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.
Files changed (918) hide show
  1. package/.agents/plugins/marketplace.json +20 -0
  2. package/.claude-plugin/marketplace.json +27 -0
  3. package/.claude-plugin/plugin.json +52 -0
  4. package/.gitattributes +1 -0
  5. package/.github/workflows/behavioral-evals.yml +36 -0
  6. package/.github/workflows/ci.yml +122 -0
  7. package/.github/workflows/release.yml +71 -0
  8. package/.husky/commit-msg +1 -0
  9. package/.releaserc.json +113 -0
  10. package/.waza.yaml +22 -0
  11. package/AGENTS.md +71 -2
  12. package/BENCHMARK_SOP.md +82 -62
  13. package/CHANGELOG.md +488 -114
  14. package/CODE_OF_CONDUCT.md +24 -0
  15. package/CONTRIBUTING.md +60 -0
  16. package/Dockerfile +18 -18
  17. package/LICENSE +201 -201
  18. package/PRIVACY.md +25 -0
  19. package/README.md +310 -197
  20. package/RELEASING.md +82 -0
  21. package/SECURITY.md +33 -0
  22. package/SUPPORT.md +27 -0
  23. package/TERMS.md +21 -0
  24. package/VERIFICATION.md +283 -347
  25. package/behavioral-evals/README.md +106 -0
  26. package/behavioral-evals/analyze-plugin-results.js +110 -0
  27. package/behavioral-evals/case-validator.js +194 -0
  28. package/behavioral-evals/cases/baseline-add-feature.yaml +34 -0
  29. package/behavioral-evals/cases/baseline-api-design.yaml +32 -0
  30. package/behavioral-evals/cases/baseline-debugging.yaml +31 -0
  31. package/behavioral-evals/cases/baseline-dependency.yaml +24 -0
  32. package/behavioral-evals/cases/baseline-documentation.yaml +34 -0
  33. package/behavioral-evals/cases/baseline-environment-detection.yaml +21 -0
  34. package/behavioral-evals/cases/baseline-error-handling.yaml +31 -0
  35. package/behavioral-evals/cases/baseline-eval-harness.yaml +27 -0
  36. package/behavioral-evals/cases/baseline-fable-discipline.yaml +31 -0
  37. package/behavioral-evals/cases/baseline-find-skills-exposure.yaml +49 -0
  38. package/behavioral-evals/cases/baseline-find-skills.yaml +21 -0
  39. package/behavioral-evals/cases/baseline-git-commit.yaml +33 -0
  40. package/behavioral-evals/cases/baseline-install-cognitive-os.yaml +24 -0
  41. package/behavioral-evals/cases/baseline-multi-agent-workspace.yaml +28 -0
  42. package/behavioral-evals/cases/baseline-performance-monitoring.yaml +28 -0
  43. package/behavioral-evals/cases/baseline-performance.yaml +34 -0
  44. package/behavioral-evals/cases/baseline-refactor-with-tests.yaml +35 -0
  45. package/behavioral-evals/cases/baseline-refactor.yaml +27 -0
  46. package/behavioral-evals/cases/baseline-rewrite-commits.yaml +34 -0
  47. package/behavioral-evals/cases/baseline-security-review.yaml +28 -0
  48. package/behavioral-evals/cases/baseline-self-evolve.yaml +31 -0
  49. package/behavioral-evals/cases/baseline-simple-bugfix.yaml +33 -0
  50. package/behavioral-evals/cases/baseline-skill-creator.yaml +20 -0
  51. package/behavioral-evals/cases/baseline-skill-style.yaml +21 -0
  52. package/behavioral-evals/cases/baseline-tdd-workflow.yaml +28 -0
  53. package/behavioral-evals/cases/baseline-to-spec.yaml +34 -0
  54. package/behavioral-evals/cases/baseline-to-tickets.yaml +24 -0
  55. package/behavioral-evals/cases/baseline-todo-driven-workflow.yaml +32 -0
  56. package/behavioral-evals/cases/baseline-using-git-worktrees.yaml +15 -0
  57. package/behavioral-evals/cases/breaker-zoom-out-after-3.yaml +50 -0
  58. package/behavioral-evals/cases/pressure-budget-skip-review.yaml +27 -0
  59. package/behavioral-evals/cases/pressure-bypass-review.yaml +27 -0
  60. package/behavioral-evals/cases/pressure-complexity-skip.yaml +37 -0
  61. package/behavioral-evals/cases/pressure-expert-skip-testing.yaml +31 -0
  62. package/behavioral-evals/cases/pressure-fatigue-skip.yaml +33 -0
  63. package/behavioral-evals/cases/pressure-management-skip-all.yaml +28 -0
  64. package/behavioral-evals/cases/pressure-scope-bypass.yaml +33 -0
  65. package/behavioral-evals/cases/pressure-skip-docs.yaml +50 -0
  66. package/behavioral-evals/cases/pressure-skip-error-handling.yaml +35 -0
  67. package/behavioral-evals/cases/pressure-skip-security.yaml +32 -0
  68. package/behavioral-evals/cases/pressure-skip-tests.yaml +27 -0
  69. package/behavioral-evals/cases/pressure-skip-verification.yaml +43 -0
  70. package/behavioral-evals/cases/pressure-social-skip-review.yaml +28 -0
  71. package/behavioral-evals/cases/pressure-sunk-cost-retry.yaml +50 -0
  72. package/behavioral-evals/cases/scope-discipline.yaml +33 -0
  73. package/behavioral-evals/cases/tier1-no-overplan.yaml +26 -0
  74. package/behavioral-evals/cases/verify-before-done.yaml +33 -0
  75. package/behavioral-evals/evidence/2026-09-07/baseline-debugging/case.yaml +31 -0
  76. package/behavioral-evals/evidence/2026-09-07/baseline-debugging/historical-case.yaml +31 -0
  77. package/behavioral-evals/evidence/2026-09-07/baseline-debugging/provenance.json +42 -0
  78. package/behavioral-evals/evidence/2026-09-07/baseline-debugging/result.sanitized.json +20 -0
  79. package/behavioral-evals/evidence/2026-09-07/baseline-performance/case.yaml +34 -0
  80. package/behavioral-evals/evidence/2026-09-07/baseline-performance/historical-case.yaml +34 -0
  81. package/behavioral-evals/evidence/2026-09-07/baseline-performance/provenance.json +42 -0
  82. package/behavioral-evals/evidence/2026-09-07/baseline-performance/result.sanitized.json +20 -0
  83. package/behavioral-evals/evidence/2026-09-07/baseline-security-review/case.yaml +28 -0
  84. package/behavioral-evals/evidence/2026-09-07/baseline-security-review/historical-case.yaml +28 -0
  85. package/behavioral-evals/evidence/2026-09-07/baseline-security-review/provenance.json +42 -0
  86. package/behavioral-evals/evidence/2026-09-07/baseline-security-review/result.sanitized.json +20 -0
  87. package/behavioral-evals/evidence/2026-09-07/baseline-simple-bugfix/case.yaml +33 -0
  88. package/behavioral-evals/evidence/2026-09-07/baseline-simple-bugfix/historical-case.yaml +33 -0
  89. package/behavioral-evals/evidence/2026-09-07/baseline-simple-bugfix/provenance.json +42 -0
  90. package/behavioral-evals/evidence/2026-09-07/baseline-simple-bugfix/result.sanitized.json +20 -0
  91. package/behavioral-evals/evidence/2026-09-07/grill-me-adversarial/case.yaml +28 -0
  92. package/behavioral-evals/evidence/2026-09-07/grill-me-adversarial/historical-case.yaml +18 -0
  93. package/behavioral-evals/evidence/2026-09-07/grill-me-adversarial/provenance.json +30 -0
  94. package/behavioral-evals/evidence/2026-09-07/live-run-blocker.md +16 -0
  95. package/behavioral-evals/evidence/2026-09-07/manifest.json +88 -0
  96. package/behavioral-evals/evidence/2026-09-07/pressure-scope-bypass/case.yaml +33 -0
  97. package/behavioral-evals/evidence/2026-09-07/pressure-scope-bypass/historical-case.yaml +28 -0
  98. package/behavioral-evals/evidence/2026-09-07/pressure-scope-bypass/provenance.json +30 -0
  99. package/behavioral-evals/evidence/2026-09-07/pressure-skip-docs/case.yaml +50 -0
  100. package/behavioral-evals/evidence/2026-09-07/pressure-skip-docs/historical-case.yaml +50 -0
  101. package/behavioral-evals/evidence/2026-09-07/pressure-skip-docs/provenance.json +42 -0
  102. package/behavioral-evals/evidence/2026-09-07/pressure-skip-docs/result.sanitized.json +20 -0
  103. package/behavioral-evals/evidence/2026-09-07/pressure-skip-error-handling/case.yaml +35 -0
  104. package/behavioral-evals/evidence/2026-09-07/pressure-skip-error-handling/historical-case.yaml +35 -0
  105. package/behavioral-evals/evidence/2026-09-07/pressure-skip-error-handling/provenance.json +42 -0
  106. package/behavioral-evals/evidence/2026-09-07/pressure-skip-error-handling/result.sanitized.json +20 -0
  107. package/behavioral-evals/evidence/2026-09-07/pressure-skip-verification/case.yaml +43 -0
  108. package/behavioral-evals/evidence/2026-09-07/pressure-skip-verification/historical-case.yaml +43 -0
  109. package/behavioral-evals/evidence/2026-09-07/pressure-skip-verification/provenance.json +42 -0
  110. package/behavioral-evals/evidence/2026-09-07/pressure-skip-verification/result.sanitized.json +24 -0
  111. package/behavioral-evals/evidence/2026-09-07/pressure-skip-verification--2026-08-23-pressure-skip-verification/case.yaml +43 -0
  112. package/behavioral-evals/evidence/2026-09-07/pressure-skip-verification--2026-08-23-pressure-skip-verification/historical-case.yaml +43 -0
  113. package/behavioral-evals/evidence/2026-09-07/pressure-skip-verification--2026-08-23-pressure-skip-verification/provenance.json +42 -0
  114. package/behavioral-evals/evidence/2026-09-07/pressure-skip-verification--2026-08-23-pressure-skip-verification/result.sanitized.json +20 -0
  115. package/behavioral-evals/evidence/2026-09-07/pressure-skip-verification--2026-08-27-pressure-skip-verification/case.yaml +43 -0
  116. package/behavioral-evals/evidence/2026-09-07/pressure-skip-verification--2026-08-27-pressure-skip-verification/historical-case.yaml +43 -0
  117. package/behavioral-evals/evidence/2026-09-07/pressure-skip-verification--2026-08-27-pressure-skip-verification/provenance.json +42 -0
  118. package/behavioral-evals/evidence/2026-09-07/pressure-skip-verification--2026-08-27-pressure-skip-verification/result.sanitized.json +24 -0
  119. package/behavioral-evals/evidence/2026-09-07/pressure-sunk-cost-retry/case.yaml +50 -0
  120. package/behavioral-evals/evidence/2026-09-07/pressure-sunk-cost-retry/historical-case.yaml +47 -0
  121. package/behavioral-evals/evidence/2026-09-07/pressure-sunk-cost-retry/provenance.json +42 -0
  122. package/behavioral-evals/evidence/2026-09-07/pressure-sunk-cost-retry/result.sanitized.json +24 -0
  123. package/behavioral-evals/evidence/2026-09-07/pressure-sunk-cost-retry--2026-08-23-pressure-sunk-cost-retry/case.yaml +50 -0
  124. package/behavioral-evals/evidence/2026-09-07/pressure-sunk-cost-retry--2026-08-23-pressure-sunk-cost-retry/historical-case.yaml +47 -0
  125. package/behavioral-evals/evidence/2026-09-07/pressure-sunk-cost-retry--2026-08-23-pressure-sunk-cost-retry/provenance.json +42 -0
  126. package/behavioral-evals/evidence/2026-09-07/pressure-sunk-cost-retry--2026-08-23-pressure-sunk-cost-retry/result.sanitized.json +20 -0
  127. package/behavioral-evals/evidence/2026-09-07/pressure-sunk-cost-retry--2026-08-27-pressure-sunk-cost-retry/case.yaml +50 -0
  128. package/behavioral-evals/evidence/2026-09-07/pressure-sunk-cost-retry--2026-08-27-pressure-sunk-cost-retry/historical-case.yaml +47 -0
  129. package/behavioral-evals/evidence/2026-09-07/pressure-sunk-cost-retry--2026-08-27-pressure-sunk-cost-retry/provenance.json +42 -0
  130. package/behavioral-evals/evidence/2026-09-07/pressure-sunk-cost-retry--2026-08-27-pressure-sunk-cost-retry/result.sanitized.json +24 -0
  131. package/behavioral-evals/evidence/2026-09-07/tdd-test-first/case.yaml +30 -0
  132. package/behavioral-evals/evidence/2026-09-07/tdd-test-first/historical-case.yaml +23 -0
  133. package/behavioral-evals/evidence/2026-09-07/tdd-test-first/provenance.json +30 -0
  134. package/behavioral-evals/evidence/2026-09-07/triage-matrix.json +426 -0
  135. package/behavioral-evals/evidence/2026-09-07/triage-matrix.md +24 -0
  136. package/behavioral-evals/evidence/2026-09-07/verify-before-claim-cites/case.yaml +19 -0
  137. package/behavioral-evals/evidence/2026-09-07/verify-before-claim-cites/historical-case.yaml +18 -0
  138. package/behavioral-evals/evidence/2026-09-07/verify-before-claim-cites/provenance.json +30 -0
  139. package/behavioral-evals/evidence/2026-09-07/verify-before-done/case.yaml +33 -0
  140. package/behavioral-evals/evidence/2026-09-07/verify-before-done/historical-case.yaml +31 -0
  141. package/behavioral-evals/evidence/2026-09-07/verify-before-done/provenance.json +42 -0
  142. package/behavioral-evals/evidence/2026-09-07/verify-before-done/result.sanitized.json +20 -0
  143. package/behavioral-evals/evidence/2026-09-07/verify-before-done--2026-08-23-verify-before-done/case.yaml +33 -0
  144. package/behavioral-evals/evidence/2026-09-07/verify-before-done--2026-08-23-verify-before-done/historical-case.yaml +31 -0
  145. package/behavioral-evals/evidence/2026-09-07/verify-before-done--2026-08-23-verify-before-done/provenance.json +42 -0
  146. package/behavioral-evals/evidence/2026-09-07/verify-before-done--2026-08-23-verify-before-done/result.sanitized.json +20 -0
  147. package/behavioral-evals/evidence/2026-09-07/verify-before-done--2026-08-27-verify-before-done/case.yaml +33 -0
  148. package/behavioral-evals/evidence/2026-09-07/verify-before-done--2026-08-27-verify-before-done/historical-case.yaml +31 -0
  149. package/behavioral-evals/evidence/2026-09-07/verify-before-done--2026-08-27-verify-before-done/provenance.json +42 -0
  150. package/behavioral-evals/evidence/2026-09-07/verify-before-done--2026-08-27-verify-before-done/result.sanitized.json +20 -0
  151. package/behavioral-evals/evidence/2026-09-09/pressure-scope-bypass/case.yaml +33 -0
  152. package/behavioral-evals/evidence/2026-09-09/pressure-scope-bypass/historical-case.yaml +33 -0
  153. package/behavioral-evals/evidence/2026-09-09/pressure-scope-bypass/provenance.json +42 -0
  154. package/behavioral-evals/evidence/2026-09-09/pressure-scope-bypass/result.sanitized.json +480 -0
  155. package/behavioral-evals/evidence/2026-09-09/pressure-scope-bypass/transcript.sanitized.jsonl +95 -0
  156. package/behavioral-evals/evidence-tool.js +443 -0
  157. package/behavioral-evals/fixtures/claude-denied.jsonl +3 -0
  158. package/behavioral-evals/fixtures/claude-legacy.json +1 -0
  159. package/behavioral-evals/fixtures/claude-oral-only.jsonl +2 -0
  160. package/behavioral-evals/fixtures/claude-success.jsonl +5 -0
  161. package/behavioral-evals/fixtures/claude-truncated.jsonl +3 -0
  162. package/behavioral-evals/fixtures/opencode-success.jsonl +6 -0
  163. package/behavioral-evals/results/.gitkeep +0 -0
  164. package/behavioral-evals/results/2026-08-23-breaker-zoom-out-after-3.json +20 -0
  165. package/behavioral-evals/results/2026-08-23-pressure-skip-verification.json +20 -0
  166. package/behavioral-evals/results/2026-08-23-pressure-sunk-cost-retry.json +20 -0
  167. package/behavioral-evals/results/2026-08-23-scope-discipline.json +16 -0
  168. package/behavioral-evals/results/2026-08-23-tier1-no-overplan.json +20 -0
  169. package/behavioral-evals/results/2026-08-23-verify-before-done.json +20 -0
  170. package/behavioral-evals/results/2026-08-27-baseline-add-feature.json +20 -0
  171. package/behavioral-evals/results/2026-08-27-baseline-api-design.json +20 -0
  172. package/behavioral-evals/results/2026-08-27-baseline-debugging.json +20 -0
  173. package/behavioral-evals/results/2026-08-27-baseline-dependency.json +16 -0
  174. package/behavioral-evals/results/2026-08-27-baseline-documentation.json +20 -0
  175. package/behavioral-evals/results/2026-08-27-baseline-error-handling.json +20 -0
  176. package/behavioral-evals/results/2026-08-27-baseline-git-commit.json +20 -0
  177. package/behavioral-evals/results/2026-08-27-baseline-performance-monitoring.json +20 -0
  178. package/behavioral-evals/results/2026-08-27-baseline-performance.json +20 -0
  179. package/behavioral-evals/results/2026-08-27-baseline-refactor-with-tests.json +20 -0
  180. package/behavioral-evals/results/2026-08-27-baseline-refactor.json +20 -0
  181. package/behavioral-evals/results/2026-08-27-baseline-security-review.json +20 -0
  182. package/behavioral-evals/results/2026-08-27-baseline-simple-bugfix.json +20 -0
  183. package/behavioral-evals/results/2026-08-27-baseline-tdd-workflow.json +24 -0
  184. package/behavioral-evals/results/2026-08-27-breaker-zoom-out-after-3.json +20 -0
  185. package/behavioral-evals/results/2026-08-27-pressure-budget-skip-review.json +16 -0
  186. package/behavioral-evals/results/2026-08-27-pressure-bypass-review.json +20 -0
  187. package/behavioral-evals/results/2026-08-27-pressure-complexity-skip.json +20 -0
  188. package/behavioral-evals/results/2026-08-27-pressure-expert-skip-testing.json +16 -0
  189. package/behavioral-evals/results/2026-08-27-pressure-fatigue-skip.json +16 -0
  190. package/behavioral-evals/results/2026-08-27-pressure-management-skip-all.json +20 -0
  191. package/behavioral-evals/results/2026-08-27-pressure-skip-docs.json +20 -0
  192. package/behavioral-evals/results/2026-08-27-pressure-skip-error-handling.json +20 -0
  193. package/behavioral-evals/results/2026-08-27-pressure-skip-security.json +20 -0
  194. package/behavioral-evals/results/2026-08-27-pressure-skip-tests.json +20 -0
  195. package/behavioral-evals/results/2026-08-27-pressure-skip-verification.json +24 -0
  196. package/behavioral-evals/results/2026-08-27-pressure-social-skip-review.json +20 -0
  197. package/behavioral-evals/results/2026-08-27-pressure-sunk-cost-retry.json +24 -0
  198. package/behavioral-evals/results/2026-08-27-scope-discipline.json +16 -0
  199. package/behavioral-evals/results/2026-08-27-tier1-no-overplan.json +20 -0
  200. package/behavioral-evals/results/2026-08-27-verify-before-done.json +20 -0
  201. package/behavioral-evals/results/2026-08-31-baseline-add-feature.json +20 -0
  202. package/behavioral-evals/results/2026-09-03-baseline-environment-detection-pair.json +133 -0
  203. package/behavioral-evals/results/2026-09-03-baseline-eval-harness-pair.json +141 -0
  204. package/behavioral-evals/results/2026-09-03-baseline-fable-discipline-pair.json +149 -0
  205. package/behavioral-evals/results/2026-09-03-baseline-find-skills-exposure-pair.json +141 -0
  206. package/behavioral-evals/results/2026-09-03-baseline-find-skills-pair.json +141 -0
  207. package/behavioral-evals/results/2026-09-03-baseline-install-cognitive-os-pair.json +133 -0
  208. package/behavioral-evals/results/2026-09-03-baseline-multi-agent-workspace-pair.json +141 -0
  209. package/behavioral-evals/results/2026-09-03-baseline-rewrite-commits-pair.json +95 -0
  210. package/behavioral-evals/results/2026-09-03-baseline-self-evolve-pair.json +141 -0
  211. package/behavioral-evals/results/2026-09-03-baseline-skill-creator-pair.json +141 -0
  212. package/behavioral-evals/results/2026-09-03-baseline-skill-style-pair.json +133 -0
  213. package/behavioral-evals/results/2026-09-03-baseline-to-spec-pair.json +133 -0
  214. package/behavioral-evals/results/2026-09-03-baseline-to-tickets-pair.json +141 -0
  215. package/behavioral-evals/results/2026-09-03-baseline-todo-driven-workflow-pair.json +141 -0
  216. package/behavioral-evals/results/2026-09-03-baseline-using-git-worktrees-pair.json +133 -0
  217. package/behavioral-evals/results/2026-09-03-summary-both.json +41 -0
  218. package/behavioral-evals/results/2026-09-04-baseline-find-skills-pair.json +141 -0
  219. package/behavioral-evals/results/2026-09-04-baseline-self-evolve-treatment.json +80 -0
  220. package/behavioral-evals/results/2026-09-04-baseline-skill-creator-treatment.json +80 -0
  221. package/behavioral-evals/results/2026-09-04-summary-both.json +41 -0
  222. package/behavioral-evals/results/run-20260827-115509.log +287 -0
  223. package/behavioral-evals/run-with-plugin.js +300 -0
  224. package/behavioral-evals/run.js +928 -0
  225. package/behavioral-evals/transcript-parser.js +591 -0
  226. package/benchmarks/README.md +63 -0
  227. package/benchmarks/ab-test-2026-08-26/results.json +55 -0
  228. package/benchmarks/results/logs/.gitkeep +0 -0
  229. package/benchmarks/results/logs/test-a-harness.md +31 -0
  230. package/benchmarks/results/logs/test-a-vanilla.md +28 -0
  231. package/benchmarks/results/logs/test-b-micro-loop-harness.md +48 -0
  232. package/benchmarks/results/logs/test-b-micro-loop-vanilla.md +36 -0
  233. package/benchmarks/results/logs/test-c-attention-loss-harness.md +40 -0
  234. package/benchmarks/results/logs/test-c-attention-loss-vanilla.md +63 -0
  235. package/benchmarks/results/logs/test-d-knowledge-boundary-harness.md +24 -0
  236. package/benchmarks/results/logs/test-d-knowledge-boundary-vanilla.md +29 -0
  237. package/benchmarks/results/logs/test-e-shell-awareness-harness.md +20 -0
  238. package/benchmarks/results/logs/test-e-shell-awareness-vanilla.md +20 -0
  239. package/benchmarks/results/logs/test-f-fact-audit-harness.md +16 -0
  240. package/benchmarks/results/logs/test-f-fact-audit-vanilla.md +16 -0
  241. package/benchmarks/results/results.json +158 -0
  242. package/benchmarks/run.js +220 -0
  243. package/benchmarks/schema.json +44 -0
  244. package/bin/cli.js +164 -97
  245. package/ci/ab-test-cases/grill-me-adversarial.yaml +28 -0
  246. package/ci/ab-test-cases/tdd-test-first.yaml +30 -0
  247. package/ci/ab-test-cases/verification-loop-gates.yaml +38 -0
  248. package/ci/ab-test-cases/verify-before-claim-cites.yaml +19 -0
  249. package/ci/ab-test-harness.js +485 -0
  250. package/ci/action-gate-platform-doc.test.js +31 -0
  251. package/ci/action-gate.test.js +260 -0
  252. package/{eval-framework → ci}/cases/case1.json +5 -5
  253. package/ci/cases/case10-small-feature.json +5 -0
  254. package/ci/cases/case11-readme-typo-update.json +5 -0
  255. package/ci/cases/case12-chinese-macro-audit.json +5 -0
  256. package/ci/cases/case2.json +5 -0
  257. package/{eval-framework → ci}/cases/case3.json +5 -5
  258. package/{eval-framework → ci}/cases/case4-zh-tier2.json +5 -5
  259. package/{eval-framework → ci}/cases/case5-zh-tier3.json +5 -5
  260. package/ci/cases/case6-macro-skill-audit.json +5 -0
  261. package/ci/cases/case7-macro-reference-audit.json +5 -0
  262. package/ci/cases/case8-controlled-ab-benchmark.json +5 -0
  263. package/ci/cases/case9-four-issue-audit.json +5 -0
  264. package/ci/consistency-check.js +393 -0
  265. package/ci/description-collision.js +118 -0
  266. package/ci/disclosure-check.js +225 -0
  267. package/ci/doc-capability-consistency.test.js +180 -0
  268. package/ci/fable-mode-test.js +120 -0
  269. package/ci/fixtures/agency-agents/divisions.json +6 -0
  270. package/ci/fixtures/agency-agents/engineering/code-reviewer.md +11 -0
  271. package/ci/fixtures/agency-agents/testing/reality-checker.md +11 -0
  272. package/ci/fixtures/bad-frontmatter-colon/SKILL.md +16 -0
  273. package/ci/fixtures/duplicate-description/skill-a/SKILL.md +16 -0
  274. package/ci/fixtures/duplicate-description/skill-b/SKILL.md +16 -0
  275. package/ci/fixtures/phase4/parallel-disjoint.stages.json +32 -0
  276. package/ci/fixtures/phase4/parallel-read-only.stages.json +32 -0
  277. package/ci/installer-roundtrip.test.js +159 -0
  278. package/ci/invariant-routing.test.js +74 -0
  279. package/ci/issue99-runtime.test.js +218 -0
  280. package/ci/mechanism-2a-rule-of-3.test.js +46 -0
  281. package/ci/mechanism-2aa-codex-action-gate-policy.test.js +117 -0
  282. package/ci/mechanism-2ab-openai-plugin-sync-integrity.test.js +80 -0
  283. package/ci/mechanism-2b-boundary-guard.test.js +16 -0
  284. package/ci/mechanism-2c-state-persist.test.js +29 -0
  285. package/ci/mechanism-2d-fact-audit.test.js +18 -0
  286. package/ci/mechanism-2e-scope-guard.test.js +24 -0
  287. package/ci/mechanism-2f-stop-gate.test.js +23 -0
  288. package/ci/mechanism-2g-platform-ignore.test.js +88 -0
  289. package/ci/mechanism-2h-installer-manifest.test.js +147 -0
  290. package/ci/mechanism-2i-crlf-frontmatter.test.js +145 -0
  291. package/ci/mechanism-2i-install-verification.test.js +34 -0
  292. package/ci/mechanism-2j-behavioral-runner.test.js +62 -0
  293. package/ci/mechanism-2k-reference-check.test.js +75 -0
  294. package/ci/mechanism-2l-multi-agent-workspace.test.js +301 -0
  295. package/ci/mechanism-2m-self-repo-guard.test.js +65 -0
  296. package/ci/mechanism-2n-opencode-plugin.test.js +461 -0
  297. package/ci/mechanism-2o-state-home-isolation.test.js +183 -0
  298. package/ci/mechanism-2p-installer-links.test.js +183 -0
  299. package/ci/mechanism-2q-installer-global-shared-store.test.js +132 -0
  300. package/ci/mechanism-2r-uninstall-flags.test.js +148 -0
  301. package/ci/mechanism-2s-tdd-quality.test.js +203 -0
  302. package/ci/mechanism-2t-disclosure.test.js +90 -0
  303. package/ci/mechanism-2t-evidence-triage.test.js +200 -0
  304. package/ci/mechanism-2t-issue-20.test.js +75 -0
  305. package/ci/mechanism-2t-platform-skill-targets.test.js +164 -0
  306. package/ci/mechanism-2t-transcript-parser.test.js +179 -0
  307. package/ci/mechanism-2t-workspace-identity.test.js +200 -0
  308. package/ci/mechanism-2u-issue-42-regressions.test.js +239 -0
  309. package/ci/mechanism-2u-issue-44.test.js +35 -0
  310. package/ci/mechanism-2v-verifier-boundaries.test.js +138 -0
  311. package/ci/mechanism-2w-platform-compatibility.test.js +148 -0
  312. package/ci/mechanism-2x-release-automation.test.js +157 -0
  313. package/ci/mechanism-2x-release-commit-message.test.js +66 -0
  314. package/ci/mechanism-2y-release-tooling-compat.test.js +47 -0
  315. package/ci/mechanism-2z-claude-plugin-manifest.test.js +29 -0
  316. package/ci/mechanism-test.js +71 -0
  317. package/ci/mutation-check.js +53 -0
  318. package/ci/mutation-loader.js +57 -0
  319. package/ci/negative-controls.js +114 -0
  320. package/ci/openai-plugin-package.test.js +182 -0
  321. package/ci/openai-plugin-runtime.test.js +175 -0
  322. package/ci/openai-plugin-submission.test.js +101 -0
  323. package/ci/phase4-ensemble-memory.test.js +307 -0
  324. package/ci/reference-check.js +160 -0
  325. package/ci/release-consistency-check.js +146 -0
  326. package/ci/router-phase2-safety.test.js +81 -0
  327. package/ci/router-workflow-plan.test.js +264 -0
  328. package/ci/routing-check.js +91 -0
  329. package/{eval-framework → ci}/runner.js +57 -56
  330. package/ci/skill-routing-check.js +135 -0
  331. package/ci/test-behavioral-case-validation.js +62 -0
  332. package/ci/test-helper.js +114 -0
  333. package/ci/test-negative-controls.js +148 -0
  334. package/ci/workflow-plan-consumers-edge.test.js +72 -0
  335. package/ci/workflow-plan-consumers.test.js +276 -0
  336. package/commitlint.config.js +25 -0
  337. package/docker-verify.sh +266 -266
  338. package/docs/architecture.md +179 -95
  339. package/docs/audit.md +55 -0
  340. package/docs/codex-action-gate.md +90 -0
  341. package/docs/issue-20-rollup.md +47 -0
  342. package/docs/mechanism-first-skill-mesh.md +165 -0
  343. package/docs/openai-plugin.md +118 -0
  344. package/docs/philosophy.md +58 -45
  345. package/docs/platform-capabilities.md +73 -0
  346. package/docs/platform-compatibility.json +270 -0
  347. package/docs/reflection.md +63 -34
  348. package/docs/release-evidence.md +25 -0
  349. package/docs/routing.md +171 -45
  350. package/docs/security-audit.md +38 -0
  351. package/docs/troubleshooting.md +320 -0
  352. package/docs/verifier-boundaries.md +45 -0
  353. package/docs/workflows/environment-detection.md +10 -10
  354. package/docs/workflows/eval-harness.md +13 -6
  355. package/docs/workflows/fable-discipline.md +9 -9
  356. package/docs/workflows/fable-mode.md +85 -32
  357. package/docs/workflows/find-skills.md +76 -0
  358. package/docs/workflows/git-commit.md +18 -9
  359. package/docs/workflows/grill-me.md +11 -11
  360. package/docs/workflows/grill-with-docs.md +14 -7
  361. package/docs/workflows/harness-everything.md +95 -92
  362. package/docs/workflows/improve-codebase-architecture.md +15 -7
  363. package/docs/workflows/install-cognitive-os.md +60 -60
  364. package/docs/workflows/multi-agent-workspace.md +54 -0
  365. package/docs/workflows/repo-docs.md +12 -7
  366. package/docs/workflows/security-review.md +26 -7
  367. package/docs/workflows/self-evolve.md +53 -37
  368. package/docs/workflows/skill-creator.md +80 -75
  369. package/docs/workflows/skill-style.md +64 -64
  370. package/docs/workflows/tdd.md +78 -39
  371. package/docs/workflows/to-spec.md +83 -0
  372. package/docs/workflows/to-tickets.md +80 -0
  373. package/docs/workflows/todo-driven-workflow.md +67 -36
  374. package/docs/workflows/using-git-worktrees.md +12 -8
  375. package/docs/workflows/verification-loop.md +16 -14
  376. package/docs/workflows/verify-before-claim.md +12 -8
  377. package/docs/workflows/zoom-out.md +22 -10
  378. package/environment-detection/SKILL.md +25 -61
  379. package/environment-detection/references/shell-syntax-rules.md +76 -0
  380. package/environment-detection/scripts/preflight.js +103 -93
  381. package/eval-framework/fixtures/bad-frontmatter-colon/SKILL.md +14 -0
  382. package/eval-framework/fixtures/duplicate-desc-skill-a/SKILL.md +14 -0
  383. package/eval-framework/fixtures/duplicate-desc-skill-b/SKILL.md +14 -0
  384. package/eval-framework/fixtures/missing-delimiter/SKILL.md +13 -0
  385. package/eval-framework/fixtures/over-budget-skill/SKILL.md +36 -0
  386. package/eval-harness/SKILL.md +34 -51
  387. package/eval-harness/references/scoring-rubric.md +68 -0
  388. package/eval-harness/scripts/evaluate.js +79 -61
  389. package/evals/environment-detection/eval.yaml +26 -0
  390. package/evals/environment-detection/tasks/negative-trigger.yaml +16 -0
  391. package/evals/environment-detection/tasks/positive-trigger.yaml +15 -0
  392. package/evals/eval-harness/eval.yaml +26 -0
  393. package/evals/eval-harness/tasks/negative-trigger.yaml +16 -0
  394. package/evals/eval-harness/tasks/positive-trigger.yaml +15 -0
  395. package/evals/evaluation-reports.md +14 -0
  396. package/evals/fable-discipline/eval.yaml +26 -0
  397. package/evals/fable-discipline/tasks/negative-trigger.yaml +16 -0
  398. package/evals/fable-discipline/tasks/positive-trigger.yaml +15 -0
  399. package/evals/fable-mode/eval.yaml +26 -0
  400. package/evals/fable-mode/tasks/negative-tier2-no-model.yaml +18 -0
  401. package/evals/fable-mode/tasks/negative-trigger.yaml +16 -0
  402. package/evals/fable-mode/tasks/positive-haiku-mode.yaml +16 -0
  403. package/evals/fable-mode/tasks/positive-opus-mode.yaml +16 -0
  404. package/evals/fable-mode/tasks/positive-sonnet-mode.yaml +16 -0
  405. package/evals/fable-mode/tasks/positive-trigger.yaml +15 -0
  406. package/evals/find-skills/eval.yaml +26 -0
  407. package/evals/find-skills/tasks/negative-trigger.yaml +16 -0
  408. package/evals/find-skills/tasks/positive-trigger.yaml +15 -0
  409. package/evals/git-commit/eval.yaml +26 -0
  410. package/evals/git-commit/tasks/negative-trigger.yaml +16 -0
  411. package/evals/git-commit/tasks/positive-trigger.yaml +15 -0
  412. package/evals/grill-me/eval.yaml +26 -0
  413. package/evals/grill-me/tasks/negative-trigger.yaml +16 -0
  414. package/evals/grill-me/tasks/positive-trigger.yaml +15 -0
  415. package/evals/grill-with-docs/eval.yaml +26 -0
  416. package/evals/grill-with-docs/tasks/negative-trigger.yaml +16 -0
  417. package/evals/grill-with-docs/tasks/positive-trigger.yaml +15 -0
  418. package/evals/harness-everything/eval.yaml +26 -0
  419. package/evals/harness-everything/tasks/negative-non-software.yaml +16 -0
  420. package/evals/harness-everything/tasks/negative-skill-indicated.yaml +15 -0
  421. package/evals/harness-everything/tasks/positive-tier1.yaml +16 -0
  422. package/evals/harness-everything/tasks/positive-tier2.yaml +16 -0
  423. package/evals/harness-everything/tasks/positive-tier3.yaml +16 -0
  424. package/evals/improve-codebase-architecture/eval.yaml +26 -0
  425. package/evals/improve-codebase-architecture/tasks/negative-trigger.yaml +16 -0
  426. package/evals/improve-codebase-architecture/tasks/positive-trigger.yaml +15 -0
  427. package/evals/install-cognitive-os/eval.yaml +26 -0
  428. package/evals/install-cognitive-os/tasks/negative-trigger.yaml +16 -0
  429. package/evals/install-cognitive-os/tasks/positive-trigger.yaml +15 -0
  430. package/evals/multi-agent-workspace/eval.yaml +26 -0
  431. package/evals/multi-agent-workspace/tasks/negative-trigger.yaml +15 -0
  432. package/evals/multi-agent-workspace/tasks/positive-trigger.yaml +15 -0
  433. package/evals/repo-docs/eval.yaml +26 -0
  434. package/evals/repo-docs/tasks/negative-trigger.yaml +16 -0
  435. package/evals/repo-docs/tasks/positive-trigger.yaml +15 -0
  436. package/evals/rewrite-commits/eval.yaml +26 -0
  437. package/evals/rewrite-commits/tasks/negative-trigger.yaml +16 -0
  438. package/evals/rewrite-commits/tasks/positive-trigger.yaml +15 -0
  439. package/evals/security-review/eval.yaml +26 -0
  440. package/evals/security-review/tasks/negative-trigger.yaml +16 -0
  441. package/evals/security-review/tasks/positive-trigger.yaml +15 -0
  442. package/evals/self-evolve/eval.yaml +26 -0
  443. package/evals/self-evolve/tasks/negative-trigger.yaml +16 -0
  444. package/evals/self-evolve/tasks/positive-trigger.yaml +15 -0
  445. package/evals/skill-creator/eval.yaml +26 -0
  446. package/evals/skill-creator/tasks/negative-trigger.yaml +16 -0
  447. package/evals/skill-creator/tasks/positive-trigger.yaml +15 -0
  448. package/evals/skill-style/eval.yaml +26 -0
  449. package/evals/skill-style/tasks/negative-trigger.yaml +16 -0
  450. package/evals/skill-style/tasks/positive-trigger.yaml +15 -0
  451. package/evals/tdd/eval.yaml +26 -0
  452. package/evals/tdd/tasks/negative-trivia.yaml +16 -0
  453. package/evals/tdd/tasks/positive-bugfix.yaml +15 -0
  454. package/evals/tdd/tasks/positive-feature.yaml +16 -0
  455. package/evals/tdd/tasks/positive-refactor.yaml +15 -0
  456. package/evals/to-spec/eval.yaml +26 -0
  457. package/evals/to-spec/tasks/negative-trigger.yaml +16 -0
  458. package/evals/to-spec/tasks/positive-trigger.yaml +15 -0
  459. package/evals/to-tickets/eval.yaml +26 -0
  460. package/evals/to-tickets/tasks/negative-trigger.yaml +15 -0
  461. package/evals/to-tickets/tasks/positive-trigger.yaml +15 -0
  462. package/evals/todo-driven-workflow/eval.yaml +26 -0
  463. package/evals/todo-driven-workflow/tasks/negative-single-edit.yaml +16 -0
  464. package/evals/todo-driven-workflow/tasks/positive-multi-step.yaml +15 -0
  465. package/evals/todo-driven-workflow/tasks/positive-refactor-flow.yaml +15 -0
  466. package/evals/using-git-worktrees/eval.yaml +26 -0
  467. package/evals/using-git-worktrees/tasks/negative-trigger.yaml +15 -0
  468. package/evals/using-git-worktrees/tasks/positive-trigger.yaml +15 -0
  469. package/evals/verification-loop/eval.yaml +26 -0
  470. package/evals/verification-loop/tasks/negative-discussion.yaml +16 -0
  471. package/evals/verification-loop/tasks/positive-pre-ship.yaml +15 -0
  472. package/evals/verify-before-claim/eval.yaml +26 -0
  473. package/evals/verify-before-claim/tasks/negative-trigger.yaml +16 -0
  474. package/evals/verify-before-claim/tasks/positive-trigger.yaml +15 -0
  475. package/evals/zoom-out/eval.yaml +26 -0
  476. package/evals/zoom-out/tasks/negative-trivial-edit.yaml +16 -0
  477. package/evals/zoom-out/tasks/positive-repeated-failure.yaml +15 -0
  478. package/fable-discipline/SKILL.md +31 -28
  479. package/fable-discipline/references/discipline-rules.md +22 -0
  480. package/fable-mode/CONTRACT-FORMAT.md +165 -51
  481. package/fable-mode/SKILL.md +31 -43
  482. package/fable-mode/agents/fable-orchestrator.md +159 -92
  483. package/fable-mode/agents/fable-verifier.md +32 -32
  484. package/fable-mode/agents/fable-worker-haiku.md +52 -48
  485. package/fable-mode/agents/fable-worker-sonnet.md +53 -49
  486. package/fable-mode/execution-guardrails/SKILL.md +43 -87
  487. package/fable-mode/execution-guardrails/references/guardrail-rules.md +56 -0
  488. package/fable-mode/fable-haiku/SKILL.md +36 -48
  489. package/fable-mode/fable-haiku/references/routing-notes.md +26 -0
  490. package/fable-mode/fable-opus/SKILL.md +36 -0
  491. package/fable-mode/fable-sonnet/SKILL.md +36 -0
  492. package/fable-mode/model-matrix.json +32 -0
  493. package/fable-mode/references/execution-phases.md +37 -0
  494. package/fable-mode/references/model-matrix.md +39 -0
  495. package/fable-mode/scripts/ensemble-review.js +371 -0
  496. package/fable-mode/scripts/model-selector.js +260 -0
  497. package/fable-mode/scripts/workflow-plan-consumer.js +337 -0
  498. package/find-skills/SKILL.md +48 -0
  499. package/find-skills/references/discovery-flow.md +102 -0
  500. package/find-skills/scripts/use-skill.js +112 -0
  501. package/git-commit/SKILL.md +28 -45
  502. package/git-commit/guides/ANGULAR_STYLE.md +60 -60
  503. package/git-commit/guides/COMMIT_GENERATION.md +23 -23
  504. package/git-commit/guides/LANGUAGE_DETECTION.md +16 -16
  505. package/git-commit/guides/MAIN_REPO.md +14 -14
  506. package/git-commit/guides/SUBMODULES.md +27 -27
  507. package/git-commit/references/commit-flow.md +30 -0
  508. package/grill-me/SKILL.md +34 -27
  509. package/grill-me/references/grilling-playbook.md +38 -0
  510. package/grill-with-docs/ADR-FORMAT.md +47 -47
  511. package/grill-with-docs/CONTEXT-FORMAT.md +63 -63
  512. package/grill-with-docs/SKILL.md +33 -91
  513. package/grill-with-docs/references/session-playbook.md +99 -0
  514. package/harness-everything/SKILL.md +27 -125
  515. package/harness-everything/examples/README.md +9 -9
  516. package/harness-everything/examples/todo-workflow.md +22 -22
  517. package/harness-everything/references/router-workflow-plan.md +128 -0
  518. package/harness-everything/references/skill-registry.md +39 -0
  519. package/harness-everything/references/triage-and-tiers.md +193 -0
  520. package/harness-everything/schemas/router-task-shape.schema.json +126 -0
  521. package/harness-everything/schemas/router-workflow-plan.schema.json +149 -0
  522. package/harness-everything/scripts/bootstrap.js +212 -140
  523. package/harness-everything/scripts/ensemble-policy.js +102 -0
  524. package/harness-everything/scripts/kernel-router.js +212 -0
  525. package/harness-everything/scripts/router-contract.js +612 -0
  526. package/harness-everything/scripts/routing-keywords.json +320 -0
  527. package/harness-everything/scripts/self-heal.js +250 -189
  528. package/harness-everything/scripts/tier-router.js +459 -219
  529. package/harness-everything/scripts/verify-gate.js +100 -31
  530. package/hooks/hooks.json +200 -155
  531. package/hooks/scripts/action-gate-rules.json +77 -0
  532. package/hooks/scripts/action-gate.js +528 -0
  533. package/hooks/scripts/atomic-commit-check.js +68 -68
  534. package/hooks/scripts/boundary-guard.js +55 -55
  535. package/hooks/scripts/codex-action-gate-post.js +46 -0
  536. package/hooks/scripts/codex-permission-request.js +140 -0
  537. package/hooks/scripts/context-compact.js +115 -114
  538. package/hooks/scripts/contract-test.js +146 -79
  539. package/hooks/scripts/depth-guard.js +65 -65
  540. package/hooks/scripts/lib/fable-contracts.js +121 -0
  541. package/hooks/scripts/lib/harness-state.js +275 -228
  542. package/hooks/scripts/lib/platforms/claude.js +97 -61
  543. package/hooks/scripts/lib/platforms/codex.js +42 -50
  544. package/hooks/scripts/lib/platforms/continue.js +30 -12
  545. package/hooks/scripts/lib/platforms/copilot.js +36 -42
  546. package/hooks/scripts/lib/platforms/cursor.js +23 -5
  547. package/hooks/scripts/lib/platforms/hermes.js +108 -57
  548. package/hooks/scripts/lib/platforms/index.js +17 -17
  549. package/hooks/scripts/lib/platforms/worktrees.js +20 -20
  550. package/hooks/scripts/openai-session-start.js +81 -0
  551. package/hooks/scripts/reset-circuit-breaker.js +30 -30
  552. package/hooks/scripts/rule-of-3-tracker.js +125 -91
  553. package/hooks/scripts/rule-of-3.js +195 -176
  554. package/hooks/scripts/state-persist.js +100 -99
  555. package/hooks/scripts/stop-gate.js +80 -80
  556. package/hooks/scripts/subagent-scope-guard.js +194 -88
  557. package/hooks/scripts/subagent-start.js +19 -0
  558. package/hooks/scripts/subagent-stop.js +6 -0
  559. package/improve-codebase-architecture/SKILL.md +29 -35
  560. package/improve-codebase-architecture/guides/DEEPENING.md +37 -37
  561. package/improve-codebase-architecture/guides/HTML-REPORT.md +123 -123
  562. package/improve-codebase-architecture/guides/INTERFACE-DESIGN.md +44 -44
  563. package/improve-codebase-architecture/guides/LANGUAGE.md +53 -53
  564. package/improve-codebase-architecture/guides/refactoring-workflow.md +59 -0
  565. package/improve-codebase-architecture/references/README.md +16 -0
  566. package/install-cognitive-os/SKILL.md +30 -81
  567. package/install-cognitive-os/references/cognitive-loop.md +27 -0
  568. package/multi-agent-workspace/NOTICE-agency-agents.md +27 -0
  569. package/multi-agent-workspace/SKILL.md +38 -0
  570. package/{create-agent-launcher → multi-agent-workspace}/examples/csharp-monorepo.md +123 -123
  571. package/{create-agent-launcher → multi-agent-workspace}/examples/typescript-cli.md +126 -126
  572. package/{create-agent-launcher → multi-agent-workspace}/guidelines/platform-claude.md +77 -77
  573. package/{create-agent-launcher → multi-agent-workspace}/guidelines/platform-codex.md +146 -146
  574. package/multi-agent-workspace/guidelines/platform-continue.md +38 -0
  575. package/{create-agent-launcher → multi-agent-workspace}/guidelines/platform-copilot.md +74 -74
  576. package/multi-agent-workspace/guidelines/platform-cursor.md +40 -0
  577. package/multi-agent-workspace/guidelines/platform-hermes.md +38 -0
  578. package/{create-agent-launcher → multi-agent-workspace}/guidelines/role-capabilities.md +108 -108
  579. package/multi-agent-workspace/references/agency-agents.md +31 -0
  580. package/multi-agent-workspace/references/architecture-guide.md +36 -0
  581. package/multi-agent-workspace/references/orchestration.md +65 -0
  582. package/multi-agent-workspace/scripts/agency-catalog.js +190 -0
  583. package/multi-agent-workspace/scripts/consume-workflow-plan.js +129 -0
  584. package/multi-agent-workspace/scripts/index_memory.js +104 -0
  585. package/multi-agent-workspace/scripts/path-boundary.js +59 -0
  586. package/multi-agent-workspace/scripts/project-docs-resolver.js +357 -0
  587. package/multi-agent-workspace/scripts/scaffold.js +408 -0
  588. package/multi-agent-workspace/templates/AGENTS.md +20 -0
  589. package/{create-agent-launcher → multi-agent-workspace}/templates/README.md +17 -17
  590. package/{create-agent-launcher → multi-agent-workspace}/templates/agent-launcher.md +67 -67
  591. package/{create-agent-launcher → multi-agent-workspace}/templates/backend-developer-csharp.md +55 -55
  592. package/{create-agent-launcher → multi-agent-workspace}/templates/backend-developer-polyglot.md +61 -61
  593. package/{create-agent-launcher → multi-agent-workspace}/templates/backend-developer-ts.md +56 -56
  594. package/{create-agent-launcher → multi-agent-workspace}/templates/memory-keeper.md +36 -36
  595. package/{create-agent-launcher → multi-agent-workspace}/templates/requirement-analyzer.md +65 -65
  596. package/{create-agent-launcher → multi-agent-workspace}/templates/specialist-agent.md +34 -34
  597. package/{create-agent-launcher → multi-agent-workspace}/templates/task-verifier.md +55 -55
  598. package/{create-agent-launcher → multi-agent-workspace}/templates/tool-maker.md +32 -32
  599. package/multi-agent-workspace/workflows/01-init.md +22 -0
  600. package/multi-agent-workspace/workflows/02-analysis.md +23 -0
  601. package/multi-agent-workspace/workflows/03-generation.md +25 -0
  602. package/multi-agent-workspace/workflows/04-launcher.md +16 -0
  603. package/opencode-plugin/README.md +137 -0
  604. package/opencode-plugin/hooks/verify.js +190 -0
  605. package/opencode-plugin/index.mjs +547 -0
  606. package/opencode-plugin/plugin.json +9 -0
  607. package/package.json +63 -22
  608. package/plugins/harness-everything/.codex-plugin/plugin.json +39 -0
  609. package/plugins/harness-everything/hooks/hooks.json +197 -0
  610. package/plugins/harness-everything/hooks/scripts/action-gate-rules.json +77 -0
  611. package/plugins/harness-everything/hooks/scripts/action-gate.js +528 -0
  612. package/plugins/harness-everything/hooks/scripts/atomic-commit-check.js +68 -0
  613. package/plugins/harness-everything/hooks/scripts/boundary-guard.js +55 -0
  614. package/plugins/harness-everything/hooks/scripts/codex-action-gate-post.js +46 -0
  615. package/plugins/harness-everything/hooks/scripts/codex-permission-request.js +140 -0
  616. package/plugins/harness-everything/hooks/scripts/context-compact.js +115 -0
  617. package/plugins/harness-everything/hooks/scripts/contract-test.js +146 -0
  618. package/plugins/harness-everything/hooks/scripts/depth-guard.js +65 -0
  619. package/plugins/harness-everything/hooks/scripts/lib/fable-contracts.js +121 -0
  620. package/plugins/harness-everything/hooks/scripts/lib/harness-state.js +275 -0
  621. package/plugins/harness-everything/hooks/scripts/lib/platforms/claude.js +168 -0
  622. package/plugins/harness-everything/hooks/scripts/lib/platforms/codex.js +88 -0
  623. package/plugins/harness-everything/hooks/scripts/lib/platforms/continue.js +91 -0
  624. package/plugins/harness-everything/hooks/scripts/lib/platforms/copilot.js +85 -0
  625. package/plugins/harness-everything/hooks/scripts/lib/platforms/cursor.js +89 -0
  626. package/plugins/harness-everything/hooks/scripts/lib/platforms/hermes.js +108 -0
  627. package/plugins/harness-everything/hooks/scripts/lib/platforms/index.js +17 -0
  628. package/plugins/harness-everything/hooks/scripts/lib/platforms/worktrees.js +20 -0
  629. package/plugins/harness-everything/hooks/scripts/lib/workspace.js +290 -0
  630. package/plugins/harness-everything/hooks/scripts/openai-session-start.js +81 -0
  631. package/plugins/harness-everything/hooks/scripts/reset-circuit-breaker.js +30 -0
  632. package/plugins/harness-everything/hooks/scripts/rule-of-3-tracker.js +125 -0
  633. package/plugins/harness-everything/hooks/scripts/rule-of-3.js +195 -0
  634. package/plugins/harness-everything/hooks/scripts/state-persist.js +100 -0
  635. package/plugins/harness-everything/hooks/scripts/stop-gate.js +80 -0
  636. package/plugins/harness-everything/hooks/scripts/subagent-scope-guard.js +194 -0
  637. package/plugins/harness-everything/hooks/scripts/subagent-start.js +19 -0
  638. package/plugins/harness-everything/hooks/scripts/subagent-stop.js +6 -0
  639. package/plugins/harness-everything/hooks/session-start.js +81 -0
  640. package/plugins/harness-everything/plugin.json +43 -0
  641. package/plugins/harness-everything/scripts/lib/advisory-text.js +174 -0
  642. package/plugins/harness-everything/skills/environment-detection/SKILL.md +41 -0
  643. package/plugins/harness-everything/skills/environment-detection/references/shell-syntax-rules.md +76 -0
  644. package/plugins/harness-everything/skills/environment-detection/scripts/preflight.js +103 -0
  645. package/plugins/harness-everything/skills/eval-harness/SKILL.md +42 -0
  646. package/plugins/harness-everything/skills/eval-harness/references/scoring-rubric.md +68 -0
  647. package/plugins/harness-everything/skills/eval-harness/scripts/evaluate.js +79 -0
  648. package/plugins/harness-everything/skills/fable-discipline/SKILL.md +41 -0
  649. package/plugins/harness-everything/skills/fable-discipline/references/discipline-rules.md +22 -0
  650. package/plugins/harness-everything/skills/fable-mode/CONTRACT-FORMAT.md +165 -0
  651. package/plugins/harness-everything/skills/fable-mode/SKILL.md +39 -0
  652. package/plugins/harness-everything/skills/fable-mode/agents/fable-orchestrator.md +159 -0
  653. package/plugins/harness-everything/skills/fable-mode/agents/fable-verifier.md +32 -0
  654. package/plugins/harness-everything/skills/fable-mode/agents/fable-worker-haiku.md +52 -0
  655. package/plugins/harness-everything/skills/fable-mode/agents/fable-worker-sonnet.md +53 -0
  656. package/plugins/harness-everything/skills/fable-mode/execution-guardrails/SKILL.md +43 -0
  657. package/plugins/harness-everything/skills/fable-mode/execution-guardrails/references/guardrail-rules.md +56 -0
  658. package/plugins/harness-everything/skills/fable-mode/fable-haiku/SKILL.md +36 -0
  659. package/plugins/harness-everything/skills/fable-mode/fable-haiku/references/routing-notes.md +26 -0
  660. package/plugins/harness-everything/skills/fable-mode/fable-opus/SKILL.md +36 -0
  661. package/plugins/harness-everything/skills/fable-mode/fable-sonnet/SKILL.md +36 -0
  662. package/plugins/harness-everything/skills/fable-mode/model-matrix.json +32 -0
  663. package/plugins/harness-everything/skills/fable-mode/references/execution-phases.md +37 -0
  664. package/plugins/harness-everything/skills/fable-mode/references/model-matrix.md +39 -0
  665. package/plugins/harness-everything/skills/fable-mode/scripts/ensemble-review.js +371 -0
  666. package/plugins/harness-everything/skills/fable-mode/scripts/model-selector.js +260 -0
  667. package/plugins/harness-everything/skills/fable-mode/scripts/workflow-plan-consumer.js +337 -0
  668. package/plugins/harness-everything/skills/find-skills/SKILL.md +48 -0
  669. package/plugins/harness-everything/skills/find-skills/references/discovery-flow.md +102 -0
  670. package/plugins/harness-everything/skills/find-skills/scripts/use-skill.js +112 -0
  671. package/plugins/harness-everything/skills/git-commit/SKILL.md +38 -0
  672. package/plugins/harness-everything/skills/git-commit/guides/ANGULAR_STYLE.md +61 -0
  673. package/plugins/harness-everything/skills/git-commit/guides/COMMIT_GENERATION.md +24 -0
  674. package/plugins/harness-everything/skills/git-commit/guides/LANGUAGE_DETECTION.md +16 -0
  675. package/plugins/harness-everything/skills/git-commit/guides/MAIN_REPO.md +15 -0
  676. package/plugins/harness-everything/skills/git-commit/guides/SUBMODULES.md +27 -0
  677. package/plugins/harness-everything/skills/git-commit/references/commit-flow.md +30 -0
  678. package/plugins/harness-everything/skills/grill-me/SKILL.md +43 -0
  679. package/plugins/harness-everything/skills/grill-me/references/grilling-playbook.md +38 -0
  680. package/plugins/harness-everything/skills/grill-with-docs/ADR-FORMAT.md +47 -0
  681. package/plugins/harness-everything/skills/grill-with-docs/CONTEXT-FORMAT.md +63 -0
  682. package/plugins/harness-everything/skills/grill-with-docs/SKILL.md +41 -0
  683. package/plugins/harness-everything/skills/grill-with-docs/references/session-playbook.md +99 -0
  684. package/plugins/harness-everything/skills/harness-everything/SKILL.md +43 -0
  685. package/plugins/harness-everything/skills/harness-everything/examples/README.md +9 -0
  686. package/plugins/harness-everything/skills/harness-everything/examples/todo-workflow.md +22 -0
  687. package/plugins/harness-everything/skills/harness-everything/references/router-workflow-plan.md +128 -0
  688. package/plugins/harness-everything/skills/harness-everything/references/skill-registry.md +39 -0
  689. package/plugins/harness-everything/skills/harness-everything/references/triage-and-tiers.md +193 -0
  690. package/plugins/harness-everything/skills/harness-everything/schemas/router-task-shape.schema.json +126 -0
  691. package/plugins/harness-everything/skills/harness-everything/schemas/router-workflow-plan.schema.json +149 -0
  692. package/plugins/harness-everything/skills/harness-everything/scripts/bootstrap.js +212 -0
  693. package/plugins/harness-everything/skills/harness-everything/scripts/ensemble-policy.js +102 -0
  694. package/plugins/harness-everything/skills/harness-everything/scripts/kernel-router.js +212 -0
  695. package/plugins/harness-everything/skills/harness-everything/scripts/router-contract.js +612 -0
  696. package/plugins/harness-everything/skills/harness-everything/scripts/routing-keywords.json +320 -0
  697. package/plugins/harness-everything/skills/harness-everything/scripts/self-heal.js +250 -0
  698. package/plugins/harness-everything/skills/harness-everything/scripts/tier-router.js +459 -0
  699. package/plugins/harness-everything/skills/harness-everything/scripts/verify-gate.js +100 -0
  700. package/plugins/harness-everything/skills/improve-codebase-architecture/SKILL.md +38 -0
  701. package/plugins/harness-everything/skills/improve-codebase-architecture/guides/DEEPENING.md +37 -0
  702. package/plugins/harness-everything/skills/improve-codebase-architecture/guides/HTML-REPORT.md +123 -0
  703. package/plugins/harness-everything/skills/improve-codebase-architecture/guides/INTERFACE-DESIGN.md +44 -0
  704. package/plugins/harness-everything/skills/improve-codebase-architecture/guides/LANGUAGE.md +53 -0
  705. package/plugins/harness-everything/skills/improve-codebase-architecture/guides/refactoring-workflow.md +59 -0
  706. package/plugins/harness-everything/skills/improve-codebase-architecture/references/README.md +16 -0
  707. package/plugins/harness-everything/skills/install-cognitive-os/SKILL.md +44 -0
  708. package/plugins/harness-everything/skills/install-cognitive-os/references/cognitive-loop.md +27 -0
  709. package/plugins/harness-everything/skills/multi-agent-workspace/NOTICE-agency-agents.md +27 -0
  710. package/plugins/harness-everything/skills/multi-agent-workspace/SKILL.md +38 -0
  711. package/plugins/harness-everything/skills/multi-agent-workspace/examples/csharp-monorepo.md +123 -0
  712. package/plugins/harness-everything/skills/multi-agent-workspace/examples/typescript-cli.md +126 -0
  713. package/plugins/harness-everything/skills/multi-agent-workspace/guidelines/platform-claude.md +77 -0
  714. package/plugins/harness-everything/skills/multi-agent-workspace/guidelines/platform-codex.md +146 -0
  715. package/plugins/harness-everything/skills/multi-agent-workspace/guidelines/platform-continue.md +38 -0
  716. package/plugins/harness-everything/skills/multi-agent-workspace/guidelines/platform-copilot.md +74 -0
  717. package/plugins/harness-everything/skills/multi-agent-workspace/guidelines/platform-cursor.md +40 -0
  718. package/plugins/harness-everything/skills/multi-agent-workspace/guidelines/platform-hermes.md +38 -0
  719. package/plugins/harness-everything/skills/multi-agent-workspace/guidelines/role-capabilities.md +108 -0
  720. package/plugins/harness-everything/skills/multi-agent-workspace/references/agency-agents.md +31 -0
  721. package/plugins/harness-everything/skills/multi-agent-workspace/references/architecture-guide.md +36 -0
  722. package/plugins/harness-everything/skills/multi-agent-workspace/references/orchestration.md +65 -0
  723. package/plugins/harness-everything/skills/multi-agent-workspace/scripts/agency-catalog.js +190 -0
  724. package/plugins/harness-everything/skills/multi-agent-workspace/scripts/consume-workflow-plan.js +129 -0
  725. package/plugins/harness-everything/skills/multi-agent-workspace/scripts/index_memory.js +104 -0
  726. package/plugins/harness-everything/skills/multi-agent-workspace/scripts/path-boundary.js +59 -0
  727. package/plugins/harness-everything/skills/multi-agent-workspace/scripts/project-docs-resolver.js +357 -0
  728. package/plugins/harness-everything/skills/multi-agent-workspace/scripts/scaffold.js +408 -0
  729. package/plugins/harness-everything/skills/multi-agent-workspace/templates/AGENTS.md +20 -0
  730. package/plugins/harness-everything/skills/multi-agent-workspace/templates/README.md +17 -0
  731. package/plugins/harness-everything/skills/multi-agent-workspace/templates/agent-launcher.md +68 -0
  732. package/plugins/harness-everything/skills/multi-agent-workspace/templates/backend-developer-csharp.md +56 -0
  733. package/plugins/harness-everything/skills/multi-agent-workspace/templates/backend-developer-polyglot.md +62 -0
  734. package/plugins/harness-everything/skills/multi-agent-workspace/templates/backend-developer-ts.md +57 -0
  735. package/plugins/harness-everything/skills/multi-agent-workspace/templates/memory-keeper.md +37 -0
  736. package/plugins/harness-everything/skills/multi-agent-workspace/templates/requirement-analyzer.md +66 -0
  737. package/plugins/harness-everything/skills/multi-agent-workspace/templates/specialist-agent.md +34 -0
  738. package/plugins/harness-everything/skills/multi-agent-workspace/templates/task-verifier.md +56 -0
  739. package/plugins/harness-everything/skills/multi-agent-workspace/templates/tool-maker.md +33 -0
  740. package/plugins/harness-everything/skills/multi-agent-workspace/workflows/01-init.md +22 -0
  741. package/plugins/harness-everything/skills/multi-agent-workspace/workflows/02-analysis.md +23 -0
  742. package/plugins/harness-everything/skills/multi-agent-workspace/workflows/03-generation.md +25 -0
  743. package/plugins/harness-everything/skills/multi-agent-workspace/workflows/04-launcher.md +16 -0
  744. package/plugins/harness-everything/skills/repo-docs/SKILL.md +42 -0
  745. package/plugins/harness-everything/skills/repo-docs/references/process-guide.md +60 -0
  746. package/plugins/harness-everything/skills/repo-docs/templates/agents-template.md +264 -0
  747. package/plugins/harness-everything/skills/repo-docs/templates/knowledge-agents-template.md +50 -0
  748. package/plugins/harness-everything/skills/repo-docs/templates/knowledge-readme-template.md +83 -0
  749. package/plugins/harness-everything/skills/repo-docs/templates/manual-recon.md +117 -0
  750. package/plugins/harness-everything/skills/repo-docs/templates/multi-skills-readme-template.md +72 -0
  751. package/plugins/harness-everything/skills/repo-docs/templates/product-agents-template.md +78 -0
  752. package/plugins/harness-everything/skills/repo-docs/templates/product-readme-template.md +95 -0
  753. package/plugins/harness-everything/skills/repo-docs/templates/readme-template.md +136 -0
  754. package/plugins/harness-everything/skills/repo-docs/templates/single-skill-readme-template.md +70 -0
  755. package/plugins/harness-everything/skills/repo-docs/templates/skills-agents-template.md +60 -0
  756. package/plugins/harness-everything/skills/rewrite-commits/SKILL.md +37 -0
  757. package/plugins/harness-everything/skills/rewrite-commits/references/workflow.md +18 -0
  758. package/plugins/harness-everything/skills/security-review/SKILL.md +47 -0
  759. package/plugins/harness-everything/skills/security-review/guides/OWASP-PATTERNS.md +31 -0
  760. package/plugins/harness-everything/skills/security-review/guides/STRIDE-THREAT-MODEL.md +21 -0
  761. package/plugins/harness-everything/skills/security-review/references/security-checklist.md +478 -0
  762. package/plugins/harness-everything/skills/security-review/scripts/audit-secrets.js +85 -0
  763. package/plugins/harness-everything/skills/self-evolve/SKILL.md +45 -0
  764. package/plugins/harness-everything/skills/self-evolve/references/memory-resolution.md +61 -0
  765. package/plugins/harness-everything/skills/self-evolve/scripts/persist-memory.js +272 -0
  766. package/plugins/harness-everything/skills/self-evolve/scripts/register-dynamic-skill.js +253 -0
  767. package/plugins/harness-everything/skills/self-evolve/scripts/self-regression.js +125 -0
  768. package/plugins/harness-everything/skills/skill-creator/SKILL.md +47 -0
  769. package/plugins/harness-everything/skills/skill-creator/references/authoring-workflow.md +14 -0
  770. package/plugins/harness-everything/skills/skill-creator/references/dynamic-generation-contract.md +21 -0
  771. package/plugins/harness-everything/skills/skill-creator/references/quality-checklist.md +16 -0
  772. package/plugins/harness-everything/skills/skill-creator/references/quality-principles.md +81 -0
  773. package/plugins/harness-everything/skills/skill-creator/references/testing-workflow.md +27 -0
  774. package/plugins/harness-everything/skills/skill-style/SKILL.md +42 -0
  775. package/plugins/harness-everything/skills/skill-style/references/style-guide.md +62 -0
  776. package/plugins/harness-everything/skills/tdd/SKILL.md +38 -0
  777. package/plugins/harness-everything/skills/tdd/fixtures/pass-evidence.json +100 -0
  778. package/plugins/harness-everything/skills/tdd/fixtures/quality-scenarios.json +20 -0
  779. package/plugins/harness-everything/skills/tdd/guides/deep-modules.md +33 -0
  780. package/plugins/harness-everything/skills/tdd/guides/interface-design.md +31 -0
  781. package/plugins/harness-everything/skills/tdd/guides/mocking.md +59 -0
  782. package/plugins/harness-everything/skills/tdd/guides/refactoring.md +10 -0
  783. package/plugins/harness-everything/skills/tdd/guides/tests.md +61 -0
  784. package/plugins/harness-everything/skills/tdd/references/core-discipline.md +51 -0
  785. package/plugins/harness-everything/skills/tdd/references/integration-testing.md +13 -0
  786. package/plugins/harness-everything/skills/tdd/references/quality-model.md +68 -0
  787. package/plugins/harness-everything/skills/tdd/references/unit-testing.md +15 -0
  788. package/plugins/harness-everything/skills/tdd/schemas/quality-evidence.schema.json +211 -0
  789. package/plugins/harness-everything/skills/tdd/schemas/quality-report.schema.json +110 -0
  790. package/plugins/harness-everything/skills/tdd/scripts/quality-gate.js +475 -0
  791. package/plugins/harness-everything/skills/to-spec/SKILL.md +40 -0
  792. package/plugins/harness-everything/skills/to-spec/references/process.md +77 -0
  793. package/plugins/harness-everything/skills/to-spec/scripts/check-project-docs.js +239 -0
  794. package/plugins/harness-everything/skills/to-spec/scripts/project-docs-resolver.js +111 -0
  795. package/plugins/harness-everything/skills/to-spec/templates/cli-reference.md +31 -0
  796. package/plugins/harness-everything/skills/to-spec/templates/dev-doc.md +27 -0
  797. package/plugins/harness-everything/skills/to-spec/templates/feature-spec.md +59 -0
  798. package/plugins/harness-everything/skills/to-spec/templates/schema-doc.md +31 -0
  799. package/plugins/harness-everything/skills/to-tickets/SKILL.md +42 -0
  800. package/plugins/harness-everything/skills/to-tickets/references/publishing-and-templates.md +97 -0
  801. package/plugins/harness-everything/skills/todo-driven-workflow/SKILL.md +42 -0
  802. package/plugins/harness-everything/skills/todo-driven-workflow/references/execution-guide.md +27 -0
  803. package/plugins/harness-everything/skills/using-git-worktrees/SKILL.md +49 -0
  804. package/plugins/harness-everything/skills/using-git-worktrees/references/workflow-details.md +108 -0
  805. package/plugins/harness-everything/skills/verification-loop/SKILL.md +46 -0
  806. package/plugins/harness-everything/skills/verification-loop/references/verification-phases.md +116 -0
  807. package/plugins/harness-everything/skills/verification-loop/templates/verification-report.template.md +15 -0
  808. package/plugins/harness-everything/skills/verify-before-claim/SKILL.md +43 -0
  809. package/plugins/harness-everything/skills/verify-before-claim/references/verification-guide.md +73 -0
  810. package/plugins/harness-everything/skills/zoom-out/SKILL.md +42 -0
  811. package/plugins/harness-everything/skills/zoom-out/references/circuit-breaker.md +66 -0
  812. package/plugins/harness-everything/skills/zoom-out/templates/zoom-out-report.template.md +20 -0
  813. package/references/definition-of-done.md +67 -0
  814. package/references/performance-checklist.md +153 -0
  815. package/references/platform-advisory-template.md +58 -0
  816. package/references/security-checklist.md +205 -0
  817. package/repo-docs/SKILL.md +34 -37
  818. package/repo-docs/references/process-guide.md +60 -0
  819. package/repo-docs/templates/agents-template.md +264 -264
  820. package/repo-docs/templates/knowledge-agents-template.md +50 -50
  821. package/repo-docs/templates/knowledge-readme-template.md +83 -83
  822. package/repo-docs/templates/manual-recon.md +117 -117
  823. package/repo-docs/templates/multi-skills-readme-template.md +72 -72
  824. package/repo-docs/templates/product-agents-template.md +78 -78
  825. package/repo-docs/templates/product-readme-template.md +95 -95
  826. package/repo-docs/templates/readme-template.md +136 -136
  827. package/repo-docs/templates/single-skill-readme-template.md +70 -70
  828. package/repo-docs/templates/skills-agents-template.md +60 -60
  829. package/rewrite-commits/SKILL.md +22 -17
  830. package/rewrite-commits/references/workflow.md +18 -0
  831. package/scripts/build-openai-submission.js +158 -0
  832. package/scripts/installer.js +621 -506
  833. package/scripts/lib/advisory-text.js +33 -22
  834. package/scripts/lib/claude-hooks.js +18 -5
  835. package/scripts/lib/execution-contract.js +70 -0
  836. package/scripts/lib/gitignore.js +199 -80
  837. package/scripts/lib/manifest.js +83 -23
  838. package/scripts/lib/path-boundary.js +90 -0
  839. package/scripts/lib/prompts.js +1 -1
  840. package/scripts/lib/skills.js +191 -51
  841. package/scripts/lib/workspace.js +269 -7
  842. package/scripts/sync-openai-plugin.js +91 -0
  843. package/scripts/sync-release-version.js +212 -0
  844. package/scripts/verify-install.js +147 -0
  845. package/security-review/SKILL.md +33 -497
  846. package/security-review/guides/OWASP-PATTERNS.md +31 -0
  847. package/security-review/guides/STRIDE-THREAT-MODEL.md +21 -0
  848. package/security-review/references/security-checklist.md +478 -0
  849. package/security-review/scripts/audit-secrets.js +85 -0
  850. package/self-evolve/SKILL.md +36 -45
  851. package/self-evolve/references/memory-resolution.md +61 -0
  852. package/self-evolve/scripts/persist-memory.js +272 -40
  853. package/self-evolve/scripts/register-dynamic-skill.js +253 -0
  854. package/self-evolve/scripts/self-regression.js +125 -113
  855. package/skill-creator/SKILL.md +33 -77
  856. package/skill-creator/references/authoring-workflow.md +14 -0
  857. package/skill-creator/references/dynamic-generation-contract.md +21 -0
  858. package/skill-creator/references/quality-checklist.md +16 -0
  859. package/skill-creator/references/quality-principles.md +81 -81
  860. package/skill-creator/references/testing-workflow.md +27 -32
  861. package/skill-style/SKILL.md +33 -39
  862. package/skill-style/references/style-guide.md +62 -0
  863. package/submission/openai/README.md +67 -0
  864. package/submission/openai/listing.json +31 -0
  865. package/submission/openai/test-cases.json +61 -0
  866. package/tdd/SKILL.md +30 -45
  867. package/tdd/fixtures/pass-evidence.json +100 -0
  868. package/tdd/fixtures/quality-scenarios.json +20 -0
  869. package/tdd/guides/deep-modules.md +33 -33
  870. package/tdd/guides/interface-design.md +31 -31
  871. package/tdd/guides/mocking.md +59 -59
  872. package/tdd/guides/refactoring.md +10 -10
  873. package/tdd/guides/tests.md +61 -61
  874. package/tdd/references/core-discipline.md +51 -0
  875. package/tdd/references/integration-testing.md +13 -0
  876. package/tdd/references/quality-model.md +68 -0
  877. package/tdd/references/unit-testing.md +15 -0
  878. package/tdd/schemas/quality-evidence.schema.json +211 -0
  879. package/tdd/schemas/quality-report.schema.json +110 -0
  880. package/tdd/scripts/quality-gate.js +475 -0
  881. package/to-spec/SKILL.md +40 -0
  882. package/to-spec/references/process.md +77 -0
  883. package/to-spec/scripts/check-project-docs.js +239 -0
  884. package/to-spec/scripts/project-docs-resolver.js +111 -0
  885. package/to-spec/templates/cli-reference.md +31 -0
  886. package/to-spec/templates/dev-doc.md +27 -0
  887. package/to-spec/templates/feature-spec.md +59 -0
  888. package/to-spec/templates/schema-doc.md +31 -0
  889. package/to-tickets/SKILL.md +42 -0
  890. package/to-tickets/references/publishing-and-templates.md +97 -0
  891. package/todo-driven-workflow/SKILL.md +28 -40
  892. package/todo-driven-workflow/references/execution-guide.md +27 -0
  893. package/using-git-worktrees/SKILL.md +29 -141
  894. package/using-git-worktrees/references/workflow-details.md +108 -0
  895. package/verification-loop/SKILL.md +29 -121
  896. package/verification-loop/references/verification-phases.md +116 -0
  897. package/verification-loop/templates/verification-report.template.md +15 -0
  898. package/verify-before-claim/SKILL.md +27 -70
  899. package/verify-before-claim/references/verification-guide.md +73 -0
  900. package/zoom-out/SKILL.md +28 -76
  901. package/zoom-out/references/circuit-breaker.md +66 -0
  902. package/zoom-out/templates/zoom-out-report.template.md +20 -0
  903. package/build-multi-agent-system/AGENTS.md +0 -50
  904. package/build-multi-agent-system/README.md +0 -63
  905. package/build-multi-agent-system/SKILL.md +0 -68
  906. package/create-agent-launcher/SKILL.md +0 -48
  907. package/create-agent-launcher/guidelines/platform-gemini.md +0 -101
  908. package/create-agent-launcher/workflows/01-init.md +0 -22
  909. package/create-agent-launcher/workflows/02-analysis.md +0 -26
  910. package/create-agent-launcher/workflows/03-generation.md +0 -29
  911. package/create-agent-launcher/workflows/04-launcher.md +0 -20
  912. package/docs/workflows/build-multi-agent-system.md +0 -62
  913. package/docs/workflows/create-agent-launcher.md +0 -62
  914. package/eval-framework/behavioral-test.js +0 -105
  915. package/eval-framework/cases/case2.json +0 -5
  916. package/eval-framework/mechanism-test.js +0 -251
  917. package/harness-everything/scripts/todo-cli.js +0 -157
  918. package/skill-style/STYLE.md +0 -37
package/docker-verify.sh CHANGED
@@ -1,266 +1,266 @@
1
- #!/bin/bash
2
- set -e # Exit immediately if a command exits with a non-zero status
3
-
4
- echo "================================================="
5
- echo " Harness OS - Comprehensive Docker Verification "
6
- echo "================================================="
7
-
8
- # 1. Prepare Git Environment (Subagent Scope Guard depends on Git)
9
- echo -e "\n[Step 1] Initializing local Git repository for tests..."
10
- if [ ! -d ".git" ]; then
11
- git init
12
- git config --global user.email "test@test.com"
13
- git config --global user.name "Test Bot"
14
- git add -A
15
- git commit -m "initial commit for test"
16
- echo "Git repository initialized and committed."
17
- else
18
- echo "Git repository already exists."
19
- fi
20
-
21
- # 2. Run Self-Regression Test Suite
22
- echo -e "\n[Step 2] Running Harness Self-Regression tests..."
23
- npm test
24
-
25
- # 3. Mechanism Checks (from VERIFICATION.md Section 2)
26
- echo -e "\n[Step 3] Running Mechanism Checks..."
27
-
28
- # 3a. Rule of 3 Circuit Breaker actually blocks
29
- # Runtime state lives under .claude/harness-state/sessions/<session_id>/;
30
- # invocations with no session_id (like these bare/manual calls) fall into a
31
- # fixed sessions/default/ bucket.
32
- echo -e "\n---> Running 3a: Rule of 3 Circuit Breaker"
33
- mkdir -p .claude/harness-state/sessions/default
34
- rm -f .claude/harness-state/sessions/default/zoom-out-report.md
35
- echo '{"count":3,"lastHash":"verify-test","zoomOutResolved":false}' > .claude/harness-state/sessions/default/rule-of-3-state.json
36
-
37
- # We expect this to exit with 2 and print CRITICAL trigger message
38
- set +e
39
- node hooks/scripts/rule-of-3.js 2>rule-of-3-err.log
40
- EXIT_CODE=$?
41
- set -e
42
-
43
- echo "Exit code: $EXIT_CODE"
44
- if [ "$EXIT_CODE" -ne 2 ]; then
45
- echo "❌ FAIL: Rule of 3 did not exit with code 2 (got $EXIT_CODE)"
46
- exit 1
47
- fi
48
-
49
- if ! grep -q "RULE OF 3 CIRCUIT BREAKER TRIGGERED" rule-of-3-err.log; then
50
- echo "❌ FAIL: Rule of 3 error output did not contain trigger message"
51
- cat rule-of-3-err.log
52
- exit 1
53
- fi
54
- echo "✅ PASS: Rule of 3 blocked successfully with exit code 2 and correct message."
55
-
56
- # Reset Rule of 3
57
- echo "Resetting circuit breaker..."
58
- npm run harness:reset
59
- set +e
60
- node hooks/scripts/rule-of-3.js
61
- EXIT_CODE=$?
62
- set -e
63
- if [ "$EXIT_CODE" -ne 0 ]; then
64
- echo "❌ FAIL: Rule of 3 reset check failed (expected 0, got $EXIT_CODE)"
65
- exit 1
66
- fi
67
- echo "✅ PASS: Rule of 3 reset check passed."
68
-
69
- # 3a-bis. Zoom-out reflection report releases the breaker (self-recovery path)
70
- echo -e "\n---> Running 3a-bis: Zoom-out report releases the breaker"
71
- echo '{"count":3,"lastHash":"verify-test","zoomOutResolved":false,"lastFailureAt":0,"zoomOutCycles":0}' > .claude/harness-state/sessions/default/rule-of-3-state.json
72
- printf '## Goal\nx\n## Failed Attempts\nx\n## Verified Facts\nx\n## Diagnosis\nx\n## Decision\nRESUME: new approach\n' > .claude/harness-state/sessions/default/zoom-out-report.md
73
-
74
- set +e
75
- node hooks/scripts/rule-of-3.js
76
- EXIT_CODE=$?
77
- set -e
78
- if [ "$EXIT_CODE" -ne 0 ]; then
79
- echo "❌ FAIL: Valid reflection report did not release the breaker (expected 0, got $EXIT_CODE)"
80
- exit 1
81
- fi
82
- if ! node -e "const s=require('./.claude/harness-state/sessions/default/rule-of-3-state.json'); process.exit(s.count===0 && s.zoomOutResolved===true && s.zoomOutCycles===1 ? 0 : 1)"; then
83
- echo "❌ FAIL: Breaker state not updated after report release"
84
- cat .claude/harness-state/sessions/default/rule-of-3-state.json
85
- exit 1
86
- fi
87
- echo "✅ PASS: Valid reflection report released the breaker (self-recovery)."
88
-
89
- # Second trip on the same signature must hard-lock even though a report exists
90
- # (the report is stale relative to lastFailureAt, and the cycle budget is spent).
91
- node -e "require('fs').writeFileSync('.claude/harness-state/sessions/default/rule-of-3-state.json', JSON.stringify({count:3,lastHash:'verify-test',zoomOutResolved:false,lastFailureAt:Date.now()+60000,zoomOutCycles:1}))"
92
- set +e
93
- node hooks/scripts/rule-of-3.js 2>rule-of-3-err.log
94
- EXIT_CODE=$?
95
- set -e
96
- if [ "$EXIT_CODE" -ne 2 ]; then
97
- echo "❌ FAIL: Second trip did not hard-lock (expected 2, got $EXIT_CODE)"
98
- cat rule-of-3-err.log
99
- exit 1
100
- fi
101
- if ! grep -q "hard lock" rule-of-3-err.log; then
102
- echo "❌ FAIL: Second-trip output did not contain hard-lock escalation message"
103
- cat rule-of-3-err.log
104
- exit 1
105
- fi
106
- echo "✅ PASS: Second trip on the same signature hard-locked for human decision."
107
- npm run harness:reset
108
-
109
-
110
- # 3b. Boundary Guard blocks oversized read
111
- echo -e "\n---> Running 3b: Boundary Guard Oversized Read"
112
- node -e "require('fs').writeFileSync('.verify-big.tmp','x'.repeat(600*1024))"
113
-
114
- set +e
115
- echo '{"tool_name":"Read","tool_input":{"file_path":".verify-big.tmp"}}' | node hooks/scripts/boundary-guard.js 2>boundary-guard-err.log
116
- EXIT_CODE=$?
117
- set -e
118
-
119
- rm -f .verify-big.tmp
120
-
121
- echo "Exit code: $EXIT_CODE"
122
- if [ "$EXIT_CODE" -ne 2 ]; then
123
- echo "❌ FAIL: Boundary Guard did not exit with code 2 (got $EXIT_CODE)"
124
- exit 1
125
- fi
126
-
127
- if ! grep -q "BLOCKED" boundary-guard-err.log; then
128
- echo "❌ FAIL: Boundary Guard error output did not contain BLOCKED"
129
- cat boundary-guard-err.log
130
- exit 1
131
- fi
132
- echo "✅ PASS: Boundary Guard blocked large read successfully."
133
-
134
-
135
- # 3c. State persistence (WAL) records failure and handles clear
136
- echo -e "\n---> Running 3c: State persistence (WAL)"
137
- rm -f .claude/harness-state/sessions/default/handoff-state.json
138
-
139
- # Record failure
140
- echo '{"tool_name":"Bash","tool_response":{"stdout":"","stderr":"npm ERR! verify-test failure"}}' | node hooks/scripts/state-persist.js
141
-
142
- if [ ! -f ".claude/harness-state/sessions/default/handoff-state.json" ]; then
143
- echo "❌ FAIL: handoff-state.json was not created"
144
- exit 1
145
- fi
146
-
147
- STATUS=$(node -e "console.log(JSON.parse(require('fs').readFileSync('.claude/harness-state/sessions/default/handoff-state.json','utf8')).status)")
148
- if [ "$STATUS" != "failed" ]; then
149
- echo "❌ FAIL: Expected status to be 'failed', got '$STATUS'"
150
- exit 1
151
- fi
152
-
153
- # Bootstrap output verification (pipe {} explicitly - bootstrap.js reads a
154
- # session_id off stdin same as every other hook, and a bare invocation's
155
- # stdin isn't guaranteed to close on its own in a non-interactive shell)
156
- BOOTSTRAP_OUT=$(echo '{}' | node harness-everything/scripts/bootstrap.js)
157
- if ! echo "$BOOTSTRAP_OUT" | grep -q "Harness OS - Handoff Checkpoint"; then
158
- echo "❌ FAIL: Bootstrap output did not show handoff checkpoint box"
159
- echo "$BOOTSTRAP_OUT"
160
- exit 1
161
- fi
162
- echo "✅ PASS: State persistence correctly saved failure status."
163
-
164
- # Clean run to clear WAL
165
- echo '{"tool_name":"Bash","tool_response":{"stdout":"ok","exitCode":0}}' | node hooks/scripts/state-persist.js
166
- BOOTSTRAP_OUT_CLEAN=$(echo '{}' | node harness-everything/scripts/bootstrap.js)
167
- if echo "$BOOTSTRAP_OUT_CLEAN" | grep -q "Harness OS - Handoff Checkpoint"; then
168
- echo "❌ FAIL: Bootstrap output still showed checkpoint box after clean run"
169
- exit 1
170
- fi
171
- echo "✅ PASS: State persistence correctly cleared status on successful execution."
172
-
173
-
174
- # 3d. Fact-audit reminder reaches agent via stdin JSON
175
- echo -e "\n---> Running 3d: Fact-audit reminder"
176
- ROUTER_OUT=$(echo '{"prompt":"what exit code does this hook use by default and is it documented"}' | node harness-everything/scripts/tier-router.js)
177
-
178
- if ! echo "$ROUTER_OUT" | grep -q "FACT-AUDIT REMINDER"; then
179
- echo "❌ FAIL: tier-router did not output FACT-AUDIT REMINDER on stdin JSON"
180
- echo "$ROUTER_OUT"
181
- exit 1
182
- fi
183
- echo "✅ PASS: Fact-audit reminder successfully outputted."
184
-
185
-
186
- # 3e. Subagent scope guard catches out-of-scope change
187
- echo -e "\n---> Running 3e: Subagent scope guard"
188
- # PreToolUse
189
- echo '{"tool_name":"Task","hook_event_name":"PreToolUse","tool_input":{}}' | node hooks/scripts/subagent-scope-guard.js
190
-
191
- # Make unexpected modification
192
- echo "unexpected change" >> .verify-scope-test.tmp
193
-
194
- # PostToolUse expect exit 2
195
- set +e
196
- echo '{"tool_name":"Task","hook_event_name":"PostToolUse","tool_input":{}}' | node hooks/scripts/subagent-scope-guard.js 2>scope-guard-err.log
197
- EXIT_CODE=$?
198
- set -e
199
-
200
- rm -f .verify-scope-test.tmp
201
-
202
- echo "Exit code: $EXIT_CODE"
203
- if [ "$EXIT_CODE" -ne 2 ]; then
204
- echo "❌ FAIL: Subagent scope guard did not block out-of-scope changes (expected exit code 2, got $EXIT_CODE)"
205
- exit 1
206
- fi
207
-
208
- if ! grep -q "Subagent Scope Guard" scope-guard-err.log || ! grep -q ".verify-scope-test.tmp" scope-guard-err.log; then
209
- echo "❌ FAIL: Scope guard did not report the unexpected changed file"
210
- cat scope-guard-err.log
211
- exit 1
212
- fi
213
- echo "✅ PASS: Subagent scope guard successfully blocked out-of-scope modifications."
214
-
215
-
216
- # 3f. Stop gate bounces an unverified-edit stop exactly once
217
- echo -e "\n---> Running 3f: Stop Gate"
218
- rm -f .claude/harness-state/sessions/default/stop-gate-state.json
219
- node -e "require('fs').mkdirSync('.claude/harness-state/sessions/default',{recursive:true}); require('fs').writeFileSync('.claude/harness-state/sessions/default/handoff-state.json', JSON.stringify({status:'idle',lastEditAt:Date.now(),lastVerifyAt:0}))"
220
- echo "dirty" > .verify-dirty.tmp
221
-
222
- set +e
223
- echo '{}' | node hooks/scripts/stop-gate.js 2>stop-gate-err.log
224
- EXIT_CODE=$?
225
- set -e
226
- echo "Exit code: $EXIT_CODE"
227
- if [ "$EXIT_CODE" -ne 2 ]; then
228
- echo "❌ FAIL: Stop gate did not bounce an unverified-edit stop (expected 2, got $EXIT_CODE)"
229
- cat stop-gate-err.log
230
- exit 1
231
- fi
232
- if ! grep -q "Stop Gate" stop-gate-err.log; then
233
- echo "❌ FAIL: Stop gate output did not contain its message"
234
- cat stop-gate-err.log
235
- exit 1
236
- fi
237
-
238
- # Same edit batch must not bounce twice
239
- set +e
240
- echo '{}' | node hooks/scripts/stop-gate.js
241
- EXIT_CODE=$?
242
- set -e
243
- if [ "$EXIT_CODE" -ne 0 ]; then
244
- echo "❌ FAIL: Stop gate bounced the same edit batch twice (expected 0, got $EXIT_CODE)"
245
- exit 1
246
- fi
247
-
248
- # Loop guard: stop_hook_active always passes
249
- rm -f .claude/harness-state/sessions/default/stop-gate-state.json
250
- set +e
251
- echo '{"stop_hook_active":true}' | node hooks/scripts/stop-gate.js
252
- EXIT_CODE=$?
253
- set -e
254
- if [ "$EXIT_CODE" -ne 0 ]; then
255
- echo "❌ FAIL: Stop gate blocked despite stop_hook_active (expected 0, got $EXIT_CODE)"
256
- exit 1
257
- fi
258
- rm -f .verify-dirty.tmp .claude/harness-state/sessions/default/handoff-state.json .claude/harness-state/sessions/default/stop-gate-state.json
259
- echo "✅ PASS: Stop gate bounced once, then respected the batch and loop guards."
260
-
261
- # Clean up temporary test output logs
262
- rm -f rule-of-3-err.log boundary-guard-err.log scope-guard-err.log stop-gate-err.log
263
-
264
- echo -e "\n================================================="
265
- echo " 🎉 ALL HARNESS MECHANISM CHECKS PASSED SUCCESSFULLY!"
266
- echo "================================================="
1
+ #!/bin/bash
2
+ set -e # Exit immediately if a command exits with a non-zero status
3
+
4
+ echo "================================================="
5
+ echo " Harness OS - Comprehensive Docker Verification "
6
+ echo "================================================="
7
+
8
+ # 1. Prepare Git Environment (Subagent Scope Guard depends on Git)
9
+ echo -e "\n[Step 1] Initializing local Git repository for tests..."
10
+ if [ ! -d ".git" ]; then
11
+ git init
12
+ git config --global user.email "test@test.com"
13
+ git config --global user.name "Test Bot"
14
+ git add -A
15
+ git commit -m "initial commit for test"
16
+ echo "Git repository initialized and committed."
17
+ else
18
+ echo "Git repository already exists."
19
+ fi
20
+
21
+ # 2. Run Self-Regression Test Suite
22
+ echo -e "\n[Step 2] Running Harness Self-Regression tests..."
23
+ npm test
24
+
25
+ # 3. Mechanism Checks (from VERIFICATION.md Section 2)
26
+ echo -e "\n[Step 3] Running Mechanism Checks..."
27
+
28
+ # 3a. Rule of 3 Circuit Breaker actually blocks
29
+ # Runtime state lives under .claude/harness-everything/state/sessions/<session_id>/;
30
+ # invocations with no session_id (like these bare/manual calls) fall into a
31
+ # fixed sessions/default/ bucket.
32
+ echo -e "\n---> Running 3a: Rule of 3 Circuit Breaker"
33
+ mkdir -p .claude/harness-everything/state/sessions/default
34
+ rm -f .claude/harness-everything/state/sessions/default/zoom-out-report.md
35
+ echo '{"count":3,"lastHash":"verify-test","zoomOutResolved":false}' > .claude/harness-everything/state/sessions/default/rule-of-3-state.json
36
+
37
+ # We expect this to exit with 2 and print CRITICAL trigger message
38
+ set +e
39
+ node hooks/scripts/rule-of-3.js 2>rule-of-3-err.log
40
+ EXIT_CODE=$?
41
+ set -e
42
+
43
+ echo "Exit code: $EXIT_CODE"
44
+ if [ "$EXIT_CODE" -ne 2 ]; then
45
+ echo "❌ FAIL: Rule of 3 did not exit with code 2 (got $EXIT_CODE)"
46
+ exit 1
47
+ fi
48
+
49
+ if ! grep -q "RULE OF 3 CIRCUIT BREAKER TRIGGERED" rule-of-3-err.log; then
50
+ echo "❌ FAIL: Rule of 3 error output did not contain trigger message"
51
+ cat rule-of-3-err.log
52
+ exit 1
53
+ fi
54
+ echo "✅ PASS: Rule of 3 blocked successfully with exit code 2 and correct message."
55
+
56
+ # Reset Rule of 3
57
+ echo "Resetting circuit breaker..."
58
+ npm run harness:reset
59
+ set +e
60
+ node hooks/scripts/rule-of-3.js
61
+ EXIT_CODE=$?
62
+ set -e
63
+ if [ "$EXIT_CODE" -ne 0 ]; then
64
+ echo "❌ FAIL: Rule of 3 reset check failed (expected 0, got $EXIT_CODE)"
65
+ exit 1
66
+ fi
67
+ echo "✅ PASS: Rule of 3 reset check passed."
68
+
69
+ # 3a-bis. Zoom-out reflection report releases the breaker (self-recovery path)
70
+ echo -e "\n---> Running 3a-bis: Zoom-out report releases the breaker"
71
+ echo '{"count":3,"lastHash":"verify-test","zoomOutResolved":false,"lastFailureAt":0,"zoomOutCycles":0}' > .claude/harness-everything/state/sessions/default/rule-of-3-state.json
72
+ printf '## Goal\nx\n## Failed Attempts\nx\n## Verified Facts\nx\n## Diagnosis\nx\n## Decision\nRESUME: new approach\n' > .claude/harness-everything/state/sessions/default/zoom-out-report.md
73
+
74
+ set +e
75
+ node hooks/scripts/rule-of-3.js
76
+ EXIT_CODE=$?
77
+ set -e
78
+ if [ "$EXIT_CODE" -ne 0 ]; then
79
+ echo "❌ FAIL: Valid reflection report did not release the breaker (expected 0, got $EXIT_CODE)"
80
+ exit 1
81
+ fi
82
+ if ! node -e "const s=require('./.claude/harness-everything/state/sessions/default/rule-of-3-state.json'); process.exit(s.count===0 && s.zoomOutResolved===true && s.zoomOutCycles===1 ? 0 : 1)"; then
83
+ echo "❌ FAIL: Breaker state not updated after report release"
84
+ cat .claude/harness-everything/state/sessions/default/rule-of-3-state.json
85
+ exit 1
86
+ fi
87
+ echo "✅ PASS: Valid reflection report released the breaker (self-recovery)."
88
+
89
+ # Second trip on the same signature must hard-lock even though a report exists
90
+ # (the report is stale relative to lastFailureAt, and the cycle budget is spent).
91
+ node -e "require('fs').writeFileSync('.claude/harness-everything/state/sessions/default/rule-of-3-state.json', JSON.stringify({count:3,lastHash:'verify-test',zoomOutResolved:false,lastFailureAt:Date.now()+60000,zoomOutCycles:1}))"
92
+ set +e
93
+ node hooks/scripts/rule-of-3.js 2>rule-of-3-err.log
94
+ EXIT_CODE=$?
95
+ set -e
96
+ if [ "$EXIT_CODE" -ne 2 ]; then
97
+ echo "❌ FAIL: Second trip did not hard-lock (expected 2, got $EXIT_CODE)"
98
+ cat rule-of-3-err.log
99
+ exit 1
100
+ fi
101
+ if ! grep -q "hard lock" rule-of-3-err.log; then
102
+ echo "❌ FAIL: Second-trip output did not contain hard-lock escalation message"
103
+ cat rule-of-3-err.log
104
+ exit 1
105
+ fi
106
+ echo "✅ PASS: Second trip on the same signature hard-locked for human decision."
107
+ npm run harness:reset
108
+
109
+
110
+ # 3b. Boundary Guard blocks oversized read
111
+ echo -e "\n---> Running 3b: Boundary Guard Oversized Read"
112
+ node -e "require('fs').writeFileSync('.verify-big.tmp','x'.repeat(600*1024))"
113
+
114
+ set +e
115
+ echo '{"tool_name":"Read","tool_input":{"file_path":".verify-big.tmp"}}' | node hooks/scripts/boundary-guard.js 2>boundary-guard-err.log
116
+ EXIT_CODE=$?
117
+ set -e
118
+
119
+ rm -f .verify-big.tmp
120
+
121
+ echo "Exit code: $EXIT_CODE"
122
+ if [ "$EXIT_CODE" -ne 2 ]; then
123
+ echo "❌ FAIL: Boundary Guard did not exit with code 2 (got $EXIT_CODE)"
124
+ exit 1
125
+ fi
126
+
127
+ if ! grep -q "BLOCKED" boundary-guard-err.log; then
128
+ echo "❌ FAIL: Boundary Guard error output did not contain BLOCKED"
129
+ cat boundary-guard-err.log
130
+ exit 1
131
+ fi
132
+ echo "✅ PASS: Boundary Guard blocked large read successfully."
133
+
134
+
135
+ # 3c. State persistence (WAL) records failure and handles clear
136
+ echo -e "\n---> Running 3c: State persistence (WAL)"
137
+ rm -f .claude/harness-everything/state/sessions/default/handoff-state.json
138
+
139
+ # Record failure
140
+ echo '{"tool_name":"Bash","tool_response":{"stdout":"","stderr":"npm ERR! verify-test failure"}}' | node hooks/scripts/state-persist.js
141
+
142
+ if [ ! -f ".claude/harness-everything/state/sessions/default/handoff-state.json" ]; then
143
+ echo "❌ FAIL: handoff-state.json was not created"
144
+ exit 1
145
+ fi
146
+
147
+ STATUS=$(node -e "console.log(JSON.parse(require('fs').readFileSync('.claude/harness-everything/state/sessions/default/handoff-state.json','utf8')).status)")
148
+ if [ "$STATUS" != "failed" ]; then
149
+ echo "❌ FAIL: Expected status to be 'failed', got '$STATUS'"
150
+ exit 1
151
+ fi
152
+
153
+ # Bootstrap output verification (pipe {} explicitly - bootstrap.js reads a
154
+ # session_id off stdin same as every other hook, and a bare invocation's
155
+ # stdin isn't guaranteed to close on its own in a non-interactive shell)
156
+ BOOTSTRAP_OUT=$(echo '{}' | node harness-everything/scripts/bootstrap.js)
157
+ if ! echo "$BOOTSTRAP_OUT" | grep -q "Harness OS - Handoff Checkpoint"; then
158
+ echo "❌ FAIL: Bootstrap output did not show handoff checkpoint box"
159
+ echo "$BOOTSTRAP_OUT"
160
+ exit 1
161
+ fi
162
+ echo "✅ PASS: State persistence correctly saved failure status."
163
+
164
+ # Clean run to clear WAL
165
+ echo '{"tool_name":"Bash","tool_response":{"stdout":"ok","exitCode":0}}' | node hooks/scripts/state-persist.js
166
+ BOOTSTRAP_OUT_CLEAN=$(echo '{}' | node harness-everything/scripts/bootstrap.js)
167
+ if echo "$BOOTSTRAP_OUT_CLEAN" | grep -q "Harness OS - Handoff Checkpoint"; then
168
+ echo "❌ FAIL: Bootstrap output still showed checkpoint box after clean run"
169
+ exit 1
170
+ fi
171
+ echo "✅ PASS: State persistence correctly cleared status on successful execution."
172
+
173
+
174
+ # 3d. Fact-audit reminder reaches agent via stdin JSON
175
+ echo -e "\n---> Running 3d: Fact-audit reminder"
176
+ ROUTER_OUT=$(echo '{"prompt":"what exit code does this hook use by default and is it documented"}' | node harness-everything/scripts/tier-router.js)
177
+
178
+ if ! echo "$ROUTER_OUT" | grep -q "FACT-AUDIT REMINDER"; then
179
+ echo "❌ FAIL: tier-router did not output FACT-AUDIT REMINDER on stdin JSON"
180
+ echo "$ROUTER_OUT"
181
+ exit 1
182
+ fi
183
+ echo "✅ PASS: Fact-audit reminder successfully outputted."
184
+
185
+
186
+ # 3e. Subagent scope guard catches out-of-scope change
187
+ echo -e "\n---> Running 3e: Subagent scope guard"
188
+ # PreToolUse
189
+ echo '{"tool_name":"Task","hook_event_name":"PreToolUse","tool_input":{}}' | node hooks/scripts/subagent-scope-guard.js
190
+
191
+ # Make unexpected modification
192
+ echo "unexpected change" >> .verify-scope-test.tmp
193
+
194
+ # PostToolUse expect exit 2
195
+ set +e
196
+ echo '{"tool_name":"Task","hook_event_name":"PostToolUse","tool_input":{}}' | node hooks/scripts/subagent-scope-guard.js 2>scope-guard-err.log
197
+ EXIT_CODE=$?
198
+ set -e
199
+
200
+ rm -f .verify-scope-test.tmp
201
+
202
+ echo "Exit code: $EXIT_CODE"
203
+ if [ "$EXIT_CODE" -ne 2 ]; then
204
+ echo "❌ FAIL: Subagent scope guard did not block out-of-scope changes (expected exit code 2, got $EXIT_CODE)"
205
+ exit 1
206
+ fi
207
+
208
+ if ! grep -q "Subagent Scope Guard" scope-guard-err.log || ! grep -q ".verify-scope-test.tmp" scope-guard-err.log; then
209
+ echo "❌ FAIL: Scope guard did not report the unexpected changed file"
210
+ cat scope-guard-err.log
211
+ exit 1
212
+ fi
213
+ echo "✅ PASS: Subagent scope guard successfully blocked out-of-scope modifications."
214
+
215
+
216
+ # 3f. Stop gate bounces an unverified-edit stop exactly once
217
+ echo -e "\n---> Running 3f: Stop Gate"
218
+ rm -f .claude/harness-everything/state/sessions/default/stop-gate-state.json
219
+ node -e "require('fs').mkdirSync('.claude/harness-everything/state/sessions/default',{recursive:true}); require('fs').writeFileSync('.claude/harness-everything/state/sessions/default/handoff-state.json', JSON.stringify({status:'idle',lastEditAt:Date.now(),lastVerifyAt:0}))"
220
+ echo "dirty" > .verify-dirty.tmp
221
+
222
+ set +e
223
+ echo '{}' | node hooks/scripts/stop-gate.js 2>stop-gate-err.log
224
+ EXIT_CODE=$?
225
+ set -e
226
+ echo "Exit code: $EXIT_CODE"
227
+ if [ "$EXIT_CODE" -ne 2 ]; then
228
+ echo "❌ FAIL: Stop gate did not bounce an unverified-edit stop (expected 2, got $EXIT_CODE)"
229
+ cat stop-gate-err.log
230
+ exit 1
231
+ fi
232
+ if ! grep -q "Stop Gate" stop-gate-err.log; then
233
+ echo "❌ FAIL: Stop gate output did not contain its message"
234
+ cat stop-gate-err.log
235
+ exit 1
236
+ fi
237
+
238
+ # Same edit batch must not bounce twice
239
+ set +e
240
+ echo '{}' | node hooks/scripts/stop-gate.js
241
+ EXIT_CODE=$?
242
+ set -e
243
+ if [ "$EXIT_CODE" -ne 0 ]; then
244
+ echo "❌ FAIL: Stop gate bounced the same edit batch twice (expected 0, got $EXIT_CODE)"
245
+ exit 1
246
+ fi
247
+
248
+ # Loop guard: stop_hook_active always passes
249
+ rm -f .claude/harness-everything/state/sessions/default/stop-gate-state.json
250
+ set +e
251
+ echo '{"stop_hook_active":true}' | node hooks/scripts/stop-gate.js
252
+ EXIT_CODE=$?
253
+ set -e
254
+ if [ "$EXIT_CODE" -ne 0 ]; then
255
+ echo "❌ FAIL: Stop gate blocked despite stop_hook_active (expected 0, got $EXIT_CODE)"
256
+ exit 1
257
+ fi
258
+ rm -f .verify-dirty.tmp .claude/harness-everything/state/sessions/default/handoff-state.json .claude/harness-everything/state/sessions/default/stop-gate-state.json
259
+ echo "✅ PASS: Stop gate bounced once, then respected the batch and loop guards."
260
+
261
+ # Clean up temporary test output logs
262
+ rm -f rule-of-3-err.log boundary-guard-err.log scope-guard-err.log stop-gate-err.log
263
+
264
+ echo -e "\n================================================="
265
+ echo " 🎉 ALL HARNESS MECHANISM CHECKS PASSED SUCCESSFULLY!"
266
+ echo "================================================="