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
@@ -0,0 +1,71 @@
1
+ #!/usr/bin/env node
2
+ /**
3
+ * Harness Mechanism Test Suite Runner (Claude Code hook contract)
4
+ *
5
+ * Dynamically discovers and executes all isolated mechanism test files (matching `mechanism-*.test.js`
6
+ * under `ci/`). This allows clean tracking of which mechanism tests are implemented,
7
+ * keeps the test code highly decoupled, and simplifies debugging and adding new test suites.
8
+ */
9
+
10
+ const { spawnSync } = require('child_process');
11
+ const fs = require('fs');
12
+ const path = require('path');
13
+
14
+ const projectRoot = path.resolve(__dirname, '..');
15
+ const evalDir = path.join(projectRoot, 'ci');
16
+
17
+ console.log('=================================================');
18
+ console.log('=== 🔧 Harness Mechanism Test Suite (VERIFICATION.md §2) ===');
19
+ console.log('=================================================');
20
+
21
+ // 1. Discover all test files matching mechanism-*.test.js
22
+ const files = fs.readdirSync(evalDir)
23
+ .filter(file => file.startsWith('mechanism-') && file.endsWith('.test.js'))
24
+ .sort(); // Run them in sorting order (2a, 2b, 2c, etc.)
25
+
26
+ console.log(`\nDiscovered ${files.length} mechanism test suite(s):`);
27
+ files.forEach(f => console.log(` - ${f}`));
28
+ console.log('\n-------------------------------------------------');
29
+
30
+ let failuresCount = 0;
31
+ const results = [];
32
+
33
+ // 2. Execute each test file sequentially
34
+ files.forEach(file => {
35
+ const filePath = path.join(evalDir, file);
36
+ console.log(`\n▶️ Running test suite: ${file}`);
37
+
38
+ const testRun = spawnSync('node', [filePath], {
39
+ stdio: 'inherit',
40
+ cwd: projectRoot,
41
+ env: { ...process.env, CLAUDE: '1' }
42
+ });
43
+
44
+ if (testRun.status !== 0) {
45
+ console.error(`❌ Test suite failed: ${file}`);
46
+ failuresCount++;
47
+ results.push({ name: file, status: 'FAILED' });
48
+ } else {
49
+ results.push({ name: file, status: 'PASSED' });
50
+ }
51
+ });
52
+
53
+ // 3. Summarize results
54
+ console.log('\n=================================================');
55
+ console.log(' Mechanism Test Summary ');
56
+ console.log('=================================================');
57
+ results.forEach(res => {
58
+ const icon = res.status === 'PASSED' ? '✅' : '❌';
59
+ console.log(` ${icon} ${res.name.padEnd(45)} [${res.status}]`);
60
+ });
61
+ console.log('-------------------------------------------------');
62
+
63
+ if (failuresCount > 0) {
64
+ console.error(` ❌ ${failuresCount} mechanism test suite(s) FAILED.`);
65
+ console.log('=================================================');
66
+ process.exit(1);
67
+ } else {
68
+ console.log(' 🎉 All VERIFICATION.md §2 mechanism checks passed.');
69
+ console.log('=================================================');
70
+ process.exit(0);
71
+ }
@@ -0,0 +1,53 @@
1
+ #!/usr/bin/env node
2
+ 'use strict';
3
+
4
+ const path = require('path');
5
+ const { spawnSync } = require('child_process');
6
+
7
+ const ROOT = path.resolve(__dirname, '..');
8
+ const loader = path.join(__dirname, 'mutation-loader.js');
9
+ const checks = [
10
+ ['physical path boundaries', 'scripts/lib/path-boundary.js', 'ci/mechanism-2v-verifier-boundaries.test.js'],
11
+ ['exact execution commands', 'scripts/lib/execution-contract.js', 'ci/mechanism-2v-verifier-boundaries.test.js'],
12
+ ['archive output boundaries', 'behavioral-evals/evidence-tool.js', 'ci/mechanism-2t-pr66-evidence-triage.test.js'],
13
+ ['legacy migration boundaries', 'multi-agent-workspace/scripts/scaffold.js', 'ci/mechanism-2l-multi-agent-workspace.test.js'],
14
+ ['global installer fallback', 'scripts/installer.js', 'ci/mechanism-2q-installer-global-shared-store.test.js'],
15
+ ['transcript command qualifiers', 'behavioral-evals/transcript-parser.js', 'ci/mechanism-2t-transcript-parser.test.js'],
16
+ ['fixture path boundaries', 'behavioral-evals/run.js', 'ci/mechanism-2j-behavioral-runner.test.js'],
17
+ ['unbound workspace identity', 'scripts/lib/workspace.js', 'ci/mechanism-2u-issue-42-regressions.test.js'],
18
+ ];
19
+
20
+ let failures = 0;
21
+ for (const [label, target, testFile] of checks) {
22
+ const result = spawnSync(process.execPath, ['--require', loader, path.join(ROOT, testFile)], {
23
+ cwd: ROOT,
24
+ env: {
25
+ ...process.env,
26
+ HARNESS_MUTATION: {
27
+ 'scripts/lib/path-boundary.js': 'path-boundary',
28
+ 'scripts/lib/execution-contract.js': 'execution-contract',
29
+ 'behavioral-evals/evidence-tool.js': 'evidence-tool',
30
+ 'multi-agent-workspace/scripts/scaffold.js': 'scaffold-migration',
31
+ 'scripts/installer.js': 'installer-global',
32
+ 'behavioral-evals/transcript-parser.js': 'transcript-command',
33
+ 'behavioral-evals/run.js': 'runner-fixture',
34
+ 'scripts/lib/workspace.js': 'workspace-identity',
35
+ }[target],
36
+ HARNESS_MUTATION_TARGET: path.join(ROOT, target),
37
+ NODE_OPTIONS: `--require=${loader}`,
38
+ },
39
+ encoding: 'utf8'
40
+ });
41
+ if (result.status === 0) {
42
+ console.error(`FAIL mutation survived: ${label} (${target})`);
43
+ failures++;
44
+ } else {
45
+ console.log(`PASS mutation was caught: ${label} (${testFile})`);
46
+ }
47
+ }
48
+
49
+ if (failures) {
50
+ console.error(`Mutation check failed: ${failures} mutation(s) survived.`);
51
+ process.exit(1);
52
+ }
53
+ console.log(`Mutation check passed: ${checks.length} production mutations were rejected by focused tests.`);
@@ -0,0 +1,57 @@
1
+ 'use strict';
2
+
3
+ const fs = require('fs');
4
+ const Module = require('module');
5
+ const path = require('path');
6
+
7
+ const target = process.env.HARNESS_MUTATION_TARGET && path.resolve(process.env.HARNESS_MUTATION_TARGET);
8
+ const mutation = process.env.HARNESS_MUTATION;
9
+ const originalLoader = Module._extensions['.js'];
10
+
11
+ const mutations = {
12
+ 'path-boundary': (source) => source.replace(
13
+ ' assertNoLinkComponents(root, candidate);',
14
+ ' // mutation: skip physical link-component validation'
15
+ ),
16
+ 'execution-contract': (source) => source.replace(
17
+ ' return left !== null && left === right;',
18
+ ' return left !== null; // mutation: accept every non-null command'
19
+ ),
20
+ 'evidence-tool': (source) => source
21
+ .replace(' const containedDestination = assertContainedPath(archiveRoot, destination).target;', ' const containedDestination = destination; // mutation: skip destination boundary')
22
+ .replace(' const contained = assertContainedPath(archiveRoot, target).target;', ' const contained = target; // mutation: skip archive write boundary'),
23
+ 'scaffold-migration': (source) => source.replace(
24
+ ' assertContainedPath(workspaceRoot, legacyRoot);',
25
+ ' // mutation: skip linked legacy-root validation'
26
+ ),
27
+ 'installer-global': (source) => source.replace(
28
+ 'const workspaceRoot = detectedWorkspaceRoot || path.resolve(process.cwd());',
29
+ 'const workspaceRoot = detectedWorkspaceRoot; // mutation: restore null global workspace'
30
+ ),
31
+ 'transcript-command': (source) => source.replace(
32
+ ' if (target.command !== undefined && (!isObject(call.input) || !commandMatches(call.input.command, target.command))) return false;',
33
+ ' // mutation: ignore command qualifiers'
34
+ ),
35
+ 'runner-fixture': (source) => source.replace(
36
+ ' const relative = normalizeRepoPath(file && file.path);',
37
+ ' const relative = file && file.path; // mutation: skip fixture path normalization'
38
+ ),
39
+ 'workspace-identity': (source) => source.replace(
40
+ ' if (!workspaceRoot) return UNBOUND_WORKSPACE_KEY;',
41
+ ' if (!workspaceRoot) return path.basename(process.cwd()); // mutation: derive unbound identity from cwd'
42
+ ),
43
+ };
44
+
45
+ if (target && mutation && mutations[mutation]) {
46
+ const transform = mutations[mutation];
47
+ Module._extensions['.js'] = function loadWithMutation(mod, filename) {
48
+ const source = fs.readFileSync(filename, 'utf8');
49
+ if (path.resolve(filename) === target) {
50
+ const mutated = transform(source);
51
+ if (mutated === source) throw new Error(`mutation was not applied to ${filename}`);
52
+ mod._compile(mutated, filename);
53
+ return;
54
+ }
55
+ originalLoader(mod, filename);
56
+ };
57
+ }
@@ -0,0 +1,114 @@
1
+ #!/usr/bin/env node
2
+ /**
3
+ * Negative control validator for quality gates.
4
+ * Tests that known-bad fixtures are correctly rejected by each gate.
5
+ *
6
+ * Usage: node ci/negative-controls.js
7
+ */
8
+ const fs = require('fs');
9
+ const path = require('path');
10
+ const { spawnSync } = require('child_process');
11
+
12
+ const ROOT = path.resolve(__dirname, '..');
13
+ let passed = 0;
14
+ let failed = 0;
15
+ let fixtureCounter = 0;
16
+
17
+ function test(name, fn) {
18
+ try {
19
+ fn();
20
+ console.log(`✅ ${name}`);
21
+ passed++;
22
+ } catch (e) {
23
+ console.error(`❌ ${name}`);
24
+ console.error(` ${e.message}`);
25
+ failed++;
26
+ }
27
+ }
28
+
29
+ function assert(condition, msg) {
30
+ if (!condition) throw new Error(msg || 'assertion failed');
31
+ }
32
+
33
+ function runGate(script) {
34
+ const result = spawnSync(process.execPath, [path.join(ROOT, 'ci', script)], {
35
+ cwd: ROOT,
36
+ encoding: 'utf8',
37
+ env: process.env,
38
+ });
39
+ return {
40
+ ...result,
41
+ output: `${result.stdout || ''}\n${result.stderr || ''}`,
42
+ };
43
+ }
44
+
45
+ function installFixture(source, label) {
46
+ const target = path.join(ROOT, `issue20-negative-${process.pid}-${++fixtureCounter}-${label}`);
47
+ assert(!fs.existsSync(target), `refusing to overwrite existing path: ${target}`);
48
+ fs.cpSync(path.join(ROOT, source), target, { recursive: true });
49
+ return target;
50
+ }
51
+
52
+ function removeFixtures(targets) {
53
+ for (const target of targets.reverse()) {
54
+ fs.rmSync(target, { recursive: true, force: true });
55
+ }
56
+ }
57
+
58
+ function expectGateFailure(name, script, fixtures, matcher) {
59
+ test(name, () => {
60
+ const targets = [];
61
+ try {
62
+ for (const fixture of fixtures) targets.push(installFixture(fixture.source, fixture.label));
63
+ const result = runGate(script);
64
+ assert(result.status !== 0, `${script} unexpectedly passed (exit ${result.status})`);
65
+ assert(matcher.test(result.output), `${script} failed without the expected diagnostic:\n${result.output}`);
66
+ console.log(` real gate: ${script} exited ${result.status}`);
67
+ } finally {
68
+ removeFixtures(targets);
69
+ }
70
+ });
71
+ }
72
+
73
+ expectGateFailure(
74
+ 'consistency-check rejects malformed frontmatter',
75
+ 'consistency-check.js',
76
+ [{ source: 'ci/fixtures/bad-frontmatter-colon', label: 'bad-frontmatter' }],
77
+ /frontmatter YAML syntax valid|YAMLException|unquoted colon/i
78
+ );
79
+
80
+ expectGateFailure(
81
+ 'consistency-check rejects a missing frontmatter delimiter',
82
+ 'consistency-check.js',
83
+ [{ source: 'eval-framework/fixtures/missing-delimiter', label: 'missing-delimiter' }],
84
+ /delimiter|frontmatter/i
85
+ );
86
+
87
+ expectGateFailure(
88
+ 'description-collision rejects duplicate descriptions',
89
+ 'description-collision.js',
90
+ [
91
+ { source: 'ci/fixtures/duplicate-description/skill-a', label: 'duplicate-a' },
92
+ { source: 'ci/fixtures/duplicate-description/skill-b', label: 'duplicate-b' },
93
+ ],
94
+ /COLLISION/i
95
+ );
96
+
97
+ expectGateFailure(
98
+ 'consistency-check rejects an over-budget skill',
99
+ 'consistency-check.js',
100
+ [{ source: 'eval-framework/fixtures/over-budget-skill', label: 'over-budget' }],
101
+ /exceeds hard limit|word count/i
102
+ );
103
+
104
+ test('normal repository passes the exercised gates', () => {
105
+ const consistency = runGate('consistency-check.js');
106
+ const collision = runGate('description-collision.js');
107
+ assert(consistency.status === 0, `consistency-check failed (exit ${consistency.status})`);
108
+ assert(collision.status === 0, `description-collision failed (exit ${collision.status})`);
109
+ console.log(' real gates: consistency-check.js and description-collision.js exited 0');
110
+ });
111
+
112
+ console.log(`\nNegative control results: ${passed} passed, ${failed} failed`);
113
+ if (failed > 0) process.exit(1);
114
+ console.log('All negative controls verified by executing the real gates.');
@@ -0,0 +1,182 @@
1
+ 'use strict';
2
+
3
+ const assert = require('assert');
4
+ const crypto = require('crypto');
5
+ const fs = require('fs');
6
+ const path = require('path');
7
+ const { spawnSync } = require('child_process');
8
+
9
+ const ROOT = path.resolve(__dirname, '..');
10
+ const PLUGIN = path.join(ROOT, 'plugins', 'harness-everything');
11
+ const SOURCE_MANIFEST = JSON.parse(fs.readFileSync(path.join(ROOT, '.claude-plugin', 'plugin.json'), 'utf8'));
12
+ const PLUGIN_MANIFEST = JSON.parse(fs.readFileSync(path.join(PLUGIN, '.codex-plugin', 'plugin.json'), 'utf8'));
13
+ const PORTABLE_MANIFEST = JSON.parse(fs.readFileSync(path.join(PLUGIN, 'plugin.json'), 'utf8'));
14
+
15
+ function sha256(file) {
16
+ const bytes = fs.readFileSync(file);
17
+ // Git stores text blobs with LF, while a Windows checkout may materialize
18
+ // canonical files with CRLF. Package parity compares canonical content, not
19
+ // platform checkout EOL bytes.
20
+ const normalized = bytes.includes(0)
21
+ ? bytes
22
+ : Buffer.from(bytes.toString('utf8').replace(/\r\n/g, '\n'), 'utf8');
23
+ return crypto.createHash('sha256').update(normalized).digest('hex');
24
+ }
25
+
26
+ function filesUnder(root, base = root, out = []) {
27
+ for (const entry of fs.readdirSync(root, { withFileTypes: true })) {
28
+ const full = path.join(root, entry.name);
29
+ if (entry.isDirectory()) filesUnder(full, base, out);
30
+ else if (entry.isFile()) out.push(path.relative(base, full).replace(/\\/g, '/'));
31
+ }
32
+ return out.sort();
33
+ }
34
+
35
+ function assertTreesEqual(source, packaged, name) {
36
+ assert.ok(fs.existsSync(packaged), `${name}: packaged skill directory missing`);
37
+ const sourceFiles = filesUnder(source);
38
+ const packagedFiles = filesUnder(packaged);
39
+ assert.deepStrictEqual(packagedFiles, sourceFiles, `${name}: packaged file list drifted from canonical source`);
40
+ for (const relative of sourceFiles) {
41
+ assert.strictEqual(
42
+ sha256(path.join(packaged, relative)),
43
+ sha256(path.join(source, relative)),
44
+ `${name}/${relative}: packaged content drifted from canonical source`
45
+ );
46
+ }
47
+ }
48
+
49
+ assert.strictEqual(PLUGIN_MANIFEST.name, 'harness-everything');
50
+ assert.strictEqual(PLUGIN_MANIFEST.version, SOURCE_MANIFEST.version, 'OpenAI plugin version must match canonical plugin version');
51
+ assert.strictEqual(PLUGIN_MANIFEST.skills, './skills/');
52
+ assert.strictEqual(PLUGIN_MANIFEST.hooks, './hooks/hooks.json');
53
+ assert.strictEqual(PORTABLE_MANIFEST.$schema, 'https://agent-plugins.org/schemas/1.0.0/plugin.schema.json');
54
+ assert.strictEqual(PORTABLE_MANIFEST.name, PLUGIN_MANIFEST.name);
55
+ assert.strictEqual(PORTABLE_MANIFEST.version, PLUGIN_MANIFEST.version);
56
+ assert.strictEqual(PORTABLE_MANIFEST.description, PLUGIN_MANIFEST.description);
57
+ assert.deepStrictEqual(PORTABLE_MANIFEST.author, PLUGIN_MANIFEST.author);
58
+ assert.strictEqual(PORTABLE_MANIFEST.homepage, PLUGIN_MANIFEST.homepage);
59
+ assert.strictEqual(PORTABLE_MANIFEST.repository, PLUGIN_MANIFEST.repository);
60
+ assert.strictEqual(PORTABLE_MANIFEST.license, PLUGIN_MANIFEST.license);
61
+ assert.deepStrictEqual(PORTABLE_MANIFEST.keywords, PLUGIN_MANIFEST.keywords);
62
+ assert.ok(!Object.prototype.hasOwnProperty.call(PORTABLE_MANIFEST, 'skills'), 'portable plugin manifest must use the root skills/ convention');
63
+ const portableOpenAI = PORTABLE_MANIFEST.extensions?.['com.openai'];
64
+ assert.ok(portableOpenAI, 'portable plugin manifest must expose the OpenAI extension');
65
+ assert.strictEqual(portableOpenAI.hooks, PLUGIN_MANIFEST.hooks);
66
+ assert.deepStrictEqual(portableOpenAI.interface, PLUGIN_MANIFEST.interface, 'portable and compatibility interface metadata drifted');
67
+ assert.ok(!Object.prototype.hasOwnProperty.call(PLUGIN_MANIFEST, 'agents'), 'OpenAI manifest must not claim an undocumented agents field');
68
+ assert.ok(!Object.prototype.hasOwnProperty.call(PLUGIN_MANIFEST, 'mcpServers'), 'skill-only package must not add MCP without a demonstrated requirement');
69
+ assert.ok(!Object.prototype.hasOwnProperty.call(PLUGIN_MANIFEST, 'apps'), 'skill-only package must not claim a connected app');
70
+
71
+ const requiredInterfaceFields = [
72
+ 'displayName',
73
+ 'shortDescription',
74
+ 'longDescription',
75
+ 'developerName',
76
+ 'category',
77
+ 'capabilities',
78
+ 'websiteURL',
79
+ 'privacyPolicyURL',
80
+ 'termsOfServiceURL',
81
+ 'defaultPrompt'
82
+ ];
83
+ for (const field of requiredInterfaceFields) {
84
+ assert.ok(Object.prototype.hasOwnProperty.call(PLUGIN_MANIFEST.interface || {}, field), `OpenAI interface missing ${field}`);
85
+ }
86
+ assert.strictEqual(PLUGIN_MANIFEST.interface.displayName, 'Harness Everything');
87
+ assert.strictEqual(PLUGIN_MANIFEST.interface.category, 'Developer Tools');
88
+ assert.deepStrictEqual(PLUGIN_MANIFEST.interface.capabilities, ['Interactive', 'Read', 'Write']);
89
+ assert.ok(Array.isArray(PLUGIN_MANIFEST.interface.defaultPrompt), 'defaultPrompt must be an array');
90
+ assert.ok(PLUGIN_MANIFEST.interface.defaultPrompt.length > 0 && PLUGIN_MANIFEST.interface.defaultPrompt.length <= 3, 'defaultPrompt must contain 1-3 entries');
91
+ for (const prompt of PLUGIN_MANIFEST.interface.defaultPrompt) {
92
+ assert.ok(prompt.length <= 128, `defaultPrompt exceeds 128 characters: ${prompt}`);
93
+ }
94
+
95
+ const publicationDocs = [
96
+ ['privacyPolicyURL', 'PRIVACY.md'],
97
+ ['termsOfServiceURL', 'TERMS.md']
98
+ ];
99
+ for (const [field, filename] of publicationDocs) {
100
+ const value = PLUGIN_MANIFEST.interface[field];
101
+ assert.match(value, /^https:\/\/github\.com\/dyphn1\/Harness-everything\/blob\/main\//, `${field} must use a public HTTPS repository URL`);
102
+ assert.ok(fs.existsSync(path.join(ROOT, filename)), `${field} target document is missing: ${filename}`);
103
+ }
104
+
105
+ const skillNames = (SOURCE_MANIFEST.skills || []).map(entry => path.basename(entry)).sort();
106
+ const packagedSkillNames = fs.readdirSync(path.join(PLUGIN, 'skills'), { withFileTypes: true })
107
+ .filter(entry => entry.isDirectory())
108
+ .map(entry => entry.name)
109
+ .sort();
110
+ assert.strictEqual(skillNames.length, 26, 'canonical suite should contain 26 skills');
111
+ assert.deepStrictEqual(packagedSkillNames, skillNames, 'OpenAI package must contain exactly the canonical skill set');
112
+ for (const entry of SOURCE_MANIFEST.skills) {
113
+ const source = path.resolve(ROOT, entry);
114
+ const name = path.basename(source);
115
+ assertTreesEqual(source, path.join(PLUGIN, 'skills', name), name);
116
+ }
117
+
118
+ const marketplace = JSON.parse(fs.readFileSync(path.join(ROOT, '.agents', 'plugins', 'marketplace.json'), 'utf8'));
119
+ assert.strictEqual(marketplace.name, 'harness-everything', 'marketplace should use the publication name, not a local-only label');
120
+ assert.strictEqual(marketplace.interface?.displayName, 'Harness Everything');
121
+ const marketEntry = marketplace.plugins.find(plugin => plugin.name === 'harness-everything');
122
+ assert.ok(marketEntry, 'repo marketplace must expose harness-everything');
123
+ assert.strictEqual(marketEntry.source.source, 'local');
124
+ assert.strictEqual(marketEntry.source.path, './plugins/harness-everything');
125
+ assert.strictEqual(marketEntry.policy?.installation, 'AVAILABLE');
126
+ assert.strictEqual(marketEntry.policy?.authentication, 'ON_INSTALL');
127
+ assert.strictEqual(marketEntry.category, 'Developer Tools');
128
+ assert.ok(fs.existsSync(path.join(ROOT, marketEntry.source.path, '.codex-plugin', 'plugin.json')), 'marketplace source must resolve to an installable plugin');
129
+
130
+ const hooks = JSON.parse(fs.readFileSync(path.join(PLUGIN, 'hooks', 'hooks.json'), 'utf8'));
131
+ assert.ok(hooks.hooks.SessionStart?.length, 'SessionStart hook missing');
132
+ assert.ok(hooks.hooks.UserPromptSubmit?.length, 'UserPromptSubmit hook missing');
133
+ assert.ok(hooks.hooks.PreToolUse?.length, 'PreToolUse hook missing');
134
+ assert.ok(hooks.hooks.PostToolUse?.length, 'PostToolUse hook missing');
135
+ assert.ok(hooks.hooks.SubagentStart?.length, 'SubagentStart hook missing');
136
+ assert.ok(hooks.hooks.SubagentStop?.length, 'SubagentStop hook missing');
137
+ assert.ok(hooks.hooks.Stop?.length, 'Stop hook missing');
138
+ const hookText = JSON.stringify(hooks);
139
+ assert.match(hookText, /PLUGIN_ROOT/, 'plugin hooks must resolve from PLUGIN_ROOT');
140
+ assert.match(hookText, /commandWindows/, 'plugin hooks need a Windows command override');
141
+
142
+ const session = spawnSync(process.execPath, [path.join(PLUGIN, 'hooks', 'session-start.js')], { encoding: 'utf8' });
143
+ assert.strictEqual(session.status, 0, session.stderr);
144
+ assert.match(session.stdout, /Cognitive OS policy is active/);
145
+ assert.match(session.stdout, /Route before execution/);
146
+ assert.match(session.stdout, /Verify before claim/);
147
+ assert.match(session.stdout, /3 same-signature failures/);
148
+
149
+ const kernel = path.join(PLUGIN, 'skills', 'harness-everything', 'scripts', 'kernel-router.js');
150
+ function route(prompt) {
151
+ return spawnSync(process.execPath, [kernel], {
152
+ input: JSON.stringify({ prompt }),
153
+ encoding: 'utf8',
154
+ cwd: ROOT,
155
+ env: { ...process.env }
156
+ });
157
+ }
158
+
159
+ const prompt = 'Fix a bug in an API endpoint and add a unit test.';
160
+ const first = route(prompt);
161
+ const second = route(prompt);
162
+ assert.strictEqual(first.status, 0, first.stderr);
163
+ assert.strictEqual(second.status, 0, second.stderr);
164
+ assert.strictEqual(first.stdout, second.stdout, 'identical routing input must produce identical output');
165
+ assert.match(first.stdout, /RECOMMENDED TIER:\s*Tier 2/i);
166
+ assert.match(first.stdout, /REQUIRED HARNESS INVARIANTS/);
167
+ assert.match(first.stdout, /SUGGESTED SKILLS \(ADVISORY/);
168
+ assert.doesNotMatch(first.stdout, /BASE EXECUTION LOOP/);
169
+ assert.match(first.stdout, /Do not enforce workflow order/);
170
+
171
+ const trivial = route('Update one README typo');
172
+ assert.strictEqual(trivial.status, 0, trivial.stderr);
173
+ assert.match(trivial.stdout, /RECOMMENDED TIER:\s*Tier 1/i);
174
+ assert.match(trivial.stdout, /No mandatory domain skill/);
175
+
176
+ const unmatched = route('frobnicate the quux');
177
+ assert.strictEqual(unmatched.status, 0, unmatched.stderr);
178
+ assert.match(unmatched.stdout, /RECOMMENDED TIER:\s*Unclassified/i);
179
+ assert.match(unmatched.stdout, /"tier":"unclassified"/i);
180
+ assert.doesNotMatch(unmatched.stdout, /RECOMMENDED TIER:\s*Tier 1/i);
181
+
182
+ console.log(`OpenAI plugin package verified: ${skillNames.length} skills, publication metadata, hooks, marketplace, deterministic routing.`);
@@ -0,0 +1,175 @@
1
+ 'use strict';
2
+
3
+ const assert = require('assert');
4
+ const fs = require('fs');
5
+ const os = require('os');
6
+ const path = require('path');
7
+ const { spawnSync } = require('child_process');
8
+
9
+ const ROOT = path.resolve(__dirname, '..');
10
+ const PLUGIN = path.join(ROOT, 'plugins', 'harness-everything');
11
+ const HOOKS_FILE = path.join(PLUGIN, 'hooks', 'hooks.json');
12
+ const hooks = JSON.parse(fs.readFileSync(HOOKS_FILE, 'utf8')).hooks;
13
+
14
+ const EXPECTED_EVENTS = [
15
+ 'SessionStart',
16
+ 'UserPromptSubmit',
17
+ 'PreToolUse',
18
+ 'PostToolUse',
19
+ 'SubagentStart',
20
+ 'SubagentStop',
21
+ 'Stop'
22
+ ];
23
+
24
+ for (const event of EXPECTED_EVENTS) {
25
+ assert.ok(Array.isArray(hooks[event]) && hooks[event].length > 0, `${event} hook missing`);
26
+ }
27
+
28
+ const hookText = JSON.stringify(hooks);
29
+ assert.match(hookText, /PLUGIN_ROOT/, 'OpenAI hooks must resolve from PLUGIN_ROOT');
30
+ assert.match(hookText, /commandWindows/, 'OpenAI hooks need Windows command overrides');
31
+
32
+ const commandPaths = [];
33
+ for (const entries of Object.values(hooks)) {
34
+ for (const entry of entries) {
35
+ for (const command of entry.hooks || []) {
36
+ assert.strictEqual(command.type, 'command');
37
+ assert.ok(command.commandWindows, 'every packaged command hook needs a Windows override');
38
+ const match = command.command.match(/\$PLUGIN_ROOT\/([^"']+)/);
39
+ if (match) commandPaths.push(path.join(PLUGIN, match[1]));
40
+ }
41
+ }
42
+ }
43
+ for (const commandPath of commandPaths) {
44
+ assert.ok(fs.existsSync(commandPath), `packaged hook command is missing: ${path.relative(PLUGIN, commandPath)}`);
45
+ }
46
+
47
+ function filesUnder(root, out = []) {
48
+ for (const entry of fs.readdirSync(root, { withFileTypes: true })) {
49
+ const full = path.join(root, entry.name);
50
+ if (entry.isDirectory()) filesUnder(full, out);
51
+ else if (entry.isFile()) out.push(full);
52
+ }
53
+ return out;
54
+ }
55
+
56
+ for (const file of filesUnder(path.join(PLUGIN, 'hooks')).filter(file => file.endsWith('.js'))) {
57
+ const syntax = spawnSync(process.execPath, ['--check', file], { encoding: 'utf8' });
58
+ assert.strictEqual(syntax.status, 0, `${path.relative(PLUGIN, file)} does not parse: ${syntax.stderr}`);
59
+ }
60
+ assert.ok(fs.existsSync(path.join(PLUGIN, 'scripts', 'lib', 'advisory-text.js')));
61
+ const packagedState = fs.readFileSync(path.join(PLUGIN, 'hooks', 'scripts', 'lib', 'harness-state.js'), 'utf8');
62
+ assert.doesNotMatch(packagedState, /\.\.\/\.\.\/\.\.\/scripts[\\/]lib[\\/]workspace/);
63
+
64
+ function runHook(relative, payload, stateHome) {
65
+ return spawnSync(process.execPath, [path.join(PLUGIN, relative)], {
66
+ input: JSON.stringify(payload),
67
+ encoding: 'utf8',
68
+ cwd: ROOT,
69
+ env: { ...process.env, HARNESS_STATE_HOME: stateHome }
70
+ });
71
+ }
72
+
73
+ const fixture = fs.mkdtempSync(path.join(os.tmpdir(), 'harness-openai-runtime-'));
74
+ const stateHome = path.join(fixture, 'state-home');
75
+ const gitInit = spawnSync('git', ['init', '--quiet'], { cwd: fixture, encoding: 'utf8' });
76
+ assert.strictEqual(gitInit.status, 0, gitInit.stderr);
77
+ const sessionPayload = {
78
+ hook_event_name: 'SessionStart',
79
+ session_id: 'openai-runtime-test',
80
+ cwd: fixture,
81
+ source: 'startup'
82
+ };
83
+
84
+ const session = runHook('hooks/session-start.js', sessionPayload, stateHome);
85
+ assert.strictEqual(session.status, 0, session.stderr);
86
+ const sessionOutput = JSON.parse(session.stdout);
87
+ assert.strictEqual(sessionOutput.hookSpecificOutput.hookEventName, 'SessionStart');
88
+ assert.match(sessionOutput.hookSpecificOutput.additionalContext, /Route before execution/);
89
+ assert.match(sessionOutput.hookSpecificOutput.additionalContext, /Verify before claim/);
90
+
91
+ const persist = runHook('hooks/scripts/state-persist.js', {
92
+ hook_event_name: 'PostToolUse',
93
+ session_id: 'openai-runtime-test',
94
+ cwd: fixture,
95
+ tool_name: 'apply_patch',
96
+ tool_input: { command: '*** Begin Patch\n*** Update File: src/example.js\n*** End Patch' },
97
+ tool_response: { exitCode: 0, stdout: '', stderr: '' }
98
+ }, stateHome);
99
+ assert.strictEqual(persist.status, 0, persist.stderr);
100
+
101
+ function findFile(root, filename) {
102
+ if (!fs.existsSync(root)) return null;
103
+ for (const entry of fs.readdirSync(root, { withFileTypes: true })) {
104
+ const full = path.join(root, entry.name);
105
+ if (entry.isDirectory()) {
106
+ const found = findFile(full, filename);
107
+ if (found) return found;
108
+ } else if (entry.isFile() && entry.name === filename) {
109
+ return full;
110
+ }
111
+ }
112
+ return null;
113
+ }
114
+
115
+ const handoff = findFile(stateHome, 'handoff-state.json');
116
+ assert.ok(handoff, 'SessionStart/state-persist should create a handoff state file');
117
+ assert.ok(JSON.parse(fs.readFileSync(handoff, 'utf8')).lastEditAt, 'apply_patch must count as an edit');
118
+
119
+ const sessionDir = path.dirname(handoff);
120
+ fs.writeFileSync(path.join(sessionDir, 'rule-of-3-state.json'), JSON.stringify({
121
+ count: 3,
122
+ threshold: 3,
123
+ lastHash: 'runtime-test-failure',
124
+ lastFailureAt: 0
125
+ }), 'utf8');
126
+ const reportFile = path.join(sessionDir, 'zoom-out-report.md');
127
+ const reportWrite = runHook('hooks/scripts/rule-of-3.js', {
128
+ hook_event_name: 'PreToolUse',
129
+ session_id: 'openai-runtime-test',
130
+ cwd: fixture,
131
+ tool_name: 'apply_patch',
132
+ tool_input: { patch: `*** Add File: ${reportFile}\n## Goal` }
133
+ }, stateHome);
134
+ assert.strictEqual(reportWrite.status, 0, 'apply_patch must be allowed to write the zoom-out report');
135
+
136
+ const mixedPatch = runHook('hooks/scripts/rule-of-3.js', {
137
+ hook_event_name: 'PreToolUse',
138
+ session_id: 'openai-runtime-test',
139
+ cwd: fixture,
140
+ tool_name: 'apply_patch',
141
+ tool_input: { patch: `*** Update File: ${reportFile}\n@@\n*** Update File: unrelated.txt\n@@` }
142
+ }, stateHome);
143
+ assert.strictEqual(mixedPatch.status, 2, 'report exemption must not allow a mixed apply_patch');
144
+
145
+ const blockedPatch = runHook('hooks/scripts/rule-of-3.js', {
146
+ hook_event_name: 'PreToolUse',
147
+ session_id: 'openai-runtime-test',
148
+ cwd: fixture,
149
+ tool_name: 'apply_patch',
150
+ tool_input: { patch: '*** Update File: unrelated.txt\n@@' }
151
+ }, stateHome);
152
+ assert.strictEqual(blockedPatch.status, 2, 'tripped circuit breaker must block unrelated apply_patch writes');
153
+
154
+ const subagentStart = runHook('hooks/scripts/subagent-start.js', {
155
+ hook_event_name: 'SubagentStart',
156
+ session_id: 'openai-runtime-test',
157
+ cwd: fixture,
158
+ agent_type: 'worker'
159
+ }, stateHome);
160
+ assert.strictEqual(subagentStart.status, 0, subagentStart.stderr);
161
+ const subagentOutput = JSON.parse(subagentStart.stdout);
162
+ assert.strictEqual(subagentOutput.hookSpecificOutput.hookEventName, 'SubagentStart');
163
+ assert.match(subagentOutput.hookSpecificOutput.additionalContext, /scope/i);
164
+
165
+ fs.writeFileSync(path.join(fixture, 'unplanned-change.txt'), 'scope regression\n', 'utf8');
166
+ const subagentStop = runHook('hooks/scripts/subagent-stop.js', {
167
+ hook_event_name: 'SubagentStop',
168
+ session_id: 'openai-runtime-test',
169
+ cwd: fixture,
170
+ agent_type: 'worker'
171
+ }, stateHome);
172
+ assert.strictEqual(subagentStop.status, 2, subagentStop.stderr);
173
+ assert.match(subagentStop.stderr, /Subagent Scope Guard/);
174
+
175
+ console.log('OpenAI plugin runtime compatibility verified.');