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 +1,183 @@
1
+ const fs = require('fs');
2
+ const os = require('os');
3
+ const path = require('path');
4
+ const helper = require('./test-helper');
5
+
6
+ console.log('\n[2p] Canonical-store link/copy install + link-aware uninstall (issue #49)...');
7
+
8
+ const skills = require('../scripts/lib/skills');
9
+ const manifest = require('../scripts/lib/manifest');
10
+
11
+ const harnessSourceDir = path.resolve(__dirname, '..');
12
+ const ws = fs.mkdtempSync(path.join(os.tmpdir(), 'harness-2p-ws-'));
13
+ const home = fs.mkdtempSync(path.join(os.tmpdir(), 'harness-2p-home-'));
14
+
15
+ function target(name, dirName) {
16
+ const home_ = path.join(ws, dirName);
17
+ return { path: path.join(home_, 'skills'), label: `${dirName}/skills/`, manifestPath: manifest.getManifestPath(home_) };
18
+ }
19
+
20
+ // --- 1. Link mode: two targets sharing one canonical copy ---
21
+ const claudeTarget = target('claude', '.claude');
22
+ const cursorTarget = target('cursor', '.cursor');
23
+ const canonicalDir = skills.getCanonicalSkillsDir({ isGlobal: false, workspaceRoot: ws, userHome: home });
24
+
25
+ skills.installSkillsToTargets({
26
+ chosenSkills: ['tdd'],
27
+ targetDirs: [claudeTarget, cursorTarget],
28
+ harnessSourceDir,
29
+ packageVersion: '9.9.9',
30
+ canonicalDir,
31
+ linkMode: 'auto',
32
+ });
33
+
34
+ const claudeSkillDir = path.join(claudeTarget.path, 'tdd');
35
+ const cursorSkillDir = path.join(cursorTarget.path, 'tdd');
36
+ const canonicalSkillDir = path.join(canonicalDir, 'tdd');
37
+
38
+ helper.check(
39
+ '2p. both platform targets are real links (symlink/junction), not physical copies',
40
+ skills.isLinkPath(claudeSkillDir) && skills.isLinkPath(cursorSkillDir),
41
+ `claude isLink=${skills.isLinkPath(claudeSkillDir)} cursor isLink=${skills.isLinkPath(cursorSkillDir)}`
42
+ );
43
+ helper.check(
44
+ '2p. the canonical copy itself is a physical directory, not a link',
45
+ fs.existsSync(canonicalSkillDir) && !skills.isLinkPath(canonicalSkillDir),
46
+ `canonical exists=${fs.existsSync(canonicalSkillDir)} isLink=${skills.isLinkPath(canonicalSkillDir)}`
47
+ );
48
+ helper.check(
49
+ '2p. content is readable through the link and matches source',
50
+ fs.existsSync(path.join(claudeSkillDir, 'SKILL.md')) &&
51
+ fs.readFileSync(path.join(claudeSkillDir, 'SKILL.md'), 'utf8') === fs.readFileSync(path.join(harnessSourceDir, 'tdd', 'SKILL.md'), 'utf8'),
52
+ 'linked SKILL.md content did not match source'
53
+ );
54
+
55
+ const claudeManifestEntry = manifest.readManifest(claudeTarget.manifestPath).skills[0];
56
+ helper.check(
57
+ '2p. manifest records kind + canonicalPath for a linked entry',
58
+ claudeManifestEntry && ['symlink', 'junction'].includes(claudeManifestEntry.kind) && claudeManifestEntry.canonicalPath === canonicalSkillDir,
59
+ `Got: ${JSON.stringify(claudeManifestEntry)}`
60
+ );
61
+
62
+ // --- 2. getInstalledSkills surfaces both links, and the canonical dir is
63
+ // NOT double-listed as its own separate skill (advisor-flagged legacy-scan
64
+ // collision fix). ---
65
+ {
66
+ const installed = skills.getInstalledSkills(ws, home);
67
+ const tddEntries = installed.filter(s => s.id === 'tdd');
68
+ helper.check(
69
+ '2p. getInstalledSkills lists exactly the 2 linked entries, not the canonical dir as a 3rd',
70
+ tddEntries.length === 2,
71
+ `Got ${tddEntries.length}: ${JSON.stringify(tddEntries.map(e => e.dirPath))}`
72
+ );
73
+ }
74
+
75
+ // --- 3. Removing one link leaves the canonical copy (still referenced by
76
+ // the other link) intact; removing the last link removes it too. ---
77
+ {
78
+ const installedBefore = skills.getInstalledSkills(ws, home);
79
+ const claudeEntry = installedBefore.find(s => s.dirPath === claudeSkillDir);
80
+ skills.removeSkill(claudeEntry, { workspaceRoot: ws, userHome: home });
81
+
82
+ helper.check(
83
+ '2p. removing one link deletes only that link, canonical copy survives (still referenced)',
84
+ !fs.existsSync(claudeSkillDir) && fs.existsSync(canonicalSkillDir),
85
+ `claude exists=${fs.existsSync(claudeSkillDir)} canonical exists=${fs.existsSync(canonicalSkillDir)}`
86
+ );
87
+
88
+ const installedAfterFirst = skills.getInstalledSkills(ws, home);
89
+ const cursorEntry = installedAfterFirst.find(s => s.dirPath === cursorSkillDir);
90
+ skills.removeSkill(cursorEntry, { workspaceRoot: ws, userHome: home });
91
+
92
+ helper.check(
93
+ '2p. removing the LAST link also removes the now-orphaned canonical copy',
94
+ !fs.existsSync(cursorSkillDir) && !fs.existsSync(canonicalSkillDir),
95
+ `cursor exists=${fs.existsSync(cursorSkillDir)} canonical exists=${fs.existsSync(canonicalSkillDir)}`
96
+ );
97
+ }
98
+
99
+ // --- 4. --copy mode produces independent physical copies, no canonical
100
+ // store, no links at all. ---
101
+ {
102
+ const copyTarget = target('codex', '.codex');
103
+ skills.installSkillsToTargets({
104
+ chosenSkills: ['tdd'],
105
+ targetDirs: [copyTarget],
106
+ harnessSourceDir,
107
+ packageVersion: '9.9.9',
108
+ canonicalDir: skills.getCanonicalSkillsDir({ isGlobal: false, workspaceRoot: ws, userHome: home }),
109
+ linkMode: 'copy',
110
+ });
111
+ const copySkillDir = path.join(copyTarget.path, 'tdd');
112
+ helper.check(
113
+ '2p. --copy mode installs a physical directory, not a link',
114
+ fs.existsSync(copySkillDir) && !skills.isLinkPath(copySkillDir),
115
+ `exists=${fs.existsSync(copySkillDir)} isLink=${skills.isLinkPath(copySkillDir)}`
116
+ );
117
+ const entry = manifest.readManifest(copyTarget.manifestPath).skills[0];
118
+ helper.check(
119
+ '2p. --copy mode manifest entry has no kind/canonicalPath',
120
+ entry && !entry.kind && !entry.canonicalPath,
121
+ `Got: ${JSON.stringify(entry)}`
122
+ );
123
+ const installedEntry = skills.getInstalledSkills(ws, home).find(s => s.dirPath === copySkillDir);
124
+ skills.removeSkill(installedEntry, { workspaceRoot: ws, userHome: home });
125
+ helper.check('2p. --copy mode entry uninstalls cleanly', !fs.existsSync(copySkillDir), 'copy-mode dir still exists after removeSkill');
126
+ }
127
+
128
+ // --- 5. Dangling link (canonical target manually deleted) is still
129
+ // detected and cleaned up rather than silently ignored. ---
130
+ {
131
+ const danglingTarget = target('continue', '.continue');
132
+ skills.installSkillsToTargets({
133
+ chosenSkills: ['tdd'],
134
+ targetDirs: [danglingTarget],
135
+ harnessSourceDir,
136
+ packageVersion: '9.9.9',
137
+ canonicalDir: skills.getCanonicalSkillsDir({ isGlobal: false, workspaceRoot: ws, userHome: home }),
138
+ linkMode: 'symlink',
139
+ });
140
+ const linkDir = path.join(danglingTarget.path, 'tdd');
141
+ const canonicalTddDir = path.join(skills.getCanonicalSkillsDir({ isGlobal: false, workspaceRoot: ws, userHome: home }), 'tdd');
142
+ fs.rmSync(canonicalTddDir, { recursive: true, force: true }); // simulate manual/out-of-band deletion
143
+
144
+ helper.check(
145
+ '2p. a dangling link is still detected as present (pathPresent) even though its target is gone',
146
+ skills.pathPresent(linkDir) && !fs.existsSync(linkDir),
147
+ `pathPresent=${skills.pathPresent(linkDir)} existsSync=${fs.existsSync(linkDir)}`
148
+ );
149
+
150
+ const installed = skills.getInstalledSkills(ws, home);
151
+ const danglingEntry = installed.find(s => s.dirPath === linkDir);
152
+ helper.check('2p. getInstalledSkills still surfaces the dangling link for cleanup', !!danglingEntry, 'dangling link entry missing from getInstalledSkills');
153
+ if (danglingEntry) {
154
+ skills.removeSkill(danglingEntry, { workspaceRoot: ws, userHome: home });
155
+ helper.check(
156
+ '2p. removeSkill cleans up the dangling link and its manifest row',
157
+ !skills.isLinkPath(linkDir) && manifest.readManifest(danglingTarget.manifestPath).skills.length === 0,
158
+ `isLink=${skills.isLinkPath(linkDir)} manifest=${JSON.stringify(manifest.readManifest(danglingTarget.manifestPath))}`
159
+ );
160
+ }
161
+ }
162
+
163
+ // --- 6. A legacy manifest entry with no canonicalPath (pre-#49 install)
164
+ // still uninstalls exactly as before - no forced migration required. ---
165
+ {
166
+ const legacyTarget = target('copilot', '.github');
167
+ fs.mkdirSync(legacyTarget.path, { recursive: true });
168
+ const legacyDir = path.join(legacyTarget.path, 'tdd');
169
+ skills.copyDir(path.join(harnessSourceDir, 'tdd'), legacyDir);
170
+ manifest.recordSkillInstall(legacyTarget.manifestPath, '0.3.0', 'tdd', legacyDir); // no linkInfo - pre-#49 shape
171
+
172
+ const entry = skills.getInstalledSkills(ws, home).find(s => s.dirPath === legacyDir);
173
+ helper.check('2p. a pre-#49 manifest entry (no kind/canonicalPath) is still discovered', !!entry, 'legacy entry not found');
174
+ if (entry) {
175
+ skills.removeSkill(entry, { workspaceRoot: ws, userHome: home });
176
+ helper.check('2p. a pre-#49 entry uninstalls cleanly with no canonical-store side effects', !fs.existsSync(legacyDir), 'legacy dir still exists');
177
+ }
178
+ }
179
+
180
+ fs.rmSync(ws, { recursive: true, force: true });
181
+ fs.rmSync(home, { recursive: true, force: true });
182
+
183
+ helper.finish();
@@ -0,0 +1,132 @@
1
+ const fs = require('fs');
2
+ const os = require('os');
3
+ const path = require('path');
4
+ const { spawnSync } = require('child_process');
5
+ const helper = require('./test-helper');
6
+
7
+ console.log('\n[2q] Global scope: canonical store reuses the pre-existing shared ~/.agents/skills (issue #49)...');
8
+
9
+ const skills = require('../scripts/lib/skills');
10
+ const manifest = require('../scripts/lib/manifest');
11
+
12
+ const harnessSourceDir = path.resolve(__dirname, '..');
13
+
14
+ const installerHome = fs.mkdtempSync(path.join(os.tmpdir(), 'harness-2q-installer-home-'));
15
+ const installerCwd = fs.mkdtempSync(path.join(os.tmpdir(), 'harness-2q-installer-cwd-'));
16
+ const installerResult = spawnSync(
17
+ process.execPath,
18
+ [path.join(harnessSourceDir, 'scripts', 'installer.js'), 'install', '--global', '--yes', '--no-skills', '--claude'],
19
+ {
20
+ cwd: installerCwd,
21
+ env: { ...process.env, HOME: installerHome, USERPROFILE: installerHome, APPDATA: path.join(installerHome, 'AppData', 'Roaming') },
22
+ encoding: 'utf8'
23
+ }
24
+ );
25
+ helper.check(
26
+ '2q. global installer works from a non-git directory',
27
+ installerResult.status === 0 && fs.existsSync(path.join(installerHome, '.claude', 'settings.json')),
28
+ `status=${installerResult.status} stdout=${installerResult.stdout} stderr=${installerResult.stderr}`
29
+ );
30
+ fs.rmSync(installerHome, { recursive: true, force: true });
31
+ fs.rmSync(installerCwd, { recursive: true, force: true });
32
+
33
+ const home = fs.mkdtempSync(path.join(os.tmpdir(), 'harness-2q-home-'));
34
+ const ws = fs.mkdtempSync(path.join(os.tmpdir(), 'harness-2q-ws-'));
35
+
36
+ // Mirrors what cursor.js/codex.js/continue.js/copilot.js/hermes.js's
37
+ // getSkillsTarget({isGlobal:true}) all independently return today: the SAME
38
+ // path + SAME manifest, deduped in installer.js's main() into one targetDirs
39
+ // entry before installSkillsToTargets ever runs.
40
+ const globalAgentsDir = path.join(home, '.agents');
41
+ const sharedTarget = {
42
+ path: path.join(globalAgentsDir, 'skills'),
43
+ label: '~/.agents/skills/',
44
+ manifestPath: manifest.getManifestPath(globalAgentsDir),
45
+ };
46
+ // claude.js's global target is separate: ~/.claude/skills/.
47
+ const claudeGlobalTarget = {
48
+ path: path.join(home, '.claude', 'skills'),
49
+ label: '~/.claude/skills/',
50
+ manifestPath: manifest.getManifestPath(path.join(home, '.claude')),
51
+ };
52
+
53
+ const canonicalDir = skills.getCanonicalSkillsDir({ isGlobal: true, workspaceRoot: ws, userHome: home });
54
+ helper.check(
55
+ '2q. global canonical dir IS the pre-existing shared ~/.agents/skills path, not a new location',
56
+ canonicalDir === sharedTarget.path,
57
+ `canonicalDir=${canonicalDir} sharedTarget.path=${sharedTarget.path}`
58
+ );
59
+
60
+ skills.installSkillsToTargets({
61
+ chosenSkills: ['tdd'],
62
+ targetDirs: [sharedTarget, claudeGlobalTarget],
63
+ harnessSourceDir,
64
+ packageVersion: '9.9.9',
65
+ canonicalDir,
66
+ linkMode: 'auto',
67
+ });
68
+
69
+ const sharedSkillDir = path.join(sharedTarget.path, 'tdd');
70
+ const claudeSkillDir = path.join(claudeGlobalTarget.path, 'tdd');
71
+
72
+ helper.check(
73
+ '2q. the shared target (cursor/codex/continue/copilot/hermes stand-in) gets the PHYSICAL copy, not a link to itself',
74
+ fs.existsSync(sharedSkillDir) && !skills.isLinkPath(sharedSkillDir),
75
+ `exists=${fs.existsSync(sharedSkillDir)} isLink=${skills.isLinkPath(sharedSkillDir)}`
76
+ );
77
+ helper.check(
78
+ "2q. Claude's separate global target links into the shared store instead of duplicating it",
79
+ skills.isLinkPath(claudeSkillDir),
80
+ `isLink=${skills.isLinkPath(claudeSkillDir)}`
81
+ );
82
+
83
+ const sharedManifestEntry = manifest.readManifest(sharedTarget.manifestPath).skills.find(s => s.id === 'tdd');
84
+ helper.check(
85
+ '2q. the shared-store manifest entry records a plain copy (no kind/canonicalPath)',
86
+ sharedManifestEntry && !sharedManifestEntry.kind && !sharedManifestEntry.canonicalPath,
87
+ `Got: ${JSON.stringify(sharedManifestEntry)}`
88
+ );
89
+
90
+ // --- Removing Claude's link must NOT delete the shared store: 5 other
91
+ // platforms still depend on it directly (their own manifest row IS that
92
+ // path), even though nothing else "links" to it. ---
93
+ {
94
+ const installed = skills.getInstalledSkills(ws, home);
95
+ const claudeEntry = installed.find(s => s.dirPath === claudeSkillDir);
96
+ skills.removeSkill(claudeEntry, { workspaceRoot: ws, userHome: home });
97
+
98
+ helper.check(
99
+ "2q. removing Claude's link leaves the shared store intact (still independently in use)",
100
+ !fs.existsSync(claudeSkillDir) && fs.existsSync(sharedSkillDir),
101
+ `claude exists=${fs.existsSync(claudeSkillDir)} shared exists=${fs.existsSync(sharedSkillDir)}`
102
+ );
103
+ }
104
+
105
+ // --- Removing the shared store directly must also clean up any dependent
106
+ // link (symmetric case - re-link Claude first to set this up). ---
107
+ {
108
+ skills.installSkillsToTargets({
109
+ chosenSkills: ['tdd'],
110
+ targetDirs: [claudeGlobalTarget],
111
+ harnessSourceDir,
112
+ packageVersion: '9.9.9',
113
+ canonicalDir,
114
+ linkMode: 'auto',
115
+ });
116
+ helper.check('2q. (setup) Claude re-linked to the shared store', skills.isLinkPath(claudeSkillDir), 'relink failed');
117
+
118
+ const installed = skills.getInstalledSkills(ws, home);
119
+ const sharedEntry = installed.find(s => s.dirPath === sharedSkillDir);
120
+ skills.removeSkill(sharedEntry, { workspaceRoot: ws, userHome: home });
121
+
122
+ helper.check(
123
+ "2q. removing the shared store directly also sweeps Claude's now-dangling dependent link",
124
+ !fs.existsSync(sharedSkillDir) && !skills.isLinkPath(claudeSkillDir) && !fs.existsSync(claudeSkillDir),
125
+ `shared exists=${fs.existsSync(sharedSkillDir)} claude isLink=${skills.isLinkPath(claudeSkillDir)} claude exists=${fs.existsSync(claudeSkillDir)}`
126
+ );
127
+ }
128
+
129
+ fs.rmSync(home, { recursive: true, force: true });
130
+ fs.rmSync(ws, { recursive: true, force: true });
131
+
132
+ helper.finish();
@@ -0,0 +1,148 @@
1
+ const fs = require('fs');
2
+ const os = require('os');
3
+ const path = require('path');
4
+ const { spawnSync } = require('child_process');
5
+ const helper = require('./test-helper');
6
+
7
+ console.log('\n[2r] Uninstall CLI flag handling: no silent no-op, combined local+global (issue #48)...');
8
+
9
+ const repoRoot = path.resolve(__dirname, '..');
10
+ const installerPath = path.join(repoRoot, 'scripts', 'installer.js');
11
+
12
+ function freshWorkspace(prefix) {
13
+ const ws = fs.mkdtempSync(path.join(os.tmpdir(), prefix));
14
+ fs.mkdirSync(path.join(ws, '.git')); // pins getWorkspaceRoot() deterministically
15
+ return ws;
16
+ }
17
+
18
+ function run(cwd, args, env) {
19
+ // spawnSync with stdio 'pipe' has no attached TTY - process.stdin.isTTY is
20
+ // undefined, exactly like a piped/non-interactive invocation (npx in CI,
21
+ // some shells, etc.) - the scenario issue #48 is about.
22
+ return spawnSync(process.execPath, [installerPath, ...args], { cwd, env, encoding: 'utf8' });
23
+ }
24
+
25
+ function seedLocalInstall(ws, skillId) {
26
+ fs.mkdirSync(path.join(ws, '.claude'), { recursive: true });
27
+ fs.writeFileSync(path.join(ws, '.claude', 'settings.json'), '{}', 'utf8');
28
+ if (skillId) {
29
+ const skillDir = path.join(ws, '.claude', 'skills', skillId);
30
+ fs.mkdirSync(skillDir, { recursive: true });
31
+ fs.writeFileSync(path.join(skillDir, 'SKILL.md'), '---\nname: x\nauthor: Miya Daniel\n---\n', 'utf8');
32
+ const manifest = require('../scripts/lib/manifest');
33
+ manifest.recordSkillInstall(manifest.getManifestPath(path.join(ws, '.claude')), require('../package.json').version, skillId, skillDir);
34
+ }
35
+ }
36
+
37
+ function seedGlobalInstall(fakeHome, skillId) {
38
+ fs.mkdirSync(path.join(fakeHome, '.claude'), { recursive: true });
39
+ fs.writeFileSync(path.join(fakeHome, '.claude', 'settings.json'), '{}', 'utf8');
40
+ if (skillId) {
41
+ const skillDir = path.join(fakeHome, '.claude', 'skills', skillId);
42
+ fs.mkdirSync(skillDir, { recursive: true });
43
+ fs.writeFileSync(path.join(skillDir, 'SKILL.md'), '---\nname: x\nauthor: Miya Daniel\n---\n', 'utf8');
44
+ const manifest = require('../scripts/lib/manifest');
45
+ manifest.recordSkillInstall(manifest.getManifestPath(path.join(fakeHome, '.claude')), require('../package.json').version, skillId, skillDir);
46
+ }
47
+ }
48
+
49
+ // --- 1. Bare `uninstall`, no flags, no TTY: must fail loudly, not silently
50
+ // claim success while removing nothing. ---
51
+ {
52
+ const ws = freshWorkspace('harness-2r-noop-');
53
+ const fakeHome = fs.mkdtempSync(path.join(os.tmpdir(), 'harness-2r-noop-home-'));
54
+ seedLocalInstall(ws, 'dummy-skill');
55
+ const env = { ...process.env, HOME: fakeHome, USERPROFILE: fakeHome };
56
+
57
+ const res = run(ws, ['uninstall'], env);
58
+
59
+ helper.check('2r. bare uninstall with no TTY exits non-zero instead of silently succeeding', res.status !== 0, `status=${res.status} stdout=${res.stdout} stderr=${res.stderr}`);
60
+ helper.check('2r. ...and prints a hint about the flags to pass', /--local|--skills|--global/.test(res.stderr), `stderr=${res.stderr}`);
61
+ helper.check(
62
+ '2r. ...and actually removed nothing (settings.json + skill both still present)',
63
+ fs.existsSync(path.join(ws, '.claude', 'settings.json')) && fs.existsSync(path.join(ws, '.claude', 'skills', 'dummy-skill')),
64
+ 'local install was modified despite no flags being passed'
65
+ );
66
+
67
+ fs.rmSync(ws, { recursive: true, force: true });
68
+ fs.rmSync(fakeHome, { recursive: true, force: true });
69
+ }
70
+
71
+ // --- 2. `uninstall --local --skills` (no -y): removes exactly local config +
72
+ // local skills, in one non-interactive run, no menu. ---
73
+ {
74
+ const ws = freshWorkspace('harness-2r-local-');
75
+ const fakeHome = fs.mkdtempSync(path.join(os.tmpdir(), 'harness-2r-local-home-'));
76
+ seedLocalInstall(ws, 'dummy-skill');
77
+ const env = { ...process.env, HOME: fakeHome, USERPROFILE: fakeHome };
78
+
79
+ const res = run(ws, ['uninstall', '--local', '--skills'], env);
80
+
81
+ helper.check('2r. `uninstall --local --skills` exits 0', res.status === 0, `status=${res.status} stdout=${res.stdout} stderr=${res.stderr}`);
82
+ helper.check(
83
+ '2r. ...and actually removed the local skill',
84
+ !fs.existsSync(path.join(ws, '.claude', 'skills', 'dummy-skill')),
85
+ `skill dir still exists, stdout=${res.stdout}`
86
+ );
87
+
88
+ fs.rmSync(ws, { recursive: true, force: true });
89
+ fs.rmSync(fakeHome, { recursive: true, force: true });
90
+ }
91
+
92
+ // --- 3. `uninstall --local --global --skills` in ONE run removes both
93
+ // scopes together (issue #48 point 3) - no need for two separate invocations. ---
94
+ {
95
+ const ws = freshWorkspace('harness-2r-both-');
96
+ const fakeHome = fs.mkdtempSync(path.join(os.tmpdir(), 'harness-2r-both-home-'));
97
+ seedLocalInstall(ws, 'local-skill');
98
+ seedGlobalInstall(fakeHome, 'global-skill');
99
+ const env = { ...process.env, HOME: fakeHome, USERPROFILE: fakeHome };
100
+
101
+ const res = run(ws, ['uninstall', '--local', '--global', '--skills'], env);
102
+
103
+ helper.check('2r. `uninstall --local --global --skills` exits 0', res.status === 0, `status=${res.status} stdout=${res.stdout} stderr=${res.stderr}`);
104
+ helper.check(
105
+ '2r. ...local skill removed',
106
+ !fs.existsSync(path.join(ws, '.claude', 'skills', 'local-skill')),
107
+ `local skill still exists, stdout=${res.stdout}`
108
+ );
109
+ helper.check(
110
+ '2r. ...AND global skill removed in the SAME run',
111
+ !fs.existsSync(path.join(fakeHome, '.claude', 'skills', 'global-skill')),
112
+ `global skill still exists, stdout=${res.stdout}`
113
+ );
114
+
115
+ fs.rmSync(ws, { recursive: true, force: true });
116
+ fs.rmSync(fakeHome, { recursive: true, force: true });
117
+ }
118
+
119
+ // --- 4. Point 2 regression: a second bare `uninstall --local` after a first
120
+ // run that intentionally left skills in place still finds and can remove
121
+ // them (was previously masked by the same no-op bug as scenario 1). ---
122
+ {
123
+ const ws = freshWorkspace('harness-2r-second-run-');
124
+ const fakeHome = fs.mkdtempSync(path.join(os.tmpdir(), 'harness-2r-second-run-home-'));
125
+ seedLocalInstall(ws, 'dummy-skill');
126
+ const env = { ...process.env, HOME: fakeHome, USERPROFILE: fakeHome };
127
+
128
+ const first = run(ws, ['uninstall', '--local'], env); // config only, skills intentionally kept
129
+ helper.check('2r. first run (--local only) exits 0', first.status === 0, `status=${first.status} stdout=${first.stdout}`);
130
+ helper.check(
131
+ '2r. ...and leaves the skill in place as requested',
132
+ fs.existsSync(path.join(ws, '.claude', 'skills', 'dummy-skill')),
133
+ 'skill was removed even though --skills was not passed'
134
+ );
135
+
136
+ const second = run(ws, ['uninstall', '--skills'], env);
137
+ helper.check('2r. second run (--skills) exits 0', second.status === 0, `status=${second.status} stdout=${second.stdout} stderr=${second.stderr}`);
138
+ helper.check(
139
+ '2r. ...and this time removes the skill left over from the first run',
140
+ !fs.existsSync(path.join(ws, '.claude', 'skills', 'dummy-skill')),
141
+ `skill still exists after second run, stdout=${second.stdout}`
142
+ );
143
+
144
+ fs.rmSync(ws, { recursive: true, force: true });
145
+ fs.rmSync(fakeHome, { recursive: true, force: true });
146
+ }
147
+
148
+ helper.finish();
@@ -0,0 +1,203 @@
1
+ const fs = require('fs');
2
+ const os = require('os');
3
+ const path = require('path');
4
+ const { spawnSync } = require('child_process');
5
+ const helper = require('./test-helper');
6
+ const { evaluateEvidence } = require('../tdd/scripts/quality-gate');
7
+
8
+ console.log('\n[2s] TDD quality evidence gate (issue #58)...');
9
+
10
+ const classes = {
11
+ inputCompleteness: ['required', 'optional', 'omitted', 'empty', 'null', 'min', 'max', 'boundary'],
12
+ outputCompleteness: ['value', 'type', 'schema', 'required-field', 'prohibited-extra-field', 'side-effect'],
13
+ errorHandling: ['error-type', 'error-code', 'message', 'state-cleanup', 'rollback', 'no-unintended-output'],
14
+ unexpectedInput: ['malformed', 'unknown', 'extra', 'unsupported', 'out-of-range', 'duplicate', 'wrong-type']
15
+ };
16
+
17
+ const scenarioCatalog = JSON.parse(fs.readFileSync(path.join(helper.root, 'tdd', 'fixtures', 'quality-scenarios.json'), 'utf8'));
18
+ const requiredScenarios = ['PASS', 'unit profile', 'integration profile', 'weighted integration failure', 'weighted unit failure', 'NON_CONFORMANT', 'SOURCE_DEFECT', 'SOURCE_CONFLICT', 'malformed input', 'unexpected input', 'error handling', 'nondeterministic output', 'auditable N/A', 'skipped test'];
19
+ helper.check('2s. fixture catalog covers every required outcome', requiredScenarios.every(name => scenarioCatalog.scenarios.some(scenario => scenario.name === name)));
20
+
21
+ function checks(prefix, names) {
22
+ return names.map((name, index) => ({
23
+ id: `${prefix}-${index + 1}`,
24
+ class: name,
25
+ status: 'PASS',
26
+ evidence: `tests/example.test.js#${prefix}-${name}`
27
+ }));
28
+ }
29
+
30
+ function validEvidence() {
31
+ const sideEffects = {
32
+ createdFiles: [], changedFiles: [], deletedFiles: [],
33
+ persistentState: {}, events: [], logs: [], externalCalls: []
34
+ };
35
+ return {
36
+ schemaVersion: '1.0.0',
37
+ requirements: [{
38
+ requirementId: 'REQ-001',
39
+ testProfile: 'integration',
40
+ source: {
41
+ type: 'SPEC', path: 'docs/example.md', section: 'Expected behavior', status: 'AUTHORITATIVE'
42
+ },
43
+ sourceConformance: 'PASS',
44
+ sourceValidity: 'VALID',
45
+ dimensions: {
46
+ positiveParameters: { applicable: true, checks: checks('positive', ['supported-partition']) },
47
+ negativeParameters: { applicable: true, checks: checks('negative', ['rejected-partition']) },
48
+ inputCompleteness: { applicable: true, checks: checks('input', classes.inputCompleteness) },
49
+ outputCompleteness: { applicable: true, checks: checks('output', classes.outputCompleteness) },
50
+ errorHandling: { applicable: true, checks: checks('error', classes.errorHandling) },
51
+ unexpectedInput: { applicable: true, checks: checks('unexpected', classes.unexpectedInput) },
52
+ determinism: {
53
+ applicable: true,
54
+ runs: [
55
+ { id: 'run-1', input: { value: 1 }, exitStatus: 0, errorCategory: null, output: { value: 2 }, sideEffects },
56
+ { id: 'run-2', input: { value: 1 }, exitStatus: 0, errorCategory: null, output: { value: 2 }, sideEffects }
57
+ ],
58
+ normalizations: []
59
+ },
60
+ sourceTraceability: { applicable: true }
61
+ },
62
+ skippedTests: [],
63
+ naDeclarations: []
64
+ }]
65
+ };
66
+ }
67
+
68
+ function clone(value) {
69
+ return JSON.parse(JSON.stringify(value));
70
+ }
71
+
72
+ function evaluate(mutator) {
73
+ const evidence = validEvidence();
74
+ if (mutator) mutator(evidence.requirements[0], evidence);
75
+ return evaluateEvidence(evidence);
76
+ }
77
+
78
+ const pass = evaluate();
79
+ helper.check('2s. complete profile-routed evidence passes with a 100 score', pass.result === 'PASS' && pass.score === 100 && pass.requirements[0].testProfile === 'integration');
80
+ helper.check(
81
+ '2s. report publishes the canonical weights and pass threshold',
82
+ pass.scoring.requiredScore === 100 && pass.scoring.weights.positiveParameters === 0.15 && pass.scoring.weights.determinism === 0.10
83
+ );
84
+
85
+ const weightedIntegration = evaluate(req => { req.dimensions.positiveParameters.checks[0].status = 'FAIL'; });
86
+ helper.check('2s. integration score follows the canonical weighted formula exactly', weightedIntegration.score === 85);
87
+
88
+ const missingProfile = evaluate(req => { delete req.testProfile; });
89
+ helper.check('2s. every requirement must declare its selected test profile', missingProfile.result === 'FAIL' && missingProfile.errors.some(e => /testProfile/.test(e)));
90
+
91
+ const unknownProfile = evaluate(req => { req.testProfile = 'contract'; });
92
+ helper.check('2s. only unit and integration profiles are accepted', unknownProfile.result === 'FAIL' && unknownProfile.errors.some(e => /unit or integration/.test(e)));
93
+
94
+ const unitProfile = evaluate(req => {
95
+ req.testProfile = 'unit';
96
+ req.dimensions.determinism = { applicable: false, reason: 'This isolated unit does not cross an integration boundary.' };
97
+ });
98
+ helper.check('2s. unit evidence may mark integration determinism N/A with a reason', unitProfile.result === 'PASS' && unitProfile.requirements[0].dimensions.determinism.applicable === false);
99
+
100
+ const weightedUnit = evaluate(req => {
101
+ req.testProfile = 'unit';
102
+ req.dimensions.determinism = { applicable: false, reason: 'This isolated unit does not cross an integration boundary.' };
103
+ req.dimensions.positiveParameters.checks[0].status = 'FAIL';
104
+ });
105
+ helper.check('2s. unit score renormalizes only the applicable canonical weights', weightedUnit.score === 83.33);
106
+
107
+ const integrationWithoutRepeat = evaluate(req => {
108
+ req.dimensions.determinism = { applicable: false, reason: 'Not measured.' };
109
+ });
110
+ helper.check('2s. integration evidence cannot bypass repeated determinism checks', integrationWithoutRepeat.result === 'FAIL' && integrationWithoutRepeat.errors.some(e => /integration.*determinism/.test(e)));
111
+ helper.check(
112
+ '2s. report exposes raw counts, percentages, and mandatory gates',
113
+ pass.dimensions.inputCompleteness.passed === 8 &&
114
+ pass.dimensions.inputCompleteness.required === 8 &&
115
+ pass.dimensions.inputCompleteness.percentage === 100 &&
116
+ Object.values(pass.mandatoryGates).every(Boolean)
117
+ );
118
+
119
+ const incompleteInput = evaluate(req => { req.dimensions.inputCompleteness.checks.pop(); });
120
+ helper.check('2s. missing an applicable input class is rejected', incompleteInput.result === 'FAIL' && incompleteInput.errors.some(e => /inputCompleteness.*boundary/.test(e)));
121
+
122
+ const incompleteOutput = evaluate(req => { req.dimensions.outputCompleteness.checks[0].status = 'FAIL'; });
123
+ helper.check('2s. incomplete output validation is rejected', incompleteOutput.result === 'FAIL' && incompleteOutput.mandatoryGates.allRequiredInputOutputValidated === false);
124
+
125
+ const failedUnexpected = evaluate(req => { req.dimensions.unexpectedInput.checks[0].status = 'FAIL'; });
126
+ helper.check('2s. failed unexpected-input behavior is rejected', failedUnexpected.result === 'FAIL' && failedUnexpected.dimensions.unexpectedInput.percentage < 100);
127
+
128
+ const untestedError = evaluate(req => { req.dimensions.errorHandling.checks[0].status = 'SKIPPED'; });
129
+ helper.check('2s. untested and skipped error paths are rejected and visible', untestedError.result === 'FAIL' && untestedError.skippedTests.length === 1 && untestedError.mandatoryGates.allDeclaredErrorPathsTested === false);
130
+
131
+ const nondeterministic = evaluate(req => { req.dimensions.determinism.runs[1].output.value = 3; });
132
+ helper.check('2s. unequal repeated integration output is rejected', nondeterministic.result === 'FAIL' && nondeterministic.dimensions.determinism.matchingRuns === 1);
133
+
134
+ const nonConformant = evaluate(req => { req.sourceConformance = 'NON_CONFORMANT'; });
135
+ helper.check('2s. implementation/source mismatches fail', nonConformant.result === 'FAIL' && nonConformant.mandatoryGates.noSourceMismatch === false);
136
+
137
+ const sourceDefect = evaluate(req => { req.sourceValidity = 'SOURCE_DEFECT'; req.sourceNotes = 'The approved example appears internally wrong.'; });
138
+ helper.check('2s. source defects are reported separately without inventing a mismatch', sourceDefect.result === 'PASS' && sourceDefect.requirements[0].sourceValidity === 'SOURCE_DEFECT');
139
+
140
+ const sourceConflict = evaluate(req => {
141
+ req.sourceValidity = 'SOURCE_CONFLICT';
142
+ req.sourceNotes = 'ADR-1 and SPEC-2 disagree.';
143
+ req.additionalSources = [{ type: 'ADR', path: 'docs/adr/001.md', section: 'Contract', status: 'AUTHORITATIVE' }];
144
+ });
145
+ helper.check('2s. unresolved authoritative source conflicts fail', sourceConflict.result === 'FAIL' && sourceConflict.mandatoryGates.noSourceConflict === false && sourceConflict.dimensions.sourceTraceability.traced === 2);
146
+
147
+ const inferred = evaluate(req => { req.source = { type: 'INFERENCE', status: 'INFERRED', inferenceBasis: 'Observed public API and user-approved acceptance criteria.' }; });
148
+ helper.check('2s. inferred expectations remain explicit and traceable', inferred.result === 'PASS' && inferred.dimensions.sourceTraceability.percentage === 100);
149
+
150
+ const badNa = evaluate(req => { req.dimensions.negativeParameters = { applicable: false }; });
151
+ helper.check('2s. N/A declarations without a written reason are rejected', badNa.result === 'FAIL' && badNa.errors.some(e => /negativeParameters.*reason/.test(e)));
152
+
153
+ const goodNa = evaluate(req => { req.dimensions.negativeParameters = { applicable: false, reason: 'This parameterless behavior has no rejected partition.' }; });
154
+ helper.check('2s. reasoned N/A stays visible and does not lower applicable scores', goodNa.result === 'PASS' && goodNa.naDeclarations.length === 1);
155
+
156
+ const normalized = evaluate(req => {
157
+ req.dimensions.determinism.runs[0].output.generatedId = 'a';
158
+ req.dimensions.determinism.runs[1].output.generatedId = 'b';
159
+ req.dimensions.determinism.normalizations = [{ path: 'output.generatedId', rule: 'generated ID may vary', sourceSection: 'docs/example.md#Volatile fields' }];
160
+ });
161
+ helper.check('2s. explicitly source-backed volatile fields may be normalized', normalized.result === 'PASS');
162
+
163
+ const malformed = clone(validEvidence());
164
+ malformed.requirements = 'not-an-array';
165
+ const malformedReport = evaluateEvidence(malformed);
166
+ helper.check('2s. malformed evidence fails closed', malformedReport.result === 'FAIL' && malformedReport.errors.length > 0);
167
+
168
+ const incompleteSideEffects = evaluate(req => { delete req.dimensions.determinism.runs[1].sideEffects.externalCalls; });
169
+ helper.check('2s. incomplete deterministic side-effect evidence is rejected', incompleteSideEffects.result === 'FAIL' && incompleteSideEffects.errors.some(e => /externalCalls is required/.test(e)));
170
+
171
+ const duplicate = clone(validEvidence());
172
+ duplicate.requirements.push(clone(duplicate.requirements[0]));
173
+ const duplicateReport = evaluateEvidence(duplicate);
174
+ helper.check('2s. duplicate requirement IDs are rejected', duplicateReport.result === 'FAIL' && duplicateReport.errors.some(e => /Duplicate requirementId/.test(e)));
175
+
176
+ const tmp = fs.mkdtempSync(path.join(os.tmpdir(), 'harness-tdd-quality-'));
177
+ const inputPath = path.join(tmp, 'evidence.json');
178
+ const outputPath = path.join(tmp, 'report.json');
179
+ fs.writeFileSync(inputPath, JSON.stringify(validEvidence()), 'utf8');
180
+ const cli = spawnSync(process.execPath, [path.join(helper.root, 'tdd', 'scripts', 'quality-gate.js'), inputPath, '--output', outputPath], { encoding: 'utf8' });
181
+ helper.check('2s. CLI emits per-requirement human scores and structured JSON', cli.status === 0 && /TDD Quality Gate: PASS/.test(cli.stdout) && /REQ-001 \[integration\]: PASS 100%/.test(cli.stdout) && fs.existsSync(outputPath));
182
+ const failInputPath = path.join(tmp, 'non-conformant.json');
183
+ const failOutputPath = path.join(tmp, 'non-conformant-report.json');
184
+ const failEvidence = validEvidence();
185
+ failEvidence.requirements[0].sourceConformance = 'NON_CONFORMANT';
186
+ fs.writeFileSync(failInputPath, JSON.stringify(failEvidence), 'utf8');
187
+ const failedCli = spawnSync(process.execPath, [path.join(helper.root, 'tdd', 'scripts', 'quality-gate.js'), failInputPath, '--output', failOutputPath], { encoding: 'utf8' });
188
+ let failedReport = null;
189
+ try { failedReport = JSON.parse(fs.readFileSync(failOutputPath, 'utf8')); } catch (e) { /* asserted below */ }
190
+ helper.check(
191
+ '2s. CLI returns exit 1 and writes structured JSON for a valid FAIL report',
192
+ failedCli.status === 1 && /TDD Quality Gate: FAIL/.test(failedCli.stdout) && failedReport && failedReport.result === 'FAIL'
193
+ );
194
+ const malformedPath = path.join(tmp, 'malformed.json');
195
+ fs.writeFileSync(malformedPath, '{', 'utf8');
196
+ const malformedCli = spawnSync(process.execPath, [path.join(helper.root, 'tdd', 'scripts', 'quality-gate.js'), malformedPath], { encoding: 'utf8' });
197
+ helper.check('2s. CLI returns exit 2 for malformed JSON', malformedCli.status === 2 && /TDD Quality Gate: FAIL/.test(malformedCli.stderr));
198
+ fs.rmSync(tmp, { recursive: true, force: true });
199
+
200
+ const shippedFixture = JSON.parse(fs.readFileSync(path.join(helper.root, 'tdd', 'fixtures', 'pass-evidence.json'), 'utf8'));
201
+ helper.check('2s. shipped PASS evidence is executable', evaluateEvidence(shippedFixture).result === 'PASS');
202
+
203
+ helper.finish();