agent-hitch 0.2.6 → 0.2.8

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 (562) hide show
  1. package/README.md +30 -3
  2. package/README.zh-CN.md +14 -2
  3. package/dist/bin/hitch.js +25 -3
  4. package/dist/bin/hitch.js.map +1 -1
  5. package/dist/scripts/canary-benchmark.js +54 -0
  6. package/dist/scripts/canary-benchmark.js.map +1 -0
  7. package/dist/scripts/canary-buildkit-secrets.js +166 -0
  8. package/dist/scripts/canary-buildkit-secrets.js.map +1 -0
  9. package/dist/scripts/canary-collision-domains.js +77 -0
  10. package/dist/scripts/canary-collision-domains.js.map +1 -0
  11. package/dist/scripts/canary-docker-image-observation.js +181 -0
  12. package/dist/scripts/canary-docker-image-observation.js.map +1 -0
  13. package/dist/scripts/canary-eval-scheduler-throughput.js +47 -0
  14. package/dist/scripts/canary-eval-scheduler-throughput.js.map +1 -0
  15. package/dist/scripts/canary-gpu-hardware.js +95 -0
  16. package/dist/scripts/canary-gpu-hardware.js.map +1 -0
  17. package/dist/scripts/canary-harbor-load.js +365 -0
  18. package/dist/scripts/canary-harbor-load.js.map +1 -0
  19. package/dist/scripts/canary-packaged-harness.js +75 -0
  20. package/dist/scripts/canary-packaged-harness.js.map +1 -0
  21. package/dist/scripts/canary-resource-load.js +188 -0
  22. package/dist/scripts/canary-resource-load.js.map +1 -0
  23. package/dist/scripts/check-architecture.js +10 -5
  24. package/dist/scripts/check-architecture.js.map +1 -1
  25. package/dist/scripts/docker-reaper-crash-child.js +21 -0
  26. package/dist/scripts/docker-reaper-crash-child.js.map +1 -0
  27. package/dist/src/adapters/catalog.js +3 -1
  28. package/dist/src/adapters/catalog.js.map +1 -1
  29. package/dist/src/adapters/discovery.js +1 -1
  30. package/dist/src/adapters/discovery.js.map +1 -1
  31. package/dist/src/adapters/providers/claude.js +6 -0
  32. package/dist/src/adapters/providers/claude.js.map +1 -1
  33. package/dist/src/adapters/providers/codex-auth.js +29 -0
  34. package/dist/src/adapters/providers/codex-auth.js.map +1 -0
  35. package/dist/src/adapters/providers/codex.js +29 -6
  36. package/dist/src/adapters/providers/codex.js.map +1 -1
  37. package/dist/src/adapters/providers/deepseek.js +6 -0
  38. package/dist/src/adapters/providers/deepseek.js.map +1 -1
  39. package/dist/src/adapters/providers/model-call.js +26 -0
  40. package/dist/src/adapters/providers/model-call.js.map +1 -0
  41. package/dist/src/adapters/providers/opencode.js +6 -0
  42. package/dist/src/adapters/providers/opencode.js.map +1 -1
  43. package/dist/src/adapters/providers/pi.js +6 -0
  44. package/dist/src/adapters/providers/pi.js.map +1 -1
  45. package/dist/src/artifacts/index.js +1 -0
  46. package/dist/src/artifacts/index.js.map +1 -1
  47. package/dist/src/artifacts/preparer.js +7 -2
  48. package/dist/src/artifacts/preparer.js.map +1 -1
  49. package/dist/src/backends/harbor/agent-budget.js +31 -0
  50. package/dist/src/backends/harbor/agent-budget.js.map +1 -0
  51. package/dist/src/backends/harbor/backend.js +59 -107
  52. package/dist/src/backends/harbor/backend.js.map +1 -1
  53. package/dist/src/backends/harbor/environment-config.js +92 -0
  54. package/dist/src/backends/harbor/environment-config.js.map +1 -0
  55. package/dist/src/backends/harbor/index.js +6 -1
  56. package/dist/src/backends/harbor/index.js.map +1 -1
  57. package/dist/src/backends/harbor/model-proxy-config.js +23 -0
  58. package/dist/src/backends/harbor/model-proxy-config.js.map +1 -0
  59. package/dist/src/backends/harbor/process.js +177 -0
  60. package/dist/src/backends/harbor/process.js.map +1 -0
  61. package/dist/src/backends/harbor/regrade.js +77 -0
  62. package/dist/src/backends/harbor/regrade.js.map +1 -0
  63. package/dist/src/backends/harbor/runtime-toolchain.js +30 -0
  64. package/dist/src/backends/harbor/runtime-toolchain.js.map +1 -0
  65. package/dist/src/backends/harbor/supervisor.js +74 -0
  66. package/dist/src/backends/harbor/supervisor.js.map +1 -0
  67. package/dist/src/backends/harbor/task-resources.js +206 -0
  68. package/dist/src/backends/harbor/task-resources.js.map +1 -0
  69. package/dist/src/backends/harbor/tools.js +6 -29
  70. package/dist/src/backends/harbor/tools.js.map +1 -1
  71. package/dist/src/backends/index.js +5 -1
  72. package/dist/src/backends/index.js.map +1 -1
  73. package/dist/src/benchmarks/index.js +3 -0
  74. package/dist/src/benchmarks/index.js.map +1 -0
  75. package/dist/src/benchmarks/loader.js +271 -0
  76. package/dist/src/benchmarks/loader.js.map +1 -0
  77. package/dist/src/benchmarks/metrics.js +16 -0
  78. package/dist/src/benchmarks/metrics.js.map +1 -0
  79. package/dist/src/benchmarks/toml.js +13 -0
  80. package/dist/src/benchmarks/toml.js.map +1 -0
  81. package/dist/src/benchmarks/validation.js +242 -0
  82. package/dist/src/benchmarks/validation.js.map +1 -0
  83. package/dist/src/cli/arguments.js +6 -4
  84. package/dist/src/cli/arguments.js.map +1 -1
  85. package/dist/src/cli/commands/benchmark.js +22 -0
  86. package/dist/src/cli/commands/benchmark.js.map +1 -0
  87. package/dist/src/cli/commands/capabilities.js +18 -0
  88. package/dist/src/cli/commands/capabilities.js.map +1 -0
  89. package/dist/src/cli/commands/daemon.js +135 -8
  90. package/dist/src/cli/commands/daemon.js.map +1 -1
  91. package/dist/src/cli/commands/eval.js +228 -6
  92. package/dist/src/cli/commands/eval.js.map +1 -1
  93. package/dist/src/cli/commands/images.js +40 -0
  94. package/dist/src/cli/commands/images.js.map +1 -0
  95. package/dist/src/cli/commands/run.js +42 -17
  96. package/dist/src/cli/commands/run.js.map +1 -1
  97. package/dist/src/cli/commands/runs.js +25 -3
  98. package/dist/src/cli/commands/runs.js.map +1 -1
  99. package/dist/src/cli/commands/trajectory.js +96 -5
  100. package/dist/src/cli/commands/trajectory.js.map +1 -1
  101. package/dist/src/cli/commands/verifier.js +36 -0
  102. package/dist/src/cli/commands/verifier.js.map +1 -0
  103. package/dist/src/cli/commands/worker.js +79 -0
  104. package/dist/src/cli/commands/worker.js.map +1 -0
  105. package/dist/src/cli/main.js +10 -0
  106. package/dist/src/cli/main.js.map +1 -1
  107. package/dist/src/cli/output.js +84 -4
  108. package/dist/src/cli/output.js.map +1 -1
  109. package/dist/src/control-plane/build-admission.js +69 -0
  110. package/dist/src/control-plane/build-admission.js.map +1 -0
  111. package/dist/src/control-plane/build-records.js +6 -0
  112. package/dist/src/control-plane/build-records.js.map +1 -0
  113. package/dist/src/control-plane/collisions.js +40 -0
  114. package/dist/src/control-plane/collisions.js.map +1 -0
  115. package/dist/src/control-plane/eval-control-state.js +22 -0
  116. package/dist/src/control-plane/eval-control-state.js.map +1 -0
  117. package/dist/src/control-plane/eval-control-work.js +44 -0
  118. package/dist/src/control-plane/eval-control-work.js.map +1 -0
  119. package/dist/src/control-plane/eval-image-resolution.js +52 -0
  120. package/dist/src/control-plane/eval-image-resolution.js.map +1 -0
  121. package/dist/src/control-plane/eval-image-services.js +30 -0
  122. package/dist/src/control-plane/eval-image-services.js.map +1 -0
  123. package/dist/src/control-plane/eval-records.js +252 -0
  124. package/dist/src/control-plane/eval-records.js.map +1 -0
  125. package/dist/src/control-plane/eval-recovery.js +181 -0
  126. package/dist/src/control-plane/eval-recovery.js.map +1 -0
  127. package/dist/src/control-plane/eval-scheduler.js +450 -0
  128. package/dist/src/control-plane/eval-scheduler.js.map +1 -0
  129. package/dist/src/control-plane/index.js +20 -0
  130. package/dist/src/control-plane/index.js.map +1 -0
  131. package/dist/src/control-plane/local-worker.js +49 -0
  132. package/dist/src/control-plane/local-worker.js.map +1 -0
  133. package/dist/src/control-plane/model-capture-planning.js +12 -0
  134. package/dist/src/control-plane/model-capture-planning.js.map +1 -0
  135. package/dist/src/control-plane/remote-result-transport.js +175 -0
  136. package/dist/src/control-plane/remote-result-transport.js.map +1 -0
  137. package/dist/src/control-plane/remote-work-coordinator.js +333 -0
  138. package/dist/src/control-plane/remote-work-coordinator.js.map +1 -0
  139. package/dist/src/control-plane/remote-work-inputs.js +266 -0
  140. package/dist/src/control-plane/remote-work-inputs.js.map +1 -0
  141. package/dist/src/control-plane/remote-work-item.js +62 -0
  142. package/dist/src/control-plane/remote-work-item.js.map +1 -0
  143. package/dist/src/control-plane/remote-work-recovery.js +220 -0
  144. package/dist/src/control-plane/remote-work-recovery.js.map +1 -0
  145. package/dist/src/control-plane/remote-worker-artifacts.js +135 -0
  146. package/dist/src/control-plane/remote-worker-artifacts.js.map +1 -0
  147. package/dist/src/control-plane/remote-worker-client.js +202 -0
  148. package/dist/src/control-plane/remote-worker-client.js.map +1 -0
  149. package/dist/src/control-plane/remote-worker-configuration.js +34 -0
  150. package/dist/src/control-plane/remote-worker-configuration.js.map +1 -0
  151. package/dist/src/control-plane/remote-worker-credentials.js +58 -0
  152. package/dist/src/control-plane/remote-worker-credentials.js.map +1 -0
  153. package/dist/src/control-plane/remote-worker-protocol.js +489 -0
  154. package/dist/src/control-plane/remote-worker-protocol.js.map +1 -0
  155. package/dist/src/control-plane/remote-worker-runner.js +214 -0
  156. package/dist/src/control-plane/remote-worker-runner.js.map +1 -0
  157. package/dist/src/control-plane/remote-workers.js +377 -0
  158. package/dist/src/control-plane/remote-workers.js.map +1 -0
  159. package/dist/src/control-plane/rerun-scheduler.js +448 -0
  160. package/dist/src/control-plane/rerun-scheduler.js.map +1 -0
  161. package/dist/src/control-plane/rerun-submission.js +62 -0
  162. package/dist/src/control-plane/rerun-submission.js.map +1 -0
  163. package/dist/src/control-plane/resources.js +182 -0
  164. package/dist/src/control-plane/resources.js.map +1 -0
  165. package/dist/src/control-plane/scheduler-eval-entry.js +32 -0
  166. package/dist/src/control-plane/scheduler-eval-entry.js.map +1 -0
  167. package/dist/src/control-plane/synthetic-result.js +21 -0
  168. package/dist/src/control-plane/synthetic-result.js.map +1 -0
  169. package/dist/src/control-plane/work-admission.js +23 -0
  170. package/dist/src/control-plane/work-admission.js.map +1 -0
  171. package/dist/src/control-plane/work-dispatcher.js +237 -0
  172. package/dist/src/control-plane/work-dispatcher.js.map +1 -0
  173. package/dist/src/controller-runtime/hash.js +19 -3
  174. package/dist/src/controller-runtime/hash.js.map +1 -1
  175. package/dist/src/controller-runtime/index.js +1 -1
  176. package/dist/src/controller-runtime/index.js.map +1 -1
  177. package/dist/src/controller-runtime/store.js +15 -6
  178. package/dist/src/controller-runtime/store.js.map +1 -1
  179. package/dist/src/daemon/auth.js +5 -0
  180. package/dist/src/daemon/auth.js.map +1 -1
  181. package/dist/src/daemon/client.js.map +1 -1
  182. package/dist/src/daemon/event-stream.js +46 -0
  183. package/dist/src/daemon/event-stream.js.map +1 -0
  184. package/dist/src/daemon/health.js +66 -0
  185. package/dist/src/daemon/health.js.map +1 -0
  186. package/dist/src/daemon/launcher.js +28 -1
  187. package/dist/src/daemon/launcher.js.map +1 -1
  188. package/dist/src/daemon/logging.js +23 -0
  189. package/dist/src/daemon/logging.js.map +1 -0
  190. package/dist/src/daemon/scheduler.js +26 -5
  191. package/dist/src/daemon/scheduler.js.map +1 -1
  192. package/dist/src/daemon/server.js +255 -98
  193. package/dist/src/daemon/server.js.map +1 -1
  194. package/dist/src/daemon/telemetry.js +163 -0
  195. package/dist/src/daemon/telemetry.js.map +1 -0
  196. package/dist/src/daemon/worker-routes.js +219 -0
  197. package/dist/src/daemon/worker-routes.js.map +1 -0
  198. package/dist/src/domain/benchmarks.js +2 -0
  199. package/dist/src/domain/benchmarks.js.map +1 -0
  200. package/dist/src/domain/bundles.js +2 -0
  201. package/dist/src/domain/bundles.js.map +1 -0
  202. package/dist/src/domain/control-plane.js +2 -0
  203. package/dist/src/domain/control-plane.js.map +1 -0
  204. package/dist/src/domain/execution-evidence.js +2 -0
  205. package/dist/src/domain/execution-evidence.js.map +1 -0
  206. package/dist/src/domain/execution-plan.js +2 -0
  207. package/dist/src/domain/execution-plan.js.map +1 -0
  208. package/dist/src/domain/images.js +2 -0
  209. package/dist/src/domain/images.js.map +1 -0
  210. package/dist/src/domain/index.js +13 -0
  211. package/dist/src/domain/index.js.map +1 -1
  212. package/dist/src/domain/interactions.js +2 -0
  213. package/dist/src/domain/interactions.js.map +1 -0
  214. package/dist/src/domain/providers.js +2 -0
  215. package/dist/src/domain/providers.js.map +1 -0
  216. package/dist/src/domain/resources.js +2 -0
  217. package/dist/src/domain/resources.js.map +1 -0
  218. package/dist/src/domain/runs.js.map +1 -1
  219. package/dist/src/domain/validation.js +11 -4
  220. package/dist/src/domain/validation.js.map +1 -1
  221. package/dist/src/domain/verifier-evidence-validation.js +214 -0
  222. package/dist/src/domain/verifier-evidence-validation.js.map +1 -0
  223. package/dist/src/domain/verifier-evidence.js +3 -0
  224. package/dist/src/domain/verifier-evidence.js.map +1 -0
  225. package/dist/src/domain/verifier-score-contract.js +232 -0
  226. package/dist/src/domain/verifier-score-contract.js.map +1 -0
  227. package/dist/src/domain/workers.js +2 -0
  228. package/dist/src/domain/workers.js.map +1 -0
  229. package/dist/src/evals/benchmark-adapter-manifest.js +219 -0
  230. package/dist/src/evals/benchmark-adapter-manifest.js.map +1 -0
  231. package/dist/src/evals/benchmark-candidate.js +36 -0
  232. package/dist/src/evals/benchmark-candidate.js.map +1 -0
  233. package/dist/src/evals/benchmark-run.js +166 -0
  234. package/dist/src/evals/benchmark-run.js.map +1 -0
  235. package/dist/src/evals/collect-only-rerun.js +183 -0
  236. package/dist/src/evals/collect-only-rerun.js.map +1 -0
  237. package/dist/src/evals/directory.js +30 -0
  238. package/dist/src/evals/directory.js.map +1 -0
  239. package/dist/src/evals/docker-engine-stats.js +89 -0
  240. package/dist/src/evals/docker-engine-stats.js.map +1 -0
  241. package/dist/src/evals/docker-ownership.js +51 -0
  242. package/dist/src/evals/docker-ownership.js.map +1 -0
  243. package/dist/src/evals/docker-reaper.js +144 -0
  244. package/dist/src/evals/docker-reaper.js.map +1 -0
  245. package/dist/src/evals/docker-resource-observer.js +207 -0
  246. package/dist/src/evals/docker-resource-observer.js.map +1 -0
  247. package/dist/src/evals/duration-estimator.js +126 -0
  248. package/dist/src/evals/duration-estimator.js.map +1 -0
  249. package/dist/src/evals/environment-image-planning.js +164 -0
  250. package/dist/src/evals/environment-image-planning.js.map +1 -0
  251. package/dist/src/evals/eval-artifact-planning.js +56 -0
  252. package/dist/src/evals/eval-artifact-planning.js.map +1 -0
  253. package/dist/src/evals/eval-finalization.js +10 -0
  254. package/dist/src/evals/eval-finalization.js.map +1 -0
  255. package/dist/src/evals/eval-lifecycle-events.js +29 -0
  256. package/dist/src/evals/eval-lifecycle-events.js.map +1 -0
  257. package/dist/src/evals/eval-logical-plan.js +51 -0
  258. package/dist/src/evals/eval-logical-plan.js.map +1 -0
  259. package/dist/src/evals/eval-result-builder.js +81 -0
  260. package/dist/src/evals/eval-result-builder.js.map +1 -0
  261. package/dist/src/evals/events.js +53 -5
  262. package/dist/src/evals/events.js.map +1 -1
  263. package/dist/src/evals/evolution-baseline.js +191 -0
  264. package/dist/src/evals/evolution-baseline.js.map +1 -0
  265. package/dist/src/evals/execution-evidence.js +116 -0
  266. package/dist/src/evals/execution-evidence.js.map +1 -0
  267. package/dist/src/evals/execution-leases.js +342 -0
  268. package/dist/src/evals/execution-leases.js.map +1 -0
  269. package/dist/src/evals/execution-plan-artifacts.js +72 -0
  270. package/dist/src/evals/execution-plan-artifacts.js.map +1 -0
  271. package/dist/src/evals/execution-plan-images.js +69 -0
  272. package/dist/src/evals/execution-plan-images.js.map +1 -0
  273. package/dist/src/evals/execution-plan-resources.js +174 -0
  274. package/dist/src/evals/execution-plan-resources.js.map +1 -0
  275. package/dist/src/evals/execution-plan.js +451 -0
  276. package/dist/src/evals/execution-plan.js.map +1 -0
  277. package/dist/src/evals/failure-classifier.js +61 -0
  278. package/dist/src/evals/failure-classifier.js.map +1 -0
  279. package/dist/src/evals/harbor-artifact-builder-image.js +99 -0
  280. package/dist/src/evals/harbor-artifact-builder-image.js.map +1 -0
  281. package/dist/src/evals/harbor-artifact-builder.js +358 -0
  282. package/dist/src/evals/harbor-artifact-builder.js.map +1 -0
  283. package/dist/src/evals/harbor-bridge-error.js +58 -3
  284. package/dist/src/evals/harbor-bridge-error.js.map +1 -1
  285. package/dist/src/evals/harbor-node-runtime.js +146 -0
  286. package/dist/src/evals/harbor-node-runtime.js.map +1 -0
  287. package/dist/src/evals/harbor-phase-timings.js +60 -0
  288. package/dist/src/evals/harbor-phase-timings.js.map +1 -0
  289. package/dist/src/evals/index.js +40 -2
  290. package/dist/src/evals/index.js.map +1 -1
  291. package/dist/src/evals/infrastructure-retry.js +208 -84
  292. package/dist/src/evals/infrastructure-retry.js.map +1 -1
  293. package/dist/src/evals/interaction-capture-import.js +49 -0
  294. package/dist/src/evals/interaction-capture-import.js.map +1 -0
  295. package/dist/src/evals/local-docker-provider.js +372 -0
  296. package/dist/src/evals/local-docker-provider.js.map +1 -0
  297. package/dist/src/evals/local-eval-planning.js +80 -0
  298. package/dist/src/evals/local-eval-planning.js.map +1 -0
  299. package/dist/src/evals/model-capture-plan.js +22 -0
  300. package/dist/src/evals/model-capture-plan.js.map +1 -0
  301. package/dist/src/evals/model-capture-runtime.js +114 -0
  302. package/dist/src/evals/model-capture-runtime.js.map +1 -0
  303. package/dist/src/evals/model-proxy-runtime-state.js +53 -0
  304. package/dist/src/evals/model-proxy-runtime-state.js.map +1 -0
  305. package/dist/src/evals/native-phase-evidence.js +440 -0
  306. package/dist/src/evals/native-phase-evidence.js.map +1 -0
  307. package/dist/src/evals/physical-retry-work.js +31 -0
  308. package/dist/src/evals/physical-retry-work.js.map +1 -0
  309. package/dist/src/evals/planned-execution-support.js +66 -0
  310. package/dist/src/evals/planned-execution-support.js.map +1 -0
  311. package/dist/src/evals/planned-execution.js +469 -0
  312. package/dist/src/evals/planned-execution.js.map +1 -0
  313. package/dist/src/evals/planned-progress-publisher.js +62 -0
  314. package/dist/src/evals/planned-progress-publisher.js.map +1 -0
  315. package/dist/src/evals/planned-remote-execution.js +40 -0
  316. package/dist/src/evals/planned-remote-execution.js.map +1 -0
  317. package/dist/src/evals/planned-retry-execution.js +174 -0
  318. package/dist/src/evals/planned-retry-execution.js.map +1 -0
  319. package/dist/src/evals/planned-retry-lifecycle.js +188 -0
  320. package/dist/src/evals/planned-retry-lifecycle.js.map +1 -0
  321. package/dist/src/evals/preparation-rerun.js +225 -0
  322. package/dist/src/evals/preparation-rerun.js.map +1 -0
  323. package/dist/src/evals/prepared-harness.js +76 -0
  324. package/dist/src/evals/prepared-harness.js.map +1 -0
  325. package/dist/src/evals/progress.js +52 -12
  326. package/dist/src/evals/progress.js.map +1 -1
  327. package/dist/src/evals/records.js +20 -3
  328. package/dist/src/evals/records.js.map +1 -1
  329. package/dist/src/evals/recovery.js +242 -0
  330. package/dist/src/evals/recovery.js.map +1 -0
  331. package/dist/src/evals/regrade-evidence.js +77 -0
  332. package/dist/src/evals/regrade-evidence.js.map +1 -0
  333. package/dist/src/evals/remote-infrastructure-retry.js +165 -0
  334. package/dist/src/evals/remote-infrastructure-retry.js.map +1 -0
  335. package/dist/src/evals/request.js +21 -3
  336. package/dist/src/evals/request.js.map +1 -1
  337. package/dist/src/evals/rerun-inputs.js +108 -0
  338. package/dist/src/evals/rerun-inputs.js.map +1 -0
  339. package/dist/src/evals/rerun-slots.js +4 -3
  340. package/dist/src/evals/rerun-slots.js.map +1 -1
  341. package/dist/src/evals/rerun-types.js +45 -0
  342. package/dist/src/evals/rerun-types.js.map +1 -0
  343. package/dist/src/evals/rerun.js +133 -88
  344. package/dist/src/evals/rerun.js.map +1 -1
  345. package/dist/src/evals/result-helpers.js +78 -0
  346. package/dist/src/evals/result-helpers.js.map +1 -0
  347. package/dist/src/evals/resume-state.js +36 -0
  348. package/dist/src/evals/resume-state.js.map +1 -0
  349. package/dist/src/evals/retry-backoff.js +20 -0
  350. package/dist/src/evals/retry-backoff.js.map +1 -0
  351. package/dist/src/evals/retry-state.js +195 -0
  352. package/dist/src/evals/retry-state.js.map +1 -0
  353. package/dist/src/evals/scheduler-metrics.js +95 -0
  354. package/dist/src/evals/scheduler-metrics.js.map +1 -0
  355. package/dist/src/evals/scheduler-trace-replay.js +107 -0
  356. package/dist/src/evals/scheduler-trace-replay.js.map +1 -0
  357. package/dist/src/evals/service-types.js +2 -0
  358. package/dist/src/evals/service-types.js.map +1 -0
  359. package/dist/src/evals/service.js +331 -347
  360. package/dist/src/evals/service.js.map +1 -1
  361. package/dist/src/evals/task-resources.js +209 -0
  362. package/dist/src/evals/task-resources.js.map +1 -0
  363. package/dist/src/evals/trial-environment-evidence.js +87 -0
  364. package/dist/src/evals/trial-environment-evidence.js.map +1 -0
  365. package/dist/src/evals/trial-execution-evidence.js +13 -0
  366. package/dist/src/evals/trial-execution-evidence.js.map +1 -0
  367. package/dist/src/evals/trial-import-identity.js +31 -0
  368. package/dist/src/evals/trial-import-identity.js.map +1 -0
  369. package/dist/src/evals/trial-import.js +183 -188
  370. package/dist/src/evals/trial-import.js.map +1 -1
  371. package/dist/src/evals/trial-publication-recovery.js +144 -0
  372. package/dist/src/evals/trial-publication-recovery.js.map +1 -0
  373. package/dist/src/evals/trial-publication.js +25 -0
  374. package/dist/src/evals/trial-publication.js.map +1 -0
  375. package/dist/src/evals/trial-reference-validation.js +49 -0
  376. package/dist/src/evals/trial-reference-validation.js.map +1 -0
  377. package/dist/src/evals/verifier-artifacts.js +317 -0
  378. package/dist/src/evals/verifier-artifacts.js.map +1 -0
  379. package/dist/src/evals/verifier-eligibility.js +25 -0
  380. package/dist/src/evals/verifier-eligibility.js.map +1 -0
  381. package/dist/src/evals/verifier-only-rerun.js +210 -0
  382. package/dist/src/evals/verifier-only-rerun.js.map +1 -0
  383. package/dist/src/evals/verifier-runtime.js +61 -0
  384. package/dist/src/evals/verifier-runtime.js.map +1 -0
  385. package/dist/src/evals/verifier-score-artifacts.js +119 -0
  386. package/dist/src/evals/verifier-score-artifacts.js.map +1 -0
  387. package/dist/src/evals/work-item-artifacts.js +23 -0
  388. package/dist/src/evals/work-item-artifacts.js.map +1 -0
  389. package/dist/src/foundation/config.js +17 -0
  390. package/dist/src/foundation/config.js.map +1 -1
  391. package/dist/src/foundation/contained-file.js +85 -0
  392. package/dist/src/foundation/contained-file.js.map +1 -0
  393. package/dist/src/foundation/credential-redaction.js +162 -0
  394. package/dist/src/foundation/credential-redaction.js.map +1 -0
  395. package/dist/src/foundation/environment-image-references.js +59 -0
  396. package/dist/src/foundation/environment-image-references.js.map +1 -0
  397. package/dist/src/foundation/executable.js +45 -5
  398. package/dist/src/foundation/executable.js.map +1 -1
  399. package/dist/src/foundation/fs.js +2 -1
  400. package/dist/src/foundation/fs.js.map +1 -1
  401. package/dist/src/foundation/index.js +6 -2
  402. package/dist/src/foundation/index.js.map +1 -1
  403. package/dist/src/foundation/process-identity.js +69 -0
  404. package/dist/src/foundation/process-identity.js.map +1 -0
  405. package/dist/src/images/context.js +47 -0
  406. package/dist/src/images/context.js.map +1 -0
  407. package/dist/src/images/docker-buildkit.js +137 -0
  408. package/dist/src/images/docker-buildkit.js.map +1 -0
  409. package/dist/src/images/dockerfile.js +67 -0
  410. package/dist/src/images/dockerfile.js.map +1 -0
  411. package/dist/src/images/gc.js +261 -0
  412. package/dist/src/images/gc.js.map +1 -0
  413. package/dist/src/images/index.js +10 -0
  414. package/dist/src/images/index.js.map +1 -0
  415. package/dist/src/images/manifest.js +109 -0
  416. package/dist/src/images/manifest.js.map +1 -0
  417. package/dist/src/images/ownership.js +5 -0
  418. package/dist/src/images/ownership.js.map +1 -0
  419. package/dist/src/images/records.js +32 -0
  420. package/dist/src/images/records.js.map +1 -0
  421. package/dist/src/images/registry.js +154 -0
  422. package/dist/src/images/registry.js.map +1 -0
  423. package/dist/src/images/service.js +224 -0
  424. package/dist/src/images/service.js.map +1 -0
  425. package/dist/src/model-access/capture.js +255 -0
  426. package/dist/src/model-access/capture.js.map +1 -0
  427. package/dist/src/model-access/index.js +5 -0
  428. package/dist/src/model-access/index.js.map +1 -0
  429. package/dist/src/model-access/policy.js +49 -0
  430. package/dist/src/model-access/policy.js.map +1 -0
  431. package/dist/src/model-access/proxy.js +336 -0
  432. package/dist/src/model-access/proxy.js.map +1 -0
  433. package/dist/src/model-access/records.js +167 -0
  434. package/dist/src/model-access/records.js.map +1 -0
  435. package/dist/src/runs/adapter-process.js +12 -0
  436. package/dist/src/runs/adapter-process.js.map +1 -0
  437. package/dist/src/runs/bundle.js +450 -0
  438. package/dist/src/runs/bundle.js.map +1 -0
  439. package/dist/src/runs/events.js +12 -1
  440. package/dist/src/runs/events.js.map +1 -1
  441. package/dist/src/runs/executor.js +46 -45
  442. package/dist/src/runs/executor.js.map +1 -1
  443. package/dist/src/runs/finalizer.js +17 -0
  444. package/dist/src/runs/finalizer.js.map +1 -1
  445. package/dist/src/runs/index.js +8 -1
  446. package/dist/src/runs/index.js.map +1 -1
  447. package/dist/src/runs/manifest.js +5 -5
  448. package/dist/src/runs/manifest.js.map +1 -1
  449. package/dist/src/runs/phase-bundle.js +86 -0
  450. package/dist/src/runs/phase-bundle.js.map +1 -0
  451. package/dist/src/runs/phase-cancellation.js +107 -0
  452. package/dist/src/runs/phase-cancellation.js.map +1 -0
  453. package/dist/src/runs/phase-group.js +113 -0
  454. package/dist/src/runs/phase-group.js.map +1 -0
  455. package/dist/src/runs/query.js +1 -1
  456. package/dist/src/runs/query.js.map +1 -1
  457. package/dist/src/runs/records.js +3 -0
  458. package/dist/src/runs/records.js.map +1 -1
  459. package/dist/src/runs/request.js +10 -3
  460. package/dist/src/runs/request.js.map +1 -1
  461. package/dist/src/runs/training-candidate.js +161 -0
  462. package/dist/src/runs/training-candidate.js.map +1 -0
  463. package/dist/src/runs/verifier-evidence-redaction.js +61 -0
  464. package/dist/src/runs/verifier-evidence-redaction.js.map +1 -0
  465. package/dist/src/runs/verifier-evidence.js +459 -0
  466. package/dist/src/runs/verifier-evidence.js.map +1 -0
  467. package/dist/src/runs/verifier-structured-evidence.js +83 -0
  468. package/dist/src/runs/verifier-structured-evidence.js.map +1 -0
  469. package/dist/src/trajectories/analysis.js +288 -0
  470. package/dist/src/trajectories/analysis.js.map +1 -0
  471. package/dist/src/trajectories/chunk-projection.js +148 -0
  472. package/dist/src/trajectories/chunk-projection.js.map +1 -0
  473. package/dist/src/trajectories/content-projection.js +389 -0
  474. package/dist/src/trajectories/content-projection.js.map +1 -0
  475. package/dist/src/trajectories/dsh-chunk-contract.js +176 -0
  476. package/dist/src/trajectories/dsh-chunk-contract.js.map +1 -0
  477. package/dist/src/trajectories/dsh-contract.js +284 -0
  478. package/dist/src/trajectories/dsh-contract.js.map +1 -0
  479. package/dist/src/trajectories/events-chunk-drill.js +95 -0
  480. package/dist/src/trajectories/events-chunk-drill.js.map +1 -0
  481. package/dist/src/trajectories/events-page.js +367 -0
  482. package/dist/src/trajectories/events-page.js.map +1 -0
  483. package/dist/src/trajectories/index.js +3 -0
  484. package/dist/src/trajectories/index.js.map +1 -1
  485. package/dist/src/trajectories/provider-capture.js +15 -24
  486. package/dist/src/trajectories/provider-capture.js.map +1 -1
  487. package/dist/src/trajectories/providers/deepseek.js +3 -3
  488. package/dist/src/trajectories/providers/deepseek.js.map +1 -1
  489. package/dist/src/trajectories/request-attempt.js +47 -0
  490. package/dist/src/trajectories/request-attempt.js.map +1 -0
  491. package/dist/src/trajectories/stream-reader.js +311 -0
  492. package/dist/src/trajectories/stream-reader.js.map +1 -0
  493. package/dist/src/trajectories/surface-fold.js +126 -0
  494. package/dist/src/trajectories/surface-fold.js.map +1 -0
  495. package/dist/src/workers/index.js +3 -0
  496. package/dist/src/workers/index.js.map +1 -0
  497. package/dist/src/workers/remote-harbor-work-spec.js +97 -0
  498. package/dist/src/workers/remote-harbor-work-spec.js.map +1 -0
  499. package/dist/src/workers/remote-harbor-worker.js +223 -0
  500. package/dist/src/workers/remote-harbor-worker.js.map +1 -0
  501. package/docs/schemas/benchmark-hook-request-v1.schema.json +65 -0
  502. package/docs/schemas/benchmark-hook-response-v1.schema.json +75 -0
  503. package/docs/schemas/benchmark-hook-v1.schema.json +33 -0
  504. package/docs/schemas/benchmark-lock-v1.schema.json +351 -0
  505. package/docs/schemas/benchmark-metric-v1.schema.json +37 -0
  506. package/docs/schemas/benchmark-package-v1.schema.json +167 -0
  507. package/docs/schemas/benchmark-phase-group.schema.json +41 -0
  508. package/docs/schemas/benchmark-profile-v1.schema.json +141 -0
  509. package/docs/schemas/benchmark-task-v1.schema.json +384 -0
  510. package/docs/schemas/benchmark-tool-result-v1.schema.json +30 -0
  511. package/docs/schemas/environment-build-index.schema.json +13 -0
  512. package/docs/schemas/environment-build-record.schema.json +25 -0
  513. package/docs/schemas/environment-image-manifest.schema.json +62 -0
  514. package/docs/schemas/error.schema.json +22 -0
  515. package/docs/schemas/eval-control.schema.json +36 -0
  516. package/docs/schemas/eval-execution-plan.schema.json +216 -0
  517. package/docs/schemas/eval-progress.schema.json +1 -15
  518. package/docs/schemas/eval-rerun-result.schema.json +58 -0
  519. package/docs/schemas/eval-rerun-state.schema.json +61 -0
  520. package/docs/schemas/eval-rerun-submission.schema.json +60 -0
  521. package/docs/schemas/eval-result.schema.json +32 -13
  522. package/docs/schemas/eval-submission-input.schema.json +14 -0
  523. package/docs/schemas/eval-submission.schema.json +68 -0
  524. package/docs/schemas/eval-trial-publication.schema.json +15 -0
  525. package/docs/schemas/eval-trial-reference.schema.json +129 -0
  526. package/docs/schemas/execution-evidence.schema.json +96 -0
  527. package/docs/schemas/execution-lease.schema.json +50 -0
  528. package/docs/schemas/execution-provider-status.schema.json +65 -0
  529. package/docs/schemas/execution-worker.schema.json +53 -0
  530. package/docs/schemas/harbor-process-exit.schema.json +14 -0
  531. package/docs/schemas/interaction-capture-ref.schema.json +45 -0
  532. package/docs/schemas/local-provider-execution.schema.json +33 -0
  533. package/docs/schemas/model-interaction.schema.json +41 -0
  534. package/docs/schemas/regrade-assessment-reference.schema.json +11 -0
  535. package/docs/schemas/remote-result-envelope.schema.json +32 -0
  536. package/docs/schemas/remote-tree-envelope.schema.json +25 -0
  537. package/docs/schemas/remote-work-offer.schema.json +58 -0
  538. package/docs/schemas/remote-work-receipt.schema.json +50 -0
  539. package/docs/schemas/remote-worker-artifact.schema.json +17 -0
  540. package/docs/schemas/remote-worker-event.schema.json +18 -0
  541. package/docs/schemas/remote-worker-heartbeat.schema.json +37 -0
  542. package/docs/schemas/remote-worker-registration.schema.json +52 -0
  543. package/docs/schemas/result-bundle-index.schema.json +133 -0
  544. package/docs/schemas/run-context.schema.json +15 -0
  545. package/docs/schemas/run-environment-images.schema.json +39 -0
  546. package/docs/schemas/run-manifest.schema.json +4 -0
  547. package/docs/schemas/run-request.schema.json +4 -0
  548. package/docs/schemas/training-data-candidate.schema.json +46 -0
  549. package/docs/schemas/trajectory-analysis.schema.json +258 -0
  550. package/docs/schemas/trajectory-events-page.schema.json +40 -0
  551. package/docs/schemas/verifier-assessment.schema.json +77 -0
  552. package/docs/schemas/verifier-evidence.schema.json +276 -0
  553. package/integrations/harbor/hitch_benchmark.py +252 -0
  554. package/integrations/harbor/hitch_candidate_recycle.py +262 -0
  555. package/integrations/harbor/hitch_harbor_agent.py +698 -568
  556. package/integrations/harbor/hitch_harbor_environment.py +469 -0
  557. package/integrations/harbor/hitch_harbor_task_resources.py +275 -0
  558. package/integrations/harbor/hitch_harbor_verifier.py +83 -2
  559. package/integrations/harbor/hitch_phase_supervisor.py +452 -0
  560. package/integrations/harbor/hitch_tool_client.mjs +91 -0
  561. package/integrations/model-call/cli.js +57 -0
  562. package/package.json +30 -6
@@ -0,0 +1 @@
1
+ {"version":3,"file":"retry-backoff.js","sourceRoot":"","sources":["../../../src/evals/retry-backoff.ts"],"names":[],"mappings":"AAAA,OAAO,EAAE,UAAU,EAAE,MAAM,aAAa,CAAC;AAEzC,MAAM,oBAAoB,GAAG,MAAM,CAAC;AAEpC;;qEAEqE;AACrE,MAAM,UAAU,cAAc,CAAC,YAAoB,EAAE,UAAkB,EAAE,IAAY;IACnF,IAAI,CAAC,MAAM,CAAC,aAAa,CAAC,YAAY,CAAC,IAAI,YAAY,GAAG,CAAC;QAAE,MAAM,IAAI,SAAS,CAAC,yCAAyC,CAAC,CAAC;IAC5H,IAAI,CAAC,MAAM,CAAC,aAAa,CAAC,UAAU,CAAC,IAAI,UAAU,GAAG,CAAC;QAAE,MAAM,IAAI,SAAS,CAAC,uCAAuC,CAAC,CAAC;IACtH,IAAI,CAAC,IAAI;QAAE,MAAM,IAAI,SAAS,CAAC,8CAA8C,CAAC,CAAC;IAC/E,IAAI,YAAY,KAAK,CAAC;QAAE,OAAO,CAAC,CAAC;IACjC,MAAM,IAAI,GAAG,IAAI,CAAC,GAAG,CAAC,YAAY,EAAE,KAAK,CAAC,CAAC;IAC3C,MAAM,GAAG,GAAG,IAAI,CAAC,GAAG,CAAC,oBAAoB,EAAE,IAAI,GAAG,CAAC,IAAI,IAAI,CAAC,GAAG,CAAC,UAAU,GAAG,CAAC,EAAE,EAAE,CAAC,CAAC,CAAC;IACrF,MAAM,OAAO,GAAG,UAAU,CAAC,QAAQ,CAAC,CAAC,MAAM,CAAC,GAAG,IAAI,KAAK,UAAU,EAAE,CAAC,CAAC,MAAM,EAAE,CAAC,YAAY,CAAC,CAAC,CAAC,CAAC;IAC/F,OAAO,OAAO,GAAG,CAAC,GAAG,GAAG,CAAC,CAAC,CAAC;AAC7B,CAAC"}
@@ -0,0 +1,195 @@
1
+ import path from "node:path";
2
+ import { atomicWriteJSON, ensureDir, readJSON, sha256JSON, withFileLock } from "../foundation/index.js";
3
+ import { classifyTrialFailure, parseFailureClassification, physicalRetryAllowed } from "./failure-classifier.js";
4
+ import { physicalRetryWorkItem, physicalRetryWorkItemForTriggerIds } from "./physical-retry-work.js";
5
+ export async function readEvalRetryState(evalDirectory, evalId) {
6
+ const value = await readJSON(path.join(evalDirectory, "retry-state.json"), null);
7
+ return value === null ? null : parseEvalRetryState(value, evalId);
8
+ }
9
+ export async function ensurePhysicalRetryDecision(input) {
10
+ const classification = classifyTrialFailure(input.trigger);
11
+ if (!classification || !physicalRetryAllowed(classification))
12
+ throw new TypeError("physical retry decision requires a retryable classification");
13
+ const retryWork = physicalRetryWorkItem(input.item, input.retryIndex, [input.trigger]);
14
+ return ensureDecision({
15
+ ...input,
16
+ classification,
17
+ disposition: "physical-retry",
18
+ retryWorkId: retryWork.work_id,
19
+ state: "planned",
20
+ });
21
+ }
22
+ export async function ensureTerminalRetryDecision(input) {
23
+ const classification = classifyTrialFailure(input.trigger);
24
+ if (!classification)
25
+ return null;
26
+ const disposition = dispositionFor(classification);
27
+ return ensureDecision({
28
+ ...input,
29
+ classification,
30
+ disposition,
31
+ state: input.exhausted && physicalRetryAllowed(classification) ? "exhausted" : "skipped",
32
+ });
33
+ }
34
+ export async function transitionRetryDecision(input) {
35
+ return mutateState(input.evalDirectory, input.evalId, (current) => {
36
+ const index = current.decisions.findIndex((decision) => decision.decision_id === input.decisionId);
37
+ if (index < 0)
38
+ throw new TypeError(`retry decision does not exist: ${input.decisionId}`);
39
+ const existing = current.decisions[index];
40
+ if (existing.state === input.state)
41
+ return { current, value: existing, changed: false };
42
+ if (!allowedTransition(existing.state, input.state))
43
+ throw new TypeError(`retry decision state transition is invalid: ${existing.state} -> ${input.state}`);
44
+ const updated = { ...existing, state: input.state, updated_at: validTimestamp(input.now ?? new Date().toISOString(), "retry decision updated_at") };
45
+ const decisions = [...current.decisions];
46
+ decisions[index] = updated;
47
+ return { current: { ...current, decisions: canonicalDecisions(decisions) }, value: updated, changed: true };
48
+ });
49
+ }
50
+ export function resolveRetryWork(plan, state, workId) {
51
+ if (!state)
52
+ return null;
53
+ const decisions = state.decisions.filter((decision) => decision.retry_work_id === workId);
54
+ if (decisions.length === 0)
55
+ return null;
56
+ const first = decisions[0];
57
+ if (decisions.some((decision) => decision.slot_id !== first.slot_id || decision.retry_index !== first.retry_index)) {
58
+ throw new TypeError(`retry work identity has conflicting decisions: ${workId}`);
59
+ }
60
+ const origin = plan.work_items.find((item) => item.slots.includes(first.slot_id));
61
+ if (!origin)
62
+ throw new TypeError(`retry decision slot is absent from execution plan: ${first.slot_id}`);
63
+ const item = physicalRetryWorkItemForTriggerIds(origin, first.retry_index, decisions.map((decision) => decision.trigger_trial_id));
64
+ if (item.work_id !== workId)
65
+ throw new TypeError(`retry decision work identity is invalid: ${workId}`);
66
+ return { item, decisions };
67
+ }
68
+ async function ensureDecision(input) {
69
+ const slotId = input.item.slots[0];
70
+ if (!slotId || input.item.slots.length !== 1 || input.item.task_ids.length !== 1)
71
+ throw new TypeError("retry decisions require one task slot work item");
72
+ if (!Number.isSafeInteger(input.retryIndex) || input.retryIndex < 1)
73
+ throw new TypeError("retry decision index is invalid");
74
+ const timestamp = validTimestamp(input.now ?? new Date().toISOString(), "retry decision created_at");
75
+ const digest = sha256JSON({ eval_id: input.evalId, slot_id: slotId, trigger_trial_id: input.trigger.trial_id, retry_index: input.retryIndex });
76
+ const decision = {
77
+ decision_id: `retry_decision_${digest.slice("sha256:".length, "sha256:".length + 32)}`,
78
+ slot_id: slotId,
79
+ trigger_trial_id: input.trigger.trial_id,
80
+ ...(input.trigger.run_id ? { trigger_run_id: input.trigger.run_id } : {}),
81
+ retry_index: input.retryIndex,
82
+ classification: input.classification,
83
+ disposition: input.disposition,
84
+ ...(input.retryWorkId ? { retry_work_id: input.retryWorkId } : {}),
85
+ ...(input.notBefore ? { not_before: validTimestamp(input.notBefore, "retry decision not_before") } : {}),
86
+ state: input.state,
87
+ created_at: timestamp,
88
+ updated_at: timestamp,
89
+ };
90
+ return mutateState(input.evalDirectory, input.evalId, (current) => {
91
+ const existing = current.decisions.find((entry) => entry.decision_id === decision.decision_id);
92
+ if (existing) {
93
+ if (JSON.stringify(immutableDecision(existing)) !== JSON.stringify(immutableDecision(decision))) {
94
+ throw new TypeError(`retry decision identity conflict: ${decision.decision_id}`);
95
+ }
96
+ return { current, value: existing, changed: false };
97
+ }
98
+ return { current: { ...current, decisions: canonicalDecisions([...current.decisions, decision]) }, value: decision, changed: true };
99
+ });
100
+ }
101
+ async function mutateState(evalDirectory, evalId, operation) {
102
+ await ensureDir(evalDirectory);
103
+ return withFileLock(path.join(evalDirectory, ".locks"), "retry-state", async () => {
104
+ const existing = await readEvalRetryState(evalDirectory, evalId);
105
+ const now = new Date().toISOString();
106
+ const state = existing ?? { schema_version: "1", eval_id: evalId, generation: 0, decisions: [], updated_at: now };
107
+ const result = operation(state);
108
+ if (result.changed) {
109
+ const next = parseEvalRetryState({ ...result.current, generation: state.generation + 1, updated_at: now }, evalId);
110
+ await atomicWriteJSON(path.join(evalDirectory, "retry-state.json"), next);
111
+ }
112
+ return result.value;
113
+ }, { timeoutCode: "eval_retry_state_locked", timeoutExitCode: 12 });
114
+ }
115
+ export function parseEvalRetryState(value, expectedEvalId) {
116
+ if (!value || typeof value !== "object" || Array.isArray(value))
117
+ throw new TypeError("eval retry state must be an object");
118
+ const record = value;
119
+ if (Object.keys(record).some((key) => !["schema_version", "eval_id", "generation", "decisions", "updated_at"].includes(key))
120
+ || record.schema_version !== "1" || typeof record.eval_id !== "string" || !/^eval_[a-f0-9]{32}$/.test(record.eval_id)
121
+ || expectedEvalId !== undefined && record.eval_id !== expectedEvalId
122
+ || !Number.isSafeInteger(record.generation) || record.generation < 0 || !Array.isArray(record.decisions)) {
123
+ throw new TypeError("eval retry state identity is invalid");
124
+ }
125
+ const decisions = record.decisions.map(parseDecision);
126
+ if (new Set(decisions.map((decision) => decision.decision_id)).size !== decisions.length)
127
+ throw new TypeError("retry decision identities are duplicated");
128
+ const sorted = canonicalDecisions(decisions);
129
+ if (sorted.some((decision, index) => decision !== decisions[index]))
130
+ throw new TypeError("retry decisions are not canonically sorted");
131
+ return {
132
+ schema_version: "1", eval_id: record.eval_id, generation: record.generation,
133
+ decisions, updated_at: validTimestamp(record.updated_at, "eval retry state updated_at"),
134
+ };
135
+ }
136
+ function parseDecision(value) {
137
+ if (!value || typeof value !== "object" || Array.isArray(value))
138
+ throw new TypeError("retry decision must be an object");
139
+ const record = value;
140
+ const optional = ["trigger_run_id", "retry_work_id", "not_before"];
141
+ const required = ["decision_id", "slot_id", "trigger_trial_id", "retry_index", "classification", "disposition", "state", "created_at", "updated_at"];
142
+ if (Object.keys(record).some((key) => !required.includes(key) && !optional.includes(key))
143
+ || typeof record.decision_id !== "string" || !/^retry_decision_[a-f0-9]{32}$/.test(record.decision_id)
144
+ || typeof record.slot_id !== "string" || !record.slot_id || typeof record.trigger_trial_id !== "string" || !record.trigger_trial_id
145
+ || !Number.isSafeInteger(record.retry_index) || record.retry_index < 1)
146
+ throw new TypeError("retry decision identity is invalid");
147
+ const classification = parseFailureClassification(record.classification);
148
+ if (!isOneOf(record.disposition, ["physical-retry", "verifier-only", "collect-only", "no-retry", "operator-required"])
149
+ || !isOneOf(record.state, ["planned", "running", "repaired", "invalid", "skipped", "exhausted"]))
150
+ throw new TypeError("retry decision action is invalid");
151
+ for (const field of optional)
152
+ if (record[field] !== undefined && (typeof record[field] !== "string" || !record[field]))
153
+ throw new TypeError(`retry decision ${field} is invalid`);
154
+ if (record.not_before !== undefined)
155
+ validTimestamp(record.not_before, "retry decision not_before");
156
+ if (record.disposition === "physical-retry" && typeof record.retry_work_id !== "string")
157
+ throw new TypeError("physical retry decision work identity is missing");
158
+ return {
159
+ decision_id: record.decision_id, slot_id: record.slot_id, trigger_trial_id: record.trigger_trial_id,
160
+ ...(record.trigger_run_id ? { trigger_run_id: record.trigger_run_id } : {}), retry_index: record.retry_index,
161
+ classification, disposition: record.disposition, ...(record.retry_work_id ? { retry_work_id: record.retry_work_id } : {}),
162
+ ...(record.not_before ? { not_before: record.not_before } : {}), state: record.state,
163
+ created_at: validTimestamp(record.created_at, "retry decision created_at"), updated_at: validTimestamp(record.updated_at, "retry decision updated_at"),
164
+ };
165
+ }
166
+ function dispositionFor(classification) {
167
+ if (classification.retryability === "verifier-only")
168
+ return "verifier-only";
169
+ if (classification.retryability === "collect-only")
170
+ return "collect-only";
171
+ if (classification.retryability === "operator-required")
172
+ return "operator-required";
173
+ return "no-retry";
174
+ }
175
+ function immutableDecision(decision) {
176
+ const { state: _state, created_at: _created, updated_at: _updated, not_before: _notBefore, ...identity } = decision;
177
+ return identity;
178
+ }
179
+ function allowedTransition(from, to) {
180
+ return from === "planned" && (to === "running" || to === "skipped")
181
+ || from === "running" && (to === "repaired" || to === "invalid" || to === "exhausted");
182
+ }
183
+ function canonicalDecisions(decisions) {
184
+ return [...decisions].sort((left, right) => Buffer.compare(Buffer.from(left.slot_id), Buffer.from(right.slot_id))
185
+ || left.retry_index - right.retry_index || Buffer.compare(Buffer.from(left.decision_id), Buffer.from(right.decision_id)));
186
+ }
187
+ function validTimestamp(value, label) {
188
+ if (typeof value !== "string" || !Number.isFinite(Date.parse(value)))
189
+ throw new TypeError(`${label} is invalid`);
190
+ return value;
191
+ }
192
+ function isOneOf(value, choices) {
193
+ return typeof value === "string" && choices.includes(value);
194
+ }
195
+ //# sourceMappingURL=retry-state.js.map
@@ -0,0 +1 @@
1
+ {"version":3,"file":"retry-state.js","sourceRoot":"","sources":["../../../src/evals/retry-state.ts"],"names":[],"mappings":"AAAA,OAAO,IAAI,MAAM,WAAW,CAAC;AAE7B,OAAO,EAAE,eAAe,EAAE,SAAS,EAAE,QAAQ,EAAE,UAAU,EAAE,YAAY,EAAE,MAAM,wBAAwB,CAAC;AAExG,OAAO,EAAE,oBAAoB,EAAE,0BAA0B,EAAE,oBAAoB,EAAE,MAAM,yBAAyB,CAAC;AACjH,OAAO,EAAE,qBAAqB,EAAE,kCAAkC,EAAE,MAAM,0BAA0B,CAAC;AAiCrG,MAAM,CAAC,KAAK,UAAU,kBAAkB,CAAC,aAAqB,EAAE,MAAc;IAC5E,MAAM,KAAK,GAAG,MAAM,QAAQ,CAAiB,IAAI,CAAC,IAAI,CAAC,aAAa,EAAE,kBAAkB,CAAC,EAAE,IAAI,CAAC,CAAC;IACjG,OAAO,KAAK,KAAK,IAAI,CAAC,CAAC,CAAC,IAAI,CAAC,CAAC,CAAC,mBAAmB,CAAC,KAAK,EAAE,MAAM,CAAC,CAAC;AACpE,CAAC;AAED,MAAM,CAAC,KAAK,UAAU,2BAA2B,CAAC,KAQjD;IACC,MAAM,cAAc,GAAG,oBAAoB,CAAC,KAAK,CAAC,OAAO,CAAC,CAAC;IAC3D,IAAI,CAAC,cAAc,IAAI,CAAC,oBAAoB,CAAC,cAAc,CAAC;QAAE,MAAM,IAAI,SAAS,CAAC,6DAA6D,CAAC,CAAC;IACjJ,MAAM,SAAS,GAAG,qBAAqB,CAAC,KAAK,CAAC,IAAI,EAAE,KAAK,CAAC,UAAU,EAAE,CAAC,KAAK,CAAC,OAAO,CAAC,CAAC,CAAC;IACvF,OAAO,cAAc,CAAC;QACpB,GAAG,KAAK;QACR,cAAc;QACd,WAAW,EAAE,gBAAgB;QAC7B,WAAW,EAAE,SAAS,CAAC,OAAO;QAC9B,KAAK,EAAE,SAAS;KACjB,CAAC,CAAC;AACL,CAAC;AAED,MAAM,CAAC,KAAK,UAAU,2BAA2B,CAAC,KAQjD;IACC,MAAM,cAAc,GAAG,oBAAoB,CAAC,KAAK,CAAC,OAAO,CAAC,CAAC;IAC3D,IAAI,CAAC,cAAc;QAAE,OAAO,IAAI,CAAC;IACjC,MAAM,WAAW,GAAG,cAAc,CAAC,cAAc,CAAC,CAAC;IACnD,OAAO,cAAc,CAAC;QACpB,GAAG,KAAK;QACR,cAAc;QACd,WAAW;QACX,KAAK,EAAE,KAAK,CAAC,SAAS,IAAI,oBAAoB,CAAC,cAAc,CAAC,CAAC,CAAC,CAAC,WAAW,CAAC,CAAC,CAAC,SAAS;KACzF,CAAC,CAAC;AACL,CAAC;AAED,MAAM,CAAC,KAAK,UAAU,uBAAuB,CAAC,KAM7C;IACC,OAAO,WAAW,CAAC,KAAK,CAAC,aAAa,EAAE,KAAK,CAAC,MAAM,EAAE,CAAC,OAAO,EAAE,EAAE;QAChE,MAAM,KAAK,GAAG,OAAO,CAAC,SAAS,CAAC,SAAS,CAAC,CAAC,QAAQ,EAAE,EAAE,CAAC,QAAQ,CAAC,WAAW,KAAK,KAAK,CAAC,UAAU,CAAC,CAAC;QACnG,IAAI,KAAK,GAAG,CAAC;YAAE,MAAM,IAAI,SAAS,CAAC,kCAAkC,KAAK,CAAC,UAAU,EAAE,CAAC,CAAC;QACzF,MAAM,QAAQ,GAAG,OAAO,CAAC,SAAS,CAAC,KAAK,CAAoB,CAAC;QAC7D,IAAI,QAAQ,CAAC,KAAK,KAAK,KAAK,CAAC,KAAK;YAAE,OAAO,EAAE,OAAO,EAAE,KAAK,EAAE,QAAQ,EAAE,OAAO,EAAE,KAAK,EAAE,CAAC;QACxF,IAAI,CAAC,iBAAiB,CAAC,QAAQ,CAAC,KAAK,EAAE,KAAK,CAAC,KAAK,CAAC;YAAE,MAAM,IAAI,SAAS,CAAC,+CAA+C,QAAQ,CAAC,KAAK,OAAO,KAAK,CAAC,KAAK,EAAE,CAAC,CAAC;QAC5J,MAAM,OAAO,GAAG,EAAE,GAAG,QAAQ,EAAE,KAAK,EAAE,KAAK,CAAC,KAAK,EAAE,UAAU,EAAE,cAAc,CAAC,KAAK,CAAC,GAAG,IAAI,IAAI,IAAI,EAAE,CAAC,WAAW,EAAE,EAAE,2BAA2B,CAAC,EAAE,CAAC;QACpJ,MAAM,SAAS,GAAG,CAAC,GAAG,OAAO,CAAC,SAAS,CAAC,CAAC;QACzC,SAAS,CAAC,KAAK,CAAC,GAAG,OAAO,CAAC;QAC3B,OAAO,EAAE,OAAO,EAAE,EAAE,GAAG,OAAO,EAAE,SAAS,EAAE,kBAAkB,CAAC,SAAS,CAAC,EAAE,EAAE,KAAK,EAAE,OAAO,EAAE,OAAO,EAAE,IAAI,EAAE,CAAC;IAC9G,CAAC,CAAC,CAAC;AACL,CAAC;AAED,MAAM,UAAU,gBAAgB,CAAC,IAAyB,EAAE,KAA8B,EAAE,MAAc;IACxG,IAAI,CAAC,KAAK;QAAE,OAAO,IAAI,CAAC;IACxB,MAAM,SAAS,GAAG,KAAK,CAAC,SAAS,CAAC,MAAM,CAAC,CAAC,QAAQ,EAAE,EAAE,CAAC,QAAQ,CAAC,aAAa,KAAK,MAAM,CAAC,CAAC;IAC1F,IAAI,SAAS,CAAC,MAAM,KAAK,CAAC;QAAE,OAAO,IAAI,CAAC;IACxC,MAAM,KAAK,GAAG,SAAS,CAAC,CAAC,CAAoB,CAAC;IAC9C,IAAI,SAAS,CAAC,IAAI,CAAC,CAAC,QAAQ,EAAE,EAAE,CAAC,QAAQ,CAAC,OAAO,KAAK,KAAK,CAAC,OAAO,IAAI,QAAQ,CAAC,WAAW,KAAK,KAAK,CAAC,WAAW,CAAC,EAAE,CAAC;QACnH,MAAM,IAAI,SAAS,CAAC,kDAAkD,MAAM,EAAE,CAAC,CAAC;IAClF,CAAC;IACD,MAAM,MAAM,GAAG,IAAI,CAAC,UAAU,CAAC,IAAI,CAAC,CAAC,IAAI,EAAE,EAAE,CAAC,IAAI,CAAC,KAAK,CAAC,QAAQ,CAAC,KAAK,CAAC,OAAO,CAAC,CAAC,CAAC;IAClF,IAAI,CAAC,MAAM;QAAE,MAAM,IAAI,SAAS,CAAC,sDAAsD,KAAK,CAAC,OAAO,EAAE,CAAC,CAAC;IACxG,MAAM,IAAI,GAAG,kCAAkC,CAAC,MAAM,EAAE,KAAK,CAAC,WAAW,EAAE,SAAS,CAAC,GAAG,CAAC,CAAC,QAAQ,EAAE,EAAE,CAAC,QAAQ,CAAC,gBAAgB,CAAC,CAAC,CAAC;IACnI,IAAI,IAAI,CAAC,OAAO,KAAK,MAAM;QAAE,MAAM,IAAI,SAAS,CAAC,4CAA4C,MAAM,EAAE,CAAC,CAAC;IACvG,OAAO,EAAE,IAAI,EAAE,SAAS,EAAE,CAAC;AAC7B,CAAC;AAED,KAAK,UAAU,cAAc,CAAC,KAY7B;IACC,MAAM,MAAM,GAAG,KAAK,CAAC,IAAI,CAAC,KAAK,CAAC,CAAC,CAAC,CAAC;IACnC,IAAI,CAAC,MAAM,IAAI,KAAK,CAAC,IAAI,CAAC,KAAK,CAAC,MAAM,KAAK,CAAC,IAAI,KAAK,CAAC,IAAI,CAAC,QAAQ,CAAC,MAAM,KAAK,CAAC;QAAE,MAAM,IAAI,SAAS,CAAC,iDAAiD,CAAC,CAAC;IACzJ,IAAI,CAAC,MAAM,CAAC,aAAa,CAAC,KAAK,CAAC,UAAU,CAAC,IAAI,KAAK,CAAC,UAAU,GAAG,CAAC;QAAE,MAAM,IAAI,SAAS,CAAC,iCAAiC,CAAC,CAAC;IAC5H,MAAM,SAAS,GAAG,cAAc,CAAC,KAAK,CAAC,GAAG,IAAI,IAAI,IAAI,EAAE,CAAC,WAAW,EAAE,EAAE,2BAA2B,CAAC,CAAC;IACrG,MAAM,MAAM,GAAG,UAAU,CAAC,EAAE,OAAO,EAAE,KAAK,CAAC,MAAM,EAAE,OAAO,EAAE,MAAM,EAAE,gBAAgB,EAAE,KAAK,CAAC,OAAO,CAAC,QAAQ,EAAE,WAAW,EAAE,KAAK,CAAC,UAAU,EAAE,CAAC,CAAC;IAC/I,MAAM,QAAQ,GAAoB;QAChC,WAAW,EAAE,kBAAkB,MAAM,CAAC,KAAK,CAAC,SAAS,CAAC,MAAM,EAAE,SAAS,CAAC,MAAM,GAAG,EAAE,CAAC,EAAE;QACtF,OAAO,EAAE,MAAM;QACf,gBAAgB,EAAE,KAAK,CAAC,OAAO,CAAC,QAAQ;QACxC,GAAG,CAAC,KAAK,CAAC,OAAO,CAAC,MAAM,CAAC,CAAC,CAAC,EAAE,cAAc,EAAE,KAAK,CAAC,OAAO,CAAC,MAAM,EAAE,CAAC,CAAC,CAAC,EAAE,CAAC;QACzE,WAAW,EAAE,KAAK,CAAC,UAAU;QAC7B,cAAc,EAAE,KAAK,CAAC,cAAc;QACpC,WAAW,EAAE,KAAK,CAAC,WAAW;QAC9B,GAAG,CAAC,KAAK,CAAC,WAAW,CAAC,CAAC,CAAC,EAAE,aAAa,EAAE,KAAK,CAAC,WAAW,EAAE,CAAC,CAAC,CAAC,EAAE,CAAC;QAClE,GAAG,CAAC,KAAK,CAAC,SAAS,CAAC,CAAC,CAAC,EAAE,UAAU,EAAE,cAAc,CAAC,KAAK,CAAC,SAAS,EAAE,2BAA2B,CAAC,EAAE,CAAC,CAAC,CAAC,EAAE,CAAC;QACxG,KAAK,EAAE,KAAK,CAAC,KAAK;QAClB,UAAU,EAAE,SAAS;QACrB,UAAU,EAAE,SAAS;KACtB,CAAC;IACF,OAAO,WAAW,CAAC,KAAK,CAAC,aAAa,EAAE,KAAK,CAAC,MAAM,EAAE,CAAC,OAAO,EAAE,EAAE;QAChE,MAAM,QAAQ,GAAG,OAAO,CAAC,SAAS,CAAC,IAAI,CAAC,CAAC,KAAK,EAAE,EAAE,CAAC,KAAK,CAAC,WAAW,KAAK,QAAQ,CAAC,WAAW,CAAC,CAAC;QAC/F,IAAI,QAAQ,EAAE,CAAC;YACb,IAAI,IAAI,CAAC,SAAS,CAAC,iBAAiB,CAAC,QAAQ,CAAC,CAAC,KAAK,IAAI,CAAC,SAAS,CAAC,iBAAiB,CAAC,QAAQ,CAAC,CAAC,EAAE,CAAC;gBAChG,MAAM,IAAI,SAAS,CAAC,qCAAqC,QAAQ,CAAC,WAAW,EAAE,CAAC,CAAC;YACnF,CAAC;YACD,OAAO,EAAE,OAAO,EAAE,KAAK,EAAE,QAAQ,EAAE,OAAO,EAAE,KAAK,EAAE,CAAC;QACtD,CAAC;QACD,OAAO,EAAE,OAAO,EAAE,EAAE,GAAG,OAAO,EAAE,SAAS,EAAE,kBAAkB,CAAC,CAAC,GAAG,OAAO,CAAC,SAAS,EAAE,QAAQ,CAAC,CAAC,EAAE,EAAE,KAAK,EAAE,QAAQ,EAAE,OAAO,EAAE,IAAI,EAAE,CAAC;IACtI,CAAC,CAAC,CAAC;AACL,CAAC;AAED,KAAK,UAAU,WAAW,CACxB,aAAqB,EACrB,MAAc,EACd,SAAiG;IAEjG,MAAM,SAAS,CAAC,aAAa,CAAC,CAAC;IAC/B,OAAO,YAAY,CAAC,IAAI,CAAC,IAAI,CAAC,aAAa,EAAE,QAAQ,CAAC,EAAE,aAAa,EAAE,KAAK,IAAI,EAAE;QAChF,MAAM,QAAQ,GAAG,MAAM,kBAAkB,CAAC,aAAa,EAAE,MAAM,CAAC,CAAC;QACjE,MAAM,GAAG,GAAG,IAAI,IAAI,EAAE,CAAC,WAAW,EAAE,CAAC;QACrC,MAAM,KAAK,GAAG,QAAQ,IAAI,EAAE,cAAc,EAAE,GAAG,EAAE,OAAO,EAAE,MAAM,EAAE,UAAU,EAAE,CAAC,EAAE,SAAS,EAAE,EAAE,EAAE,UAAU,EAAE,GAAG,EAAE,CAAC;QAClH,MAAM,MAAM,GAAG,SAAS,CAAC,KAAK,CAAC,CAAC;QAChC,IAAI,MAAM,CAAC,OAAO,EAAE,CAAC;YACnB,MAAM,IAAI,GAAG,mBAAmB,CAAC,EAAE,GAAG,MAAM,CAAC,OAAO,EAAE,UAAU,EAAE,KAAK,CAAC,UAAU,GAAG,CAAC,EAAE,UAAU,EAAE,GAAG,EAAE,EAAE,MAAM,CAAC,CAAC;YACnH,MAAM,eAAe,CAAC,IAAI,CAAC,IAAI,CAAC,aAAa,EAAE,kBAAkB,CAAC,EAAE,IAAI,CAAC,CAAC;QAC5E,CAAC;QACD,OAAO,MAAM,CAAC,KAAK,CAAC;IACtB,CAAC,EAAE,EAAE,WAAW,EAAE,yBAAyB,EAAE,eAAe,EAAE,EAAE,EAAE,CAAC,CAAC;AACtE,CAAC;AAED,MAAM,UAAU,mBAAmB,CAAC,KAAc,EAAE,cAAuB;IACzE,IAAI,CAAC,KAAK,IAAI,OAAO,KAAK,KAAK,QAAQ,IAAI,KAAK,CAAC,OAAO,CAAC,KAAK,CAAC;QAAE,MAAM,IAAI,SAAS,CAAC,oCAAoC,CAAC,CAAC;IAC3H,MAAM,MAAM,GAAG,KAAgC,CAAC;IAChD,IAAI,MAAM,CAAC,IAAI,CAAC,MAAM,CAAC,CAAC,IAAI,CAAC,CAAC,GAAG,EAAE,EAAE,CAAC,CAAC,CAAC,gBAAgB,EAAE,SAAS,EAAE,YAAY,EAAE,WAAW,EAAE,YAAY,CAAC,CAAC,QAAQ,CAAC,GAAG,CAAC,CAAC;WACvH,MAAM,CAAC,cAAc,KAAK,GAAG,IAAI,OAAO,MAAM,CAAC,OAAO,KAAK,QAAQ,IAAI,CAAC,qBAAqB,CAAC,IAAI,CAAC,MAAM,CAAC,OAAO,CAAC;WAClH,cAAc,KAAK,SAAS,IAAI,MAAM,CAAC,OAAO,KAAK,cAAc;WACjE,CAAC,MAAM,CAAC,aAAa,CAAC,MAAM,CAAC,UAAU,CAAC,IAAK,MAAM,CAAC,UAAqB,GAAG,CAAC,IAAI,CAAC,KAAK,CAAC,OAAO,CAAC,MAAM,CAAC,SAAS,CAAC,EAAE,CAAC;QACvH,MAAM,IAAI,SAAS,CAAC,sCAAsC,CAAC,CAAC;IAC9D,CAAC;IACD,MAAM,SAAS,GAAG,MAAM,CAAC,SAAS,CAAC,GAAG,CAAC,aAAa,CAAC,CAAC;IACtD,IAAI,IAAI,GAAG,CAAC,SAAS,CAAC,GAAG,CAAC,CAAC,QAAQ,EAAE,EAAE,CAAC,QAAQ,CAAC,WAAW,CAAC,CAAC,CAAC,IAAI,KAAK,SAAS,CAAC,MAAM;QAAE,MAAM,IAAI,SAAS,CAAC,0CAA0C,CAAC,CAAC;IAC1J,MAAM,MAAM,GAAG,kBAAkB,CAAC,SAAS,CAAC,CAAC;IAC7C,IAAI,MAAM,CAAC,IAAI,CAAC,CAAC,QAAQ,EAAE,KAAK,EAAE,EAAE,CAAC,QAAQ,KAAK,SAAS,CAAC,KAAK,CAAC,CAAC;QAAE,MAAM,IAAI,SAAS,CAAC,4CAA4C,CAAC,CAAC;IACvI,OAAO;QACL,cAAc,EAAE,GAAG,EAAE,OAAO,EAAE,MAAM,CAAC,OAAO,EAAE,UAAU,EAAE,MAAM,CAAC,UAAoB;QACrF,SAAS,EAAE,UAAU,EAAE,cAAc,CAAC,MAAM,CAAC,UAAU,EAAE,6BAA6B,CAAC;KACxF,CAAC;AACJ,CAAC;AAED,SAAS,aAAa,CAAC,KAAc;IACnC,IAAI,CAAC,KAAK,IAAI,OAAO,KAAK,KAAK,QAAQ,IAAI,KAAK,CAAC,OAAO,CAAC,KAAK,CAAC;QAAE,MAAM,IAAI,SAAS,CAAC,kCAAkC,CAAC,CAAC;IACzH,MAAM,MAAM,GAAG,KAAgC,CAAC;IAChD,MAAM,QAAQ,GAAG,CAAC,gBAAgB,EAAE,eAAe,EAAE,YAAY,CAAC,CAAC;IACnE,MAAM,QAAQ,GAAG,CAAC,aAAa,EAAE,SAAS,EAAE,kBAAkB,EAAE,aAAa,EAAE,gBAAgB,EAAE,aAAa,EAAE,OAAO,EAAE,YAAY,EAAE,YAAY,CAAC,CAAC;IACrJ,IAAI,MAAM,CAAC,IAAI,CAAC,MAAM,CAAC,CAAC,IAAI,CAAC,CAAC,GAAG,EAAE,EAAE,CAAC,CAAC,QAAQ,CAAC,QAAQ,CAAC,GAAG,CAAC,IAAI,CAAC,QAAQ,CAAC,QAAQ,CAAC,GAAG,CAAC,CAAC;WACpF,OAAO,MAAM,CAAC,WAAW,KAAK,QAAQ,IAAI,CAAC,+BAA+B,CAAC,IAAI,CAAC,MAAM,CAAC,WAAW,CAAC;WACnG,OAAO,MAAM,CAAC,OAAO,KAAK,QAAQ,IAAI,CAAC,MAAM,CAAC,OAAO,IAAI,OAAO,MAAM,CAAC,gBAAgB,KAAK,QAAQ,IAAI,CAAC,MAAM,CAAC,gBAAgB;WAChI,CAAC,MAAM,CAAC,aAAa,CAAC,MAAM,CAAC,WAAW,CAAC,IAAK,MAAM,CAAC,WAAsB,GAAG,CAAC;QAAE,MAAM,IAAI,SAAS,CAAC,oCAAoC,CAAC,CAAC;IAChJ,MAAM,cAAc,GAAG,0BAA0B,CAAC,MAAM,CAAC,cAAc,CAAC,CAAC;IACzE,IAAI,CAAC,OAAO,CAAC,MAAM,CAAC,WAAW,EAAE,CAAC,gBAAgB,EAAE,eAAe,EAAE,cAAc,EAAE,UAAU,EAAE,mBAAmB,CAAC,CAAC;WACjH,CAAC,OAAO,CAAC,MAAM,CAAC,KAAK,EAAE,CAAC,SAAS,EAAE,SAAS,EAAE,UAAU,EAAE,SAAS,EAAE,SAAS,EAAE,WAAW,CAAC,CAAC;QAAE,MAAM,IAAI,SAAS,CAAC,kCAAkC,CAAC,CAAC;IAC5J,KAAK,MAAM,KAAK,IAAI,QAAQ;QAAE,IAAI,MAAM,CAAC,KAAK,CAAC,KAAK,SAAS,IAAI,CAAC,OAAO,MAAM,CAAC,KAAK,CAAC,KAAK,QAAQ,IAAI,CAAC,MAAM,CAAC,KAAK,CAAC,CAAC;YAAE,MAAM,IAAI,SAAS,CAAC,kBAAkB,KAAK,aAAa,CAAC,CAAC;IAClL,IAAI,MAAM,CAAC,UAAU,KAAK,SAAS;QAAE,cAAc,CAAC,MAAM,CAAC,UAAU,EAAE,2BAA2B,CAAC,CAAC;IACpG,IAAI,MAAM,CAAC,WAAW,KAAK,gBAAgB,IAAI,OAAO,MAAM,CAAC,aAAa,KAAK,QAAQ;QAAE,MAAM,IAAI,SAAS,CAAC,kDAAkD,CAAC,CAAC;IACjK,OAAO;QACL,WAAW,EAAE,MAAM,CAAC,WAAW,EAAE,OAAO,EAAE,MAAM,CAAC,OAAO,EAAE,gBAAgB,EAAE,MAAM,CAAC,gBAAgB;QACnG,GAAG,CAAC,MAAM,CAAC,cAAc,CAAC,CAAC,CAAC,EAAE,cAAc,EAAE,MAAM,CAAC,cAAwB,EAAE,CAAC,CAAC,CAAC,EAAE,CAAC,EAAE,WAAW,EAAE,MAAM,CAAC,WAAqB;QAChI,cAAc,EAAE,WAAW,EAAE,MAAM,CAAC,WAAW,EAAE,GAAG,CAAC,MAAM,CAAC,aAAa,CAAC,CAAC,CAAC,EAAE,aAAa,EAAE,MAAM,CAAC,aAAuB,EAAE,CAAC,CAAC,CAAC,EAAE,CAAC;QACnI,GAAG,CAAC,MAAM,CAAC,UAAU,CAAC,CAAC,CAAC,EAAE,UAAU,EAAE,MAAM,CAAC,UAAoB,EAAE,CAAC,CAAC,CAAC,EAAE,CAAC,EAAE,KAAK,EAAE,MAAM,CAAC,KAAK;QAC9F,UAAU,EAAE,cAAc,CAAC,MAAM,CAAC,UAAU,EAAE,2BAA2B,CAAC,EAAE,UAAU,EAAE,cAAc,CAAC,MAAM,CAAC,UAAU,EAAE,2BAA2B,CAAC;KACvJ,CAAC;AACJ,CAAC;AAED,SAAS,cAAc,CAAC,cAAuC;IAC7D,IAAI,cAAc,CAAC,YAAY,KAAK,eAAe;QAAE,OAAO,eAAe,CAAC;IAC5E,IAAI,cAAc,CAAC,YAAY,KAAK,cAAc;QAAE,OAAO,cAAc,CAAC;IAC1E,IAAI,cAAc,CAAC,YAAY,KAAK,mBAAmB;QAAE,OAAO,mBAAmB,CAAC;IACpF,OAAO,UAAU,CAAC;AACpB,CAAC;AAED,SAAS,iBAAiB,CAAC,QAAyB;IAClD,MAAM,EAAE,KAAK,EAAE,MAAM,EAAE,UAAU,EAAE,QAAQ,EAAE,UAAU,EAAE,QAAQ,EAAE,UAAU,EAAE,UAAU,EAAE,GAAG,QAAQ,EAAE,GAAG,QAAQ,CAAC;IACpH,OAAO,QAAQ,CAAC;AAClB,CAAC;AAED,SAAS,iBAAiB,CAAC,IAA0B,EAAE,EAAwB;IAC7E,OAAO,IAAI,KAAK,SAAS,IAAI,CAAC,EAAE,KAAK,SAAS,IAAI,EAAE,KAAK,SAAS,CAAC;WAC9D,IAAI,KAAK,SAAS,IAAI,CAAC,EAAE,KAAK,UAAU,IAAI,EAAE,KAAK,SAAS,IAAI,EAAE,KAAK,WAAW,CAAC,CAAC;AAC3F,CAAC;AAED,SAAS,kBAAkB,CAAC,SAA4B;IACtD,OAAO,CAAC,GAAG,SAAS,CAAC,CAAC,IAAI,CAAC,CAAC,IAAI,EAAE,KAAK,EAAE,EAAE,CAAC,MAAM,CAAC,OAAO,CAAC,MAAM,CAAC,IAAI,CAAC,IAAI,CAAC,OAAO,CAAC,EAAE,MAAM,CAAC,IAAI,CAAC,KAAK,CAAC,OAAO,CAAC,CAAC;WAC5G,IAAI,CAAC,WAAW,GAAG,KAAK,CAAC,WAAW,IAAI,MAAM,CAAC,OAAO,CAAC,MAAM,CAAC,IAAI,CAAC,IAAI,CAAC,WAAW,CAAC,EAAE,MAAM,CAAC,IAAI,CAAC,KAAK,CAAC,WAAW,CAAC,CAAC,CAAC,CAAC;AAC9H,CAAC;AAED,SAAS,cAAc,CAAC,KAAc,EAAE,KAAa;IACnD,IAAI,OAAO,KAAK,KAAK,QAAQ,IAAI,CAAC,MAAM,CAAC,QAAQ,CAAC,IAAI,CAAC,KAAK,CAAC,KAAK,CAAC,CAAC;QAAE,MAAM,IAAI,SAAS,CAAC,GAAG,KAAK,aAAa,CAAC,CAAC;IACjH,OAAO,KAAK,CAAC;AACf,CAAC;AAED,SAAS,OAAO,CAAmB,KAAc,EAAE,OAAqB;IACtE,OAAO,OAAO,KAAK,KAAK,QAAQ,IAAI,OAAO,CAAC,QAAQ,CAAC,KAAU,CAAC,CAAC;AACnE,CAAC"}
@@ -0,0 +1,95 @@
1
+ import { performance } from "node:perf_hooks";
2
+ export class EvalSchedulerMetrics {
3
+ startedAt = performance.now();
4
+ lastTransitionAt = this.startedAt;
5
+ active = 0;
6
+ weightedActiveMs = 0;
7
+ maxActive = 0;
8
+ finalSingleTailStartedAt;
9
+ finalSingleTailMs = 0;
10
+ initialWorkMs = 0;
11
+ retryWorkMs = 0;
12
+ verifierWorkMs = 0;
13
+ resourceBlockedMs = 0;
14
+ collisionBlockedMs = 0;
15
+ backoffBlockedMs = 0;
16
+ running = new Map();
17
+ startWork(workId, kind) {
18
+ if (this.running.has(workId))
19
+ return;
20
+ const now = performance.now();
21
+ this.transition(now);
22
+ this.active += 1;
23
+ this.maxActive = Math.max(this.maxActive, this.active);
24
+ if (this.active === 1)
25
+ this.finalSingleTailStartedAt = now;
26
+ else
27
+ this.finalSingleTailStartedAt = undefined;
28
+ this.running.set(workId, { kind, startedAt: now });
29
+ }
30
+ finishWork(workId) {
31
+ const entry = this.running.get(workId);
32
+ if (!entry)
33
+ return 0;
34
+ const now = performance.now();
35
+ this.transition(now);
36
+ const duration = Math.max(0, now - entry.startedAt);
37
+ if (entry.kind === "initial")
38
+ this.initialWorkMs += duration;
39
+ else
40
+ this.retryWorkMs += duration;
41
+ this.running.delete(workId);
42
+ this.active = Math.max(0, this.active - 1);
43
+ if (this.active === 1)
44
+ this.finalSingleTailStartedAt = now;
45
+ else if (this.active === 0 && this.finalSingleTailStartedAt !== undefined) {
46
+ this.finalSingleTailMs = now - this.finalSingleTailStartedAt;
47
+ }
48
+ return Math.max(1, Math.round(duration));
49
+ }
50
+ addVerifier(durationMs) {
51
+ if (Number.isFinite(durationMs) && durationMs >= 0)
52
+ this.verifierWorkMs += durationMs;
53
+ }
54
+ addBlocked(kind, durationMs) {
55
+ if (!Number.isFinite(durationMs) || durationMs < 0)
56
+ return;
57
+ if (kind === "resource")
58
+ this.resourceBlockedMs += durationMs;
59
+ else if (kind === "collision")
60
+ this.collisionBlockedMs += durationMs;
61
+ else
62
+ this.backoffBlockedMs += durationMs;
63
+ }
64
+ summary(input) {
65
+ const now = performance.now();
66
+ this.transition(now);
67
+ const makespan = Math.max(1, now - this.startedAt);
68
+ const physical = this.initialWorkMs + this.retryWorkMs;
69
+ return {
70
+ policy: input.prioritized ? "critical-path-lpt-v1" : "fifo-compat",
71
+ makespan_ms: Math.round(makespan),
72
+ physical_work_ms: Math.round(physical),
73
+ initial_work_ms: Math.round(this.initialWorkMs),
74
+ retry_work_ms: Math.round(this.retryWorkMs),
75
+ verifier_work_ms: Math.round(this.verifierWorkMs),
76
+ max_active: this.maxActive,
77
+ effective_parallelism: rounded(this.weightedActiveMs / makespan),
78
+ slot_utilization: rounded(this.weightedActiveMs / (makespan * Math.max(1, input.maxParallelism))),
79
+ single_active_tail_ms: Math.round(this.active === 1 && this.finalSingleTailStartedAt !== undefined
80
+ ? now - this.finalSingleTailStartedAt : this.finalSingleTailMs),
81
+ resource_blocked_ms: Math.round(this.resourceBlockedMs),
82
+ collision_blocked_ms: Math.round(this.collisionBlockedMs),
83
+ backoff_blocked_ms: Math.round(this.backoffBlockedMs),
84
+ verifier_skipped: input.verifierSkipped,
85
+ };
86
+ }
87
+ transition(now) {
88
+ this.weightedActiveMs += Math.max(0, now - this.lastTransitionAt) * this.active;
89
+ this.lastTransitionAt = now;
90
+ }
91
+ }
92
+ function rounded(value) {
93
+ return Math.round(value * 1_000) / 1_000;
94
+ }
95
+ //# sourceMappingURL=scheduler-metrics.js.map
@@ -0,0 +1 @@
1
+ {"version":3,"file":"scheduler-metrics.js","sourceRoot":"","sources":["../../../src/evals/scheduler-metrics.ts"],"names":[],"mappings":"AAAA,OAAO,EAAE,WAAW,EAAE,MAAM,iBAAiB,CAAC;AAG9C,MAAM,OAAO,oBAAoB;IACd,SAAS,GAAG,WAAW,CAAC,GAAG,EAAE,CAAC;IACvC,gBAAgB,GAAG,IAAI,CAAC,SAAS,CAAC;IAClC,MAAM,GAAG,CAAC,CAAC;IACX,gBAAgB,GAAG,CAAC,CAAC;IACrB,SAAS,GAAG,CAAC,CAAC;IACd,wBAAwB,CAAqB;IAC7C,iBAAiB,GAAG,CAAC,CAAC;IACtB,aAAa,GAAG,CAAC,CAAC;IAClB,WAAW,GAAG,CAAC,CAAC;IAChB,cAAc,GAAG,CAAC,CAAC;IACnB,iBAAiB,GAAG,CAAC,CAAC;IACtB,kBAAkB,GAAG,CAAC,CAAC;IACvB,gBAAgB,GAAG,CAAC,CAAC;IACZ,OAAO,GAAG,IAAI,GAAG,EAA4D,CAAC;IAE/F,SAAS,CAAC,MAAc,EAAE,IAAyB;QACjD,IAAI,IAAI,CAAC,OAAO,CAAC,GAAG,CAAC,MAAM,CAAC;YAAE,OAAO;QACrC,MAAM,GAAG,GAAG,WAAW,CAAC,GAAG,EAAE,CAAC;QAC9B,IAAI,CAAC,UAAU,CAAC,GAAG,CAAC,CAAC;QACrB,IAAI,CAAC,MAAM,IAAI,CAAC,CAAC;QACjB,IAAI,CAAC,SAAS,GAAG,IAAI,CAAC,GAAG,CAAC,IAAI,CAAC,SAAS,EAAE,IAAI,CAAC,MAAM,CAAC,CAAC;QACvD,IAAI,IAAI,CAAC,MAAM,KAAK,CAAC;YAAE,IAAI,CAAC,wBAAwB,GAAG,GAAG,CAAC;;YACtD,IAAI,CAAC,wBAAwB,GAAG,SAAS,CAAC;QAC/C,IAAI,CAAC,OAAO,CAAC,GAAG,CAAC,MAAM,EAAE,EAAE,IAAI,EAAE,SAAS,EAAE,GAAG,EAAE,CAAC,CAAC;IACrD,CAAC;IAED,UAAU,CAAC,MAAc;QACvB,MAAM,KAAK,GAAG,IAAI,CAAC,OAAO,CAAC,GAAG,CAAC,MAAM,CAAC,CAAC;QACvC,IAAI,CAAC,KAAK;YAAE,OAAO,CAAC,CAAC;QACrB,MAAM,GAAG,GAAG,WAAW,CAAC,GAAG,EAAE,CAAC;QAC9B,IAAI,CAAC,UAAU,CAAC,GAAG,CAAC,CAAC;QACrB,MAAM,QAAQ,GAAG,IAAI,CAAC,GAAG,CAAC,CAAC,EAAE,GAAG,GAAG,KAAK,CAAC,SAAS,CAAC,CAAC;QACpD,IAAI,KAAK,CAAC,IAAI,KAAK,SAAS;YAAE,IAAI,CAAC,aAAa,IAAI,QAAQ,CAAC;;YACxD,IAAI,CAAC,WAAW,IAAI,QAAQ,CAAC;QAClC,IAAI,CAAC,OAAO,CAAC,MAAM,CAAC,MAAM,CAAC,CAAC;QAC5B,IAAI,CAAC,MAAM,GAAG,IAAI,CAAC,GAAG,CAAC,CAAC,EAAE,IAAI,CAAC,MAAM,GAAG,CAAC,CAAC,CAAC;QAC3C,IAAI,IAAI,CAAC,MAAM,KAAK,CAAC;YAAE,IAAI,CAAC,wBAAwB,GAAG,GAAG,CAAC;aACtD,IAAI,IAAI,CAAC,MAAM,KAAK,CAAC,IAAI,IAAI,CAAC,wBAAwB,KAAK,SAAS,EAAE,CAAC;YAC1E,IAAI,CAAC,iBAAiB,GAAG,GAAG,GAAG,IAAI,CAAC,wBAAwB,CAAC;QAC/D,CAAC;QACD,OAAO,IAAI,CAAC,GAAG,CAAC,CAAC,EAAE,IAAI,CAAC,KAAK,CAAC,QAAQ,CAAC,CAAC,CAAC;IAC3C,CAAC;IAED,WAAW,CAAC,UAAkB;QAC5B,IAAI,MAAM,CAAC,QAAQ,CAAC,UAAU,CAAC,IAAI,UAAU,IAAI,CAAC;YAAE,IAAI,CAAC,cAAc,IAAI,UAAU,CAAC;IACxF,CAAC;IAED,UAAU,CAAC,IAA0C,EAAE,UAAkB;QACvE,IAAI,CAAC,MAAM,CAAC,QAAQ,CAAC,UAAU,CAAC,IAAI,UAAU,GAAG,CAAC;YAAE,OAAO;QAC3D,IAAI,IAAI,KAAK,UAAU;YAAE,IAAI,CAAC,iBAAiB,IAAI,UAAU,CAAC;aACzD,IAAI,IAAI,KAAK,WAAW;YAAE,IAAI,CAAC,kBAAkB,IAAI,UAAU,CAAC;;YAChE,IAAI,CAAC,gBAAgB,IAAI,UAAU,CAAC;IAC3C,CAAC;IAED,OAAO,CAAC,KAAgF;QACtF,MAAM,GAAG,GAAG,WAAW,CAAC,GAAG,EAAE,CAAC;QAC9B,IAAI,CAAC,UAAU,CAAC,GAAG,CAAC,CAAC;QACrB,MAAM,QAAQ,GAAG,IAAI,CAAC,GAAG,CAAC,CAAC,EAAE,GAAG,GAAG,IAAI,CAAC,SAAS,CAAC,CAAC;QACnD,MAAM,QAAQ,GAAG,IAAI,CAAC,aAAa,GAAG,IAAI,CAAC,WAAW,CAAC;QACvD,OAAO;YACL,MAAM,EAAE,KAAK,CAAC,WAAW,CAAC,CAAC,CAAC,sBAAsB,CAAC,CAAC,CAAC,aAAa;YAClE,WAAW,EAAE,IAAI,CAAC,KAAK,CAAC,QAAQ,CAAC;YACjC,gBAAgB,EAAE,IAAI,CAAC,KAAK,CAAC,QAAQ,CAAC;YACtC,eAAe,EAAE,IAAI,CAAC,KAAK,CAAC,IAAI,CAAC,aAAa,CAAC;YAC/C,aAAa,EAAE,IAAI,CAAC,KAAK,CAAC,IAAI,CAAC,WAAW,CAAC;YAC3C,gBAAgB,EAAE,IAAI,CAAC,KAAK,CAAC,IAAI,CAAC,cAAc,CAAC;YACjD,UAAU,EAAE,IAAI,CAAC,SAAS;YAC1B,qBAAqB,EAAE,OAAO,CAAC,IAAI,CAAC,gBAAgB,GAAG,QAAQ,CAAC;YAChE,gBAAgB,EAAE,OAAO,CAAC,IAAI,CAAC,gBAAgB,GAAG,CAAC,QAAQ,GAAG,IAAI,CAAC,GAAG,CAAC,CAAC,EAAE,KAAK,CAAC,cAAc,CAAC,CAAC,CAAC;YACjG,qBAAqB,EAAE,IAAI,CAAC,KAAK,CAAC,IAAI,CAAC,MAAM,KAAK,CAAC,IAAI,IAAI,CAAC,wBAAwB,KAAK,SAAS;gBAChG,CAAC,CAAC,GAAG,GAAG,IAAI,CAAC,wBAAwB,CAAC,CAAC,CAAC,IAAI,CAAC,iBAAiB,CAAC;YACjE,mBAAmB,EAAE,IAAI,CAAC,KAAK,CAAC,IAAI,CAAC,iBAAiB,CAAC;YACvD,oBAAoB,EAAE,IAAI,CAAC,KAAK,CAAC,IAAI,CAAC,kBAAkB,CAAC;YACzD,kBAAkB,EAAE,IAAI,CAAC,KAAK,CAAC,IAAI,CAAC,gBAAgB,CAAC;YACrD,gBAAgB,EAAE,KAAK,CAAC,eAAe;SACxC,CAAC;IACJ,CAAC;IAEO,UAAU,CAAC,GAAW;QAC5B,IAAI,CAAC,gBAAgB,IAAI,IAAI,CAAC,GAAG,CAAC,CAAC,EAAE,GAAG,GAAG,IAAI,CAAC,gBAAgB,CAAC,GAAG,IAAI,CAAC,MAAM,CAAC;QAChF,IAAI,CAAC,gBAAgB,GAAG,GAAG,CAAC;IAC9B,CAAC;CACF;AAED,SAAS,OAAO,CAAC,KAAa;IAC5B,OAAO,IAAI,CAAC,KAAK,CAAC,KAAK,GAAG,KAAK,CAAC,GAAG,KAAK,CAAC;AAC3C,CAAC"}
@@ -0,0 +1,107 @@
1
+ export function replaySchedulerTrace(work, options) {
2
+ if (!Number.isSafeInteger(options.slots) || options.slots < 1)
3
+ throw new TypeError("trace replay slots are invalid");
4
+ const chains = traceChains(work);
5
+ const replay = options.retryScheduling === "batch-v1"
6
+ ? replayBatch(chains, options.slots, options.policy)
7
+ : replayImmediate(chains, options.slots, options.policy);
8
+ const physicalWorkMs = work.reduce((sum, item) => sum + item.duration_ms, 0);
9
+ return {
10
+ makespan_ms: replay.makespanMs,
11
+ physical_work_ms: physicalWorkMs,
12
+ max_active: replay.maxActive,
13
+ slot_utilization: rounded(physicalWorkMs / (Math.max(1, replay.makespanMs) * options.slots)),
14
+ runnable_idle_ms: 0,
15
+ };
16
+ }
17
+ function traceChains(work) {
18
+ const byTask = new Map();
19
+ const ids = new Set();
20
+ for (const item of work) {
21
+ validateTraceWork(item);
22
+ if (ids.has(item.work_id))
23
+ throw new TypeError(`trace work identity is duplicated: ${item.work_id}`);
24
+ ids.add(item.work_id);
25
+ const stages = byTask.get(item.task_id) ?? [];
26
+ stages.push(item);
27
+ byTask.set(item.task_id, stages);
28
+ }
29
+ const chains = [...byTask.entries()].map(([taskId, stages]) => ({
30
+ taskId,
31
+ stages: stages.sort((left, right) => left.retry_index - right.retry_index),
32
+ }));
33
+ for (const chain of chains) {
34
+ if (chain.stages[0]?.execution_kind !== "initial" || chain.stages[0]?.retry_index !== 0
35
+ || chain.stages.some((stage, index) => stage.retry_index !== index
36
+ || (index > 0 && stage.execution_kind !== "physical-infrastructure-retry"))) {
37
+ throw new TypeError(`trace work chain is invalid: ${chain.taskId}`);
38
+ }
39
+ }
40
+ return chains.sort((left, right) => compareBytes(left.taskId, right.taskId));
41
+ }
42
+ function replayBatch(chains, slots, policy) {
43
+ let now = 0;
44
+ let maxActive = 0;
45
+ const stages = Math.max(0, ...chains.map((chain) => chain.stages.length));
46
+ for (let index = 0; index < stages; index += 1) {
47
+ const ready = chains.flatMap((chain) => chain.stages[index]
48
+ ? [{ chain, index, remainingMs: remaining(chain, index) }]
49
+ : []);
50
+ maxActive = Math.max(maxActive, Math.min(slots, ready.length));
51
+ now = scheduleReady(ready, slots, policy, now);
52
+ }
53
+ return { makespanMs: now, maxActive };
54
+ }
55
+ function replayImmediate(chains, slots, policy) {
56
+ const ready = chains.map((chain) => ({ chain, index: 0, remainingMs: remaining(chain, 0) }));
57
+ const running = [];
58
+ let now = 0;
59
+ let maxActive = 0;
60
+ while (ready.length > 0 || running.length > 0) {
61
+ ready.sort((left, right) => compareReady(left, right, policy));
62
+ while (running.length < slots && ready.length > 0) {
63
+ const item = ready.shift();
64
+ running.push({ ...item, completedAt: now + (item.chain.stages[item.index]?.duration_ms ?? 0) });
65
+ }
66
+ maxActive = Math.max(maxActive, running.length);
67
+ running.sort((left, right) => left.completedAt - right.completedAt || compareBytes(left.chain.taskId, right.chain.taskId));
68
+ const completed = running.shift();
69
+ if (!completed)
70
+ break;
71
+ now = completed.completedAt;
72
+ const next = completed.index + 1;
73
+ if (completed.chain.stages[next])
74
+ ready.push({ chain: completed.chain, index: next, remainingMs: remaining(completed.chain, next) });
75
+ }
76
+ return { makespanMs: now, maxActive };
77
+ }
78
+ function scheduleReady(ready, slots, policy, start) {
79
+ ready.sort((left, right) => compareReady(left, right, policy));
80
+ const available = Array.from({ length: slots }, () => start);
81
+ for (const item of ready) {
82
+ const earliest = Math.min(...available);
83
+ const slot = available.indexOf(earliest);
84
+ available[slot] = earliest + (item.chain.stages[item.index]?.duration_ms ?? 0);
85
+ }
86
+ return Math.max(start, ...available);
87
+ }
88
+ function compareReady(left, right, policy) {
89
+ if (policy === "critical-path-v1" && left.remainingMs !== right.remainingMs)
90
+ return right.remainingMs - left.remainingMs;
91
+ const leftStage = left.chain.stages[left.index];
92
+ const rightStage = right.chain.stages[right.index];
93
+ return leftStage.queued_order - rightStage.queued_order || compareBytes(leftStage.work_id, rightStage.work_id);
94
+ }
95
+ function remaining(chain, index) {
96
+ return chain.stages.slice(index).reduce((sum, stage) => sum + stage.duration_ms, 0);
97
+ }
98
+ function validateTraceWork(item) {
99
+ if (!item.work_id || !item.task_id || !Number.isSafeInteger(item.duration_ms) || item.duration_ms < 1
100
+ || !Number.isSafeInteger(item.queued_order) || item.queued_order < 0
101
+ || !Number.isSafeInteger(item.retry_index) || item.retry_index < 0
102
+ || (item.execution_kind === "initial") !== (item.retry_index === 0))
103
+ throw new TypeError("scheduler trace work is invalid");
104
+ }
105
+ function rounded(value) { return Math.round(value * 1_000) / 1_000; }
106
+ function compareBytes(left, right) { return Buffer.compare(Buffer.from(left), Buffer.from(right)); }
107
+ //# sourceMappingURL=scheduler-trace-replay.js.map
@@ -0,0 +1 @@
1
+ {"version":3,"file":"scheduler-trace-replay.js","sourceRoot":"","sources":["../../../src/evals/scheduler-trace-replay.ts"],"names":[],"mappings":"AAiBA,MAAM,UAAU,oBAAoB,CAClC,IAAqC,EACrC,OAIC;IAED,IAAI,CAAC,MAAM,CAAC,aAAa,CAAC,OAAO,CAAC,KAAK,CAAC,IAAI,OAAO,CAAC,KAAK,GAAG,CAAC;QAAE,MAAM,IAAI,SAAS,CAAC,gCAAgC,CAAC,CAAC;IACrH,MAAM,MAAM,GAAG,WAAW,CAAC,IAAI,CAAC,CAAC;IACjC,MAAM,MAAM,GAAG,OAAO,CAAC,eAAe,KAAK,UAAU;QACnD,CAAC,CAAC,WAAW,CAAC,MAAM,EAAE,OAAO,CAAC,KAAK,EAAE,OAAO,CAAC,MAAM,CAAC;QACpD,CAAC,CAAC,eAAe,CAAC,MAAM,EAAE,OAAO,CAAC,KAAK,EAAE,OAAO,CAAC,MAAM,CAAC,CAAC;IAC3D,MAAM,cAAc,GAAG,IAAI,CAAC,MAAM,CAAC,CAAC,GAAG,EAAE,IAAI,EAAE,EAAE,CAAC,GAAG,GAAG,IAAI,CAAC,WAAW,EAAE,CAAC,CAAC,CAAC;IAC7E,OAAO;QACL,WAAW,EAAE,MAAM,CAAC,UAAU;QAC9B,gBAAgB,EAAE,cAAc;QAChC,UAAU,EAAE,MAAM,CAAC,SAAS;QAC5B,gBAAgB,EAAE,OAAO,CAAC,cAAc,GAAG,CAAC,IAAI,CAAC,GAAG,CAAC,CAAC,EAAE,MAAM,CAAC,UAAU,CAAC,GAAG,OAAO,CAAC,KAAK,CAAC,CAAC;QAC5F,gBAAgB,EAAE,CAAC;KACpB,CAAC;AACJ,CAAC;AAkBD,SAAS,WAAW,CAAC,IAAqC;IACxD,MAAM,MAAM,GAAG,IAAI,GAAG,EAAkC,CAAC;IACzD,MAAM,GAAG,GAAG,IAAI,GAAG,EAAU,CAAC;IAC9B,KAAK,MAAM,IAAI,IAAI,IAAI,EAAE,CAAC;QACxB,iBAAiB,CAAC,IAAI,CAAC,CAAC;QACxB,IAAI,GAAG,CAAC,GAAG,CAAC,IAAI,CAAC,OAAO,CAAC;YAAE,MAAM,IAAI,SAAS,CAAC,sCAAsC,IAAI,CAAC,OAAO,EAAE,CAAC,CAAC;QACrG,GAAG,CAAC,GAAG,CAAC,IAAI,CAAC,OAAO,CAAC,CAAC;QACtB,MAAM,MAAM,GAAG,MAAM,CAAC,GAAG,CAAC,IAAI,CAAC,OAAO,CAAC,IAAI,EAAE,CAAC;QAC9C,MAAM,CAAC,IAAI,CAAC,IAAI,CAAC,CAAC;QAClB,MAAM,CAAC,GAAG,CAAC,IAAI,CAAC,OAAO,EAAE,MAAM,CAAC,CAAC;IACnC,CAAC;IACD,MAAM,MAAM,GAAG,CAAC,GAAG,MAAM,CAAC,OAAO,EAAE,CAAC,CAAC,GAAG,CAAC,CAAC,CAAC,MAAM,EAAE,MAAM,CAAC,EAAE,EAAE,CAAC,CAAC;QAC9D,MAAM;QACN,MAAM,EAAE,MAAM,CAAC,IAAI,CAAC,CAAC,IAAI,EAAE,KAAK,EAAE,EAAE,CAAC,IAAI,CAAC,WAAW,GAAG,KAAK,CAAC,WAAW,CAAC;KAC3E,CAAC,CAAC,CAAC;IACJ,KAAK,MAAM,KAAK,IAAI,MAAM,EAAE,CAAC;QAC3B,IAAI,KAAK,CAAC,MAAM,CAAC,CAAC,CAAC,EAAE,cAAc,KAAK,SAAS,IAAI,KAAK,CAAC,MAAM,CAAC,CAAC,CAAC,EAAE,WAAW,KAAK,CAAC;eAClF,KAAK,CAAC,MAAM,CAAC,IAAI,CAAC,CAAC,KAAK,EAAE,KAAK,EAAE,EAAE,CAAC,KAAK,CAAC,WAAW,KAAK,KAAK;mBAC7D,CAAC,KAAK,GAAG,CAAC,IAAI,KAAK,CAAC,cAAc,KAAK,+BAA+B,CAAC,CAAC,EAAE,CAAC;YAChF,MAAM,IAAI,SAAS,CAAC,gCAAgC,KAAK,CAAC,MAAM,EAAE,CAAC,CAAC;QACtE,CAAC;IACH,CAAC;IACD,OAAO,MAAM,CAAC,IAAI,CAAC,CAAC,IAAI,EAAE,KAAK,EAAE,EAAE,CAAC,YAAY,CAAC,IAAI,CAAC,MAAM,EAAE,KAAK,CAAC,MAAM,CAAC,CAAC,CAAC;AAC/E,CAAC;AAED,SAAS,WAAW,CAAC,MAA6B,EAAE,KAAa,EAAE,MAAsC;IACvG,IAAI,GAAG,GAAG,CAAC,CAAC;IACZ,IAAI,SAAS,GAAG,CAAC,CAAC;IAClB,MAAM,MAAM,GAAG,IAAI,CAAC,GAAG,CAAC,CAAC,EAAE,GAAG,MAAM,CAAC,GAAG,CAAC,CAAC,KAAK,EAAE,EAAE,CAAC,KAAK,CAAC,MAAM,CAAC,MAAM,CAAC,CAAC,CAAC;IAC1E,KAAK,IAAI,KAAK,GAAG,CAAC,EAAE,KAAK,GAAG,MAAM,EAAE,KAAK,IAAI,CAAC,EAAE,CAAC;QAC/C,MAAM,KAAK,GAAG,MAAM,CAAC,OAAO,CAAC,CAAC,KAAK,EAAE,EAAE,CAAC,KAAK,CAAC,MAAM,CAAC,KAAK,CAAC;YACzD,CAAC,CAAC,CAAC,EAAE,KAAK,EAAE,KAAK,EAAE,WAAW,EAAE,SAAS,CAAC,KAAK,EAAE,KAAK,CAAC,EAAE,CAAC;YAC1D,CAAC,CAAC,EAAE,CAAC,CAAC;QACR,SAAS,GAAG,IAAI,CAAC,GAAG,CAAC,SAAS,EAAE,IAAI,CAAC,GAAG,CAAC,KAAK,EAAE,KAAK,CAAC,MAAM,CAAC,CAAC,CAAC;QAC/D,GAAG,GAAG,aAAa,CAAC,KAAK,EAAE,KAAK,EAAE,MAAM,EAAE,GAAG,CAAC,CAAC;IACjD,CAAC;IACD,OAAO,EAAE,UAAU,EAAE,GAAG,EAAE,SAAS,EAAE,CAAC;AACxC,CAAC;AAED,SAAS,eAAe,CAAC,MAA6B,EAAE,KAAa,EAAE,MAAsC;IAC3G,MAAM,KAAK,GAAiB,MAAM,CAAC,GAAG,CAAC,CAAC,KAAK,EAAE,EAAE,CAAC,CAAC,EAAE,KAAK,EAAE,KAAK,EAAE,CAAC,EAAE,WAAW,EAAE,SAAS,CAAC,KAAK,EAAE,CAAC,CAAC,EAAE,CAAC,CAAC,CAAC;IAC3G,MAAM,OAAO,GAAgD,EAAE,CAAC;IAChE,IAAI,GAAG,GAAG,CAAC,CAAC;IACZ,IAAI,SAAS,GAAG,CAAC,CAAC;IAClB,OAAO,KAAK,CAAC,MAAM,GAAG,CAAC,IAAI,OAAO,CAAC,MAAM,GAAG,CAAC,EAAE,CAAC;QAC9C,KAAK,CAAC,IAAI,CAAC,CAAC,IAAI,EAAE,KAAK,EAAE,EAAE,CAAC,YAAY,CAAC,IAAI,EAAE,KAAK,EAAE,MAAM,CAAC,CAAC,CAAC;QAC/D,OAAO,OAAO,CAAC,MAAM,GAAG,KAAK,IAAI,KAAK,CAAC,MAAM,GAAG,CAAC,EAAE,CAAC;YAClD,MAAM,IAAI,GAAG,KAAK,CAAC,KAAK,EAAgB,CAAC;YACzC,OAAO,CAAC,IAAI,CAAC,EAAE,GAAG,IAAI,EAAE,WAAW,EAAE,GAAG,GAAG,CAAC,IAAI,CAAC,KAAK,CAAC,MAAM,CAAC,IAAI,CAAC,KAAK,CAAC,EAAE,WAAW,IAAI,CAAC,CAAC,EAAE,CAAC,CAAC;QAClG,CAAC;QACD,SAAS,GAAG,IAAI,CAAC,GAAG,CAAC,SAAS,EAAE,OAAO,CAAC,MAAM,CAAC,CAAC;QAChD,OAAO,CAAC,IAAI,CAAC,CAAC,IAAI,EAAE,KAAK,EAAE,EAAE,CAAC,IAAI,CAAC,WAAW,GAAG,KAAK,CAAC,WAAW,IAAI,YAAY,CAAC,IAAI,CAAC,KAAK,CAAC,MAAM,EAAE,KAAK,CAAC,KAAK,CAAC,MAAM,CAAC,CAAC,CAAC;QAC3H,MAAM,SAAS,GAAG,OAAO,CAAC,KAAK,EAAE,CAAC;QAClC,IAAI,CAAC,SAAS;YAAE,MAAM;QACtB,GAAG,GAAG,SAAS,CAAC,WAAW,CAAC;QAC5B,MAAM,IAAI,GAAG,SAAS,CAAC,KAAK,GAAG,CAAC,CAAC;QACjC,IAAI,SAAS,CAAC,KAAK,CAAC,MAAM,CAAC,IAAI,CAAC;YAAE,KAAK,CAAC,IAAI,CAAC,EAAE,KAAK,EAAE,SAAS,CAAC,KAAK,EAAE,KAAK,EAAE,IAAI,EAAE,WAAW,EAAE,SAAS,CAAC,SAAS,CAAC,KAAK,EAAE,IAAI,CAAC,EAAE,CAAC,CAAC;IACvI,CAAC;IACD,OAAO,EAAE,UAAU,EAAE,GAAG,EAAE,SAAS,EAAE,CAAC;AACxC,CAAC;AAED,SAAS,aAAa,CAAC,KAAmB,EAAE,KAAa,EAAE,MAAsC,EAAE,KAAa;IAC9G,KAAK,CAAC,IAAI,CAAC,CAAC,IAAI,EAAE,KAAK,EAAE,EAAE,CAAC,YAAY,CAAC,IAAI,EAAE,KAAK,EAAE,MAAM,CAAC,CAAC,CAAC;IAC/D,MAAM,SAAS,GAAG,KAAK,CAAC,IAAI,CAAC,EAAE,MAAM,EAAE,KAAK,EAAE,EAAE,GAAG,EAAE,CAAC,KAAK,CAAC,CAAC;IAC7D,KAAK,MAAM,IAAI,IAAI,KAAK,EAAE,CAAC;QACzB,MAAM,QAAQ,GAAG,IAAI,CAAC,GAAG,CAAC,GAAG,SAAS,CAAC,CAAC;QACxC,MAAM,IAAI,GAAG,SAAS,CAAC,OAAO,CAAC,QAAQ,CAAC,CAAC;QACzC,SAAS,CAAC,IAAI,CAAC,GAAG,QAAQ,GAAG,CAAC,IAAI,CAAC,KAAK,CAAC,MAAM,CAAC,IAAI,CAAC,KAAK,CAAC,EAAE,WAAW,IAAI,CAAC,CAAC,CAAC;IACjF,CAAC;IACD,OAAO,IAAI,CAAC,GAAG,CAAC,KAAK,EAAE,GAAG,SAAS,CAAC,CAAC;AACvC,CAAC;AAED,SAAS,YAAY,CAAC,IAAgB,EAAE,KAAiB,EAAE,MAAsC;IAC/F,IAAI,MAAM,KAAK,kBAAkB,IAAI,IAAI,CAAC,WAAW,KAAK,KAAK,CAAC,WAAW;QAAE,OAAO,KAAK,CAAC,WAAW,GAAG,IAAI,CAAC,WAAW,CAAC;IACzH,MAAM,SAAS,GAAG,IAAI,CAAC,KAAK,CAAC,MAAM,CAAC,IAAI,CAAC,KAAK,CAAyB,CAAC;IACxE,MAAM,UAAU,GAAG,KAAK,CAAC,KAAK,CAAC,MAAM,CAAC,KAAK,CAAC,KAAK,CAAyB,CAAC;IAC3E,OAAO,SAAS,CAAC,YAAY,GAAG,UAAU,CAAC,YAAY,IAAI,YAAY,CAAC,SAAS,CAAC,OAAO,EAAE,UAAU,CAAC,OAAO,CAAC,CAAC;AACjH,CAAC;AAED,SAAS,SAAS,CAAC,KAAiB,EAAE,KAAa;IACjD,OAAO,KAAK,CAAC,MAAM,CAAC,KAAK,CAAC,KAAK,CAAC,CAAC,MAAM,CAAC,CAAC,GAAG,EAAE,KAAK,EAAE,EAAE,CAAC,GAAG,GAAG,KAAK,CAAC,WAAW,EAAE,CAAC,CAAC,CAAC;AACtF,CAAC;AAED,SAAS,iBAAiB,CAAC,IAA0B;IACnD,IAAI,CAAC,IAAI,CAAC,OAAO,IAAI,CAAC,IAAI,CAAC,OAAO,IAAI,CAAC,MAAM,CAAC,aAAa,CAAC,IAAI,CAAC,WAAW,CAAC,IAAI,IAAI,CAAC,WAAW,GAAG,CAAC;WAChG,CAAC,MAAM,CAAC,aAAa,CAAC,IAAI,CAAC,YAAY,CAAC,IAAI,IAAI,CAAC,YAAY,GAAG,CAAC;WACjE,CAAC,MAAM,CAAC,aAAa,CAAC,IAAI,CAAC,WAAW,CAAC,IAAI,IAAI,CAAC,WAAW,GAAG,CAAC;WAC/D,CAAC,IAAI,CAAC,cAAc,KAAK,SAAS,CAAC,KAAK,CAAC,IAAI,CAAC,WAAW,KAAK,CAAC,CAAC;QAAE,MAAM,IAAI,SAAS,CAAC,iCAAiC,CAAC,CAAC;AAChI,CAAC;AAED,SAAS,OAAO,CAAC,KAAa,IAAY,OAAO,IAAI,CAAC,KAAK,CAAC,KAAK,GAAG,KAAK,CAAC,GAAG,KAAK,CAAC,CAAC,CAAC;AACrF,SAAS,YAAY,CAAC,IAAY,EAAE,KAAa,IAAY,OAAO,MAAM,CAAC,OAAO,CAAC,MAAM,CAAC,IAAI,CAAC,IAAI,CAAC,EAAE,MAAM,CAAC,IAAI,CAAC,KAAK,CAAC,CAAC,CAAC,CAAC,CAAC"}
@@ -0,0 +1,2 @@
1
+ export {};
2
+ //# sourceMappingURL=service-types.js.map
@@ -0,0 +1 @@
1
+ {"version":3,"file":"service-types.js","sourceRoot":"","sources":["../../../src/evals/service-types.ts"],"names":[],"mappings":""}