false;
|
|
128
|
+
try {
|
|
129
|
+
advisory.installContinueRule(continueRule, '.continue/rules/harness.md');
|
|
130
|
+
} catch (error) {
|
|
131
|
+
rejected = /Refusing to overwrite pre-existing non-Harness/.test(error.message);
|
|
132
|
+
}
|
|
133
|
+
helper.check('2t. Continue refuses to overwrite a user-owned harness.md', rejected, 'collision was not rejected');
|
|
134
|
+
helper.check('2t. Continue collision leaves original bytes unchanged', fs.readFileSync(continueRule, 'utf8') === original, 'user rule changed');
|
|
135
|
+
}
|
|
136
|
+
|
|
137
|
+
{
|
|
138
|
+
const promptsDir = path.join(home, 'prompts');
|
|
139
|
+
fs.mkdirSync(promptsDir, { recursive: true });
|
|
140
|
+
const getUserPromptsDir = () => promptsDir;
|
|
141
|
+
const cases = [
|
|
142
|
+
['copilot', 'harness.instructions.md'],
|
|
143
|
+
['codex', 'harness.agent.md'],
|
|
144
|
+
];
|
|
145
|
+
for (const [name, fileName] of cases) {
|
|
146
|
+
const p = platform(name);
|
|
147
|
+
const target = path.join(promptsDir, fileName);
|
|
148
|
+
const original = `user-owned ${name}\n`;
|
|
149
|
+
fs.writeFileSync(target, original, 'utf8');
|
|
150
|
+
let rejected = false;
|
|
151
|
+
try {
|
|
152
|
+
p.install({ isGlobal: true, targetWorkspaceRoot: ws, getUserPromptsDir, advisory });
|
|
153
|
+
} catch (error) {
|
|
154
|
+
rejected = /Refusing to overwrite pre-existing non-Harness/.test(error.message);
|
|
155
|
+
}
|
|
156
|
+
helper.check(`2t. ${name} global install refuses a user-owned fixed prompt file`, rejected, 'collision was not rejected');
|
|
157
|
+
p.uninstall({ removeLocal: false, removeGlobal: true, workspaceRoot: ws, userHome: home, getUserPromptsDir, cleanEmptyDirs() {} });
|
|
158
|
+
helper.check(`2t. ${name} global uninstall preserves non-Harness file`, fs.readFileSync(target, 'utf8') === original, 'user-owned file was changed or deleted');
|
|
159
|
+
}
|
|
160
|
+
}
|
|
161
|
+
|
|
162
|
+
fs.rmSync(ws, { recursive: true, force: true });
|
|
163
|
+
fs.rmSync(home, { recursive: true, force: true });
|
|
164
|
+
helper.finish();
|