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":"physical-retry-work.js","sourceRoot":"","sources":["../../../src/evals/physical-retry-work.ts"],"names":[],"mappings":"AACA,OAAO,EAAE,UAAU,EAAE,MAAM,wBAAwB,CAAC;AAEpD;sEACsE;AACtE,MAAM,UAAU,qBAAqB,CACnC,IAAuB,EACvB,KAAa,EACb,QAAmC;IAEnC,OAAO,kCAAkC,CAAC,IAAI,EAAE,KAAK,EAAE,QAAQ,CAAC,GAAG,CAAC,CAAC,KAAK,EAAE,EAAE,CAAC,KAAK,CAAC,QAAQ,CAAC,CAAC,CAAC;AAClG,CAAC;AAED,MAAM,UAAU,kCAAkC,CAChD,IAAuB,EACvB,KAAa,EACb,eAAkC;IAElC,IAAI,CAAC,MAAM,CAAC,aAAa,CAAC,KAAK,CAAC,IAAI,KAAK,GAAG,CAAC;QAAE,MAAM,IAAI,SAAS,CAAC,sDAAsD,CAAC,CAAC;IAC3H,MAAM,aAAa,GAAG,CAAC,GAAG,IAAI,GAAG,CAAC,eAAe,CAAC,CAAC,CAAC,IAAI,CAAC,YAAY,CAAC,CAAC;IACvE,IAAI,aAAa,CAAC,MAAM,KAAK,CAAC;QAAE,MAAM,IAAI,SAAS,CAAC,oDAAoD,CAAC,CAAC;IAC1G,MAAM,QAAQ,GAAG,UAAU,CAAC;QAC1B,cAAc,EAAE,IAAI,CAAC,OAAO;QAC5B,KAAK,EAAE,IAAI,CAAC,KAAK;QACjB,cAAc,EAAE,+BAA+B;QAC/C,KAAK;QACL,cAAc,EAAE,aAAa;KAC9B,CAAC,CAAC;IACH,OAAO;QACL,GAAG,IAAI;QACP,OAAO,EAAE,QAAQ,QAAQ,CAAC,KAAK,CAAC,SAAS,CAAC,MAAM,EAAE,SAAS,CAAC,MAAM,GAAG,EAAE,CAAC,EAAE;QAC1E,GAAG,CAAC,IAAI,CAAC,UAAU,CAAC,CAAC,CAAC;YACpB,UAAU,EAAE,EAAE,GAAG,IAAI,CAAC,UAAU,EAAE,iBAAiB,EAAE,IAAI,CAAC,UAAU,CAAC,qBAAqB,EAAE;SAC7F,CAAC,CAAC,CAAC,EAAE,CAAC;KACR,CAAC;AACJ,CAAC;AAED,SAAS,YAAY,CAAC,IAAY,EAAE,KAAa;IAC/C,OAAO,MAAM,CAAC,OAAO,CAAC,MAAM,CAAC,IAAI,CAAC,IAAI,CAAC,EAAE,MAAM,CAAC,IAAI,CAAC,KAAK,CAAC,CAAC,CAAC;AAC/D,CAAC"}
@@ -0,0 +1,66 @@
1
+ export class PrioritySemaphore {
2
+ available;
3
+ waiters = [];
4
+ sequence = 0;
5
+ constructor(capacity) {
6
+ this.available = capacity;
7
+ }
8
+ async acquire(signal, priority = 0) {
9
+ if (signal?.aborted)
10
+ throw new Error("eval work item scheduling was cancelled");
11
+ if (!Number.isFinite(priority) || priority < 0)
12
+ throw new TypeError("eval work item scheduling priority is invalid");
13
+ if (this.available > 0) {
14
+ this.available -= 1;
15
+ return this.releaseFunction();
16
+ }
17
+ await new Promise((resolve, reject) => {
18
+ const waiter = { priority, sequence: this.sequence++, ready: () => {
19
+ signal?.removeEventListener("abort", aborted);
20
+ resolve();
21
+ } };
22
+ const aborted = () => {
23
+ const index = this.waiters.indexOf(waiter);
24
+ if (index >= 0)
25
+ this.waiters.splice(index, 1);
26
+ reject(new Error("eval work item scheduling was cancelled"));
27
+ };
28
+ this.waiters.push(waiter);
29
+ signal?.addEventListener("abort", aborted, { once: true });
30
+ });
31
+ return this.releaseFunction();
32
+ }
33
+ releaseFunction() {
34
+ let released = false;
35
+ return () => {
36
+ if (released)
37
+ return;
38
+ released = true;
39
+ const next = this.waiters.sort(compareWaiters).shift();
40
+ if (next)
41
+ next.ready();
42
+ else
43
+ this.available += 1;
44
+ };
45
+ }
46
+ }
47
+ /** Compatibility name for callers that do not supply priorities. */
48
+ export class FairSemaphore extends PrioritySemaphore {
49
+ }
50
+ export function workSchedulingPriority(item) {
51
+ return item.scheduling?.remaining_path_ms ?? 0;
52
+ }
53
+ function compareWaiters(left, right) {
54
+ return right.priority - left.priority || left.sequence - right.sequence;
55
+ }
56
+ export function assertTaskSlotPlan(plan) {
57
+ if (plan.membership !== "known" || plan.work_items.length === 0
58
+ || plan.work_items.some((item) => item.opaque_membership || item.task_ids.length !== 1 || item.slots.length !== 1
59
+ || item.logical_attempt === null || item.requested_parallelism !== 1)) {
60
+ throw new TypeError("planned Harbor execution requires one known trial slot per work item");
61
+ }
62
+ }
63
+ export function workOrder(plan, workId) {
64
+ return plan.work_items.findIndex((item) => item.work_id === workId);
65
+ }
66
+ //# sourceMappingURL=planned-execution-support.js.map
@@ -0,0 +1 @@
1
+ {"version":3,"file":"planned-execution-support.js","sourceRoot":"","sources":["../../../src/evals/planned-execution-support.ts"],"names":[],"mappings":"AAeA,MAAM,OAAO,iBAAiB;IACpB,SAAS,CAAS;IACT,OAAO,GAAqE,EAAE,CAAC;IACxF,QAAQ,GAAG,CAAC,CAAC;IAErB,YAAY,QAAgB;QAC1B,IAAI,CAAC,SAAS,GAAG,QAAQ,CAAC;IAC5B,CAAC;IAED,KAAK,CAAC,OAAO,CAAC,MAAoB,EAAE,QAAQ,GAAG,CAAC;QAC9C,IAAI,MAAM,EAAE,OAAO;YAAE,MAAM,IAAI,KAAK,CAAC,yCAAyC,CAAC,CAAC;QAChF,IAAI,CAAC,MAAM,CAAC,QAAQ,CAAC,QAAQ,CAAC,IAAI,QAAQ,GAAG,CAAC;YAAE,MAAM,IAAI,SAAS,CAAC,+CAA+C,CAAC,CAAC;QACrH,IAAI,IAAI,CAAC,SAAS,GAAG,CAAC,EAAE,CAAC;YACvB,IAAI,CAAC,SAAS,IAAI,CAAC,CAAC;YACpB,OAAO,IAAI,CAAC,eAAe,EAAE,CAAC;QAChC,CAAC;QACD,MAAM,IAAI,OAAO,CAAO,CAAC,OAAO,EAAE,MAAM,EAAE,EAAE;YAC1C,MAAM,MAAM,GAAG,EAAE,QAAQ,EAAE,QAAQ,EAAE,IAAI,CAAC,QAAQ,EAAE,EAAE,KAAK,EAAE,GAAG,EAAE;oBAChE,MAAM,EAAE,mBAAmB,CAAC,OAAO,EAAE,OAAO,CAAC,CAAC;oBAC9C,OAAO,EAAE,CAAC;gBACZ,CAAC,EAAE,CAAC;YACJ,MAAM,OAAO,GAAG,GAAG,EAAE;gBACnB,MAAM,KAAK,GAAG,IAAI,CAAC,OAAO,CAAC,OAAO,CAAC,MAAM,CAAC,CAAC;gBAC3C,IAAI,KAAK,IAAI,CAAC;oBAAE,IAAI,CAAC,OAAO,CAAC,MAAM,CAAC,KAAK,EAAE,CAAC,CAAC,CAAC;gBAC9C,MAAM,CAAC,IAAI,KAAK,CAAC,yCAAyC,CAAC,CAAC,CAAC;YAC/D,CAAC,CAAC;YACF,IAAI,CAAC,OAAO,CAAC,IAAI,CAAC,MAAM,CAAC,CAAC;YAC1B,MAAM,EAAE,gBAAgB,CAAC,OAAO,EAAE,OAAO,EAAE,EAAE,IAAI,EAAE,IAAI,EAAE,CAAC,CAAC;QAC7D,CAAC,CAAC,CAAC;QACH,OAAO,IAAI,CAAC,eAAe,EAAE,CAAC;IAChC,CAAC;IAEO,eAAe;QACrB,IAAI,QAAQ,GAAG,KAAK,CAAC;QACrB,OAAO,GAAG,EAAE;YACV,IAAI,QAAQ;gBAAE,OAAO;YACrB,QAAQ,GAAG,IAAI,CAAC;YAChB,MAAM,IAAI,GAAG,IAAI,CAAC,OAAO,CAAC,IAAI,CAAC,cAAc,CAAC,CAAC,KAAK,EAAE,CAAC;YACvD,IAAI,IAAI;gBAAE,IAAI,CAAC,KAAK,EAAE,CAAC;;gBAClB,IAAI,CAAC,SAAS,IAAI,CAAC,CAAC;QAC3B,CAAC,CAAC;IACJ,CAAC;CACF;AAED,oEAAoE;AACpE,MAAM,OAAO,aAAc,SAAQ,iBAAiB;CAAG;AAEvD,MAAM,UAAU,sBAAsB,CAAC,IAAuB;IAC5D,OAAO,IAAI,CAAC,UAAU,EAAE,iBAAiB,IAAI,CAAC,CAAC;AACjD,CAAC;AAED,SAAS,cAAc,CAAC,IAA4C,EAAE,KAA6C;IACjH,OAAO,KAAK,CAAC,QAAQ,GAAG,IAAI,CAAC,QAAQ,IAAI,IAAI,CAAC,QAAQ,GAAG,KAAK,CAAC,QAAQ,CAAC;AAC1E,CAAC;AAED,MAAM,UAAU,kBAAkB,CAAC,IAAyB;IAC1D,IAAI,IAAI,CAAC,UAAU,KAAK,OAAO,IAAI,IAAI,CAAC,UAAU,CAAC,MAAM,KAAK,CAAC;WAC1D,IAAI,CAAC,UAAU,CAAC,IAAI,CAAC,CAAC,IAAI,EAAE,EAAE,CAAC,IAAI,CAAC,iBAAiB,IAAI,IAAI,CAAC,QAAQ,CAAC,MAAM,KAAK,CAAC,IAAI,IAAI,CAAC,KAAK,CAAC,MAAM,KAAK,CAAC;eAC5G,IAAI,CAAC,eAAe,KAAK,IAAI,IAAI,IAAI,CAAC,qBAAqB,KAAK,CAAC,CAAC,EAAE,CAAC;QAC1E,MAAM,IAAI,SAAS,CAAC,sEAAsE,CAAC,CAAC;IAC9F,CAAC;AACH,CAAC;AAED,MAAM,UAAU,SAAS,CAAC,IAAyB,EAAE,MAAc;IACjE,OAAO,IAAI,CAAC,UAAU,CAAC,SAAS,CAAC,CAAC,IAAI,EAAE,EAAE,CAAC,IAAI,CAAC,OAAO,KAAK,MAAM,CAAC,CAAC;AACtE,CAAC"}
@@ -0,0 +1,469 @@
1
+ import { ProgressPublisher } from "./planned-progress-publisher.js";
2
+ import path from "node:path";
3
+ import { runHarborBackend } from "../backends/index.js";
4
+ import { HitchError } from "../foundation/index.js";
5
+ import { DEFAULT_EXECUTION_LEASE_HEARTBEAT_MS, DEFAULT_EXECUTION_LEASE_TTL_MS, createExecutionLease } from "./execution-leases.js";
6
+ import { recordLocalDockerProcessExit, recordLocalDockerProcessStart, releaseLocalDockerProcessRecord } from "./local-docker-provider.js";
7
+ import { dockerResourceOwnership } from "./docker-ownership.js";
8
+ import { assertBackendTrialSet, localSourceBackendFailure } from "./result-helpers.js";
9
+ import { importEvalTrialRun, importEvalTrialRuns, TrialBundlePendingError } from "./trial-import.js";
10
+ import { executeRemotePlannedWorkItem } from "./planned-remote-execution.js";
11
+ import { runtimeResourcesForTask } from "./execution-plan-resources.js";
12
+ import { startDockerResourceObserver } from "./docker-resource-observer.js";
13
+ import { resolvedImageMapping } from "./environment-image-planning.js";
14
+ import { loadTrialEnvironmentImages, prebuiltTaskImage, verifyTrialEnvironmentImageExecution } from "./trial-environment-evidence.js";
15
+ import { PrioritySemaphore, assertTaskSlotPlan, workOrder, workSchedulingPriority } from "./planned-execution-support.js";
16
+ import { harborPhaseTimingEvents } from "./harbor-phase-timings.js";
17
+ import { preparedArtifactForWorkItem } from "./work-item-artifacts.js";
18
+ import { recordTaskDuration } from "./duration-estimator.js";
19
+ import { classifyTrialFailure, physicalRetryAllowed } from "./failure-classifier.js";
20
+ import { EvalSchedulerMetrics } from "./scheduler-metrics.js";
21
+ import { pendingRetryWork, reconcilePersistedInitialRetryDecisions, runPlannedInfrastructureRetries, runPlannedInfrastructureRetriesFromSource } from "./planned-retry-execution.js";
22
+ export async function executePlannedHarborTasks(options) {
23
+ assertTaskSlotPlan(options.plan);
24
+ const publisher = new ProgressPublisher(options.progress, options);
25
+ const semaphore = options.admission ? undefined : new PrioritySemaphore(options.plan.max_parallelism);
26
+ const metrics = new EvalSchedulerMetrics();
27
+ await reconcilePersistedInitialRetryDecisions(options);
28
+ const pendingRetriesByTask = await pendingRetryWork(options);
29
+ const byTask = new Map();
30
+ for (const item of options.plan.work_items) {
31
+ const taskId = item.task_ids[0];
32
+ if (options.progress.trials.some((trial) => trial.task_id === taskId && trial.attempt === item.logical_attempt))
33
+ continue;
34
+ const items = byTask.get(taskId) || [];
35
+ items.push(item);
36
+ byTask.set(taskId, items);
37
+ }
38
+ for (const items of byTask.values())
39
+ items.sort((left, right) => left.logical_attempt - right.logical_attempt);
40
+ const results = [];
41
+ const infrastructureRetryRuns = [];
42
+ let stopDispatch = false;
43
+ let executionFailure;
44
+ const taskIds = [...new Set([...byTask.keys(), ...pendingRetriesByTask.keys()])];
45
+ const taskEntries = taskIds.map((taskId) => ({
46
+ taskId,
47
+ items: byTask.get(taskId) ?? [],
48
+ retries: pendingRetriesByTask.get(taskId) ?? [],
49
+ })).sort((left, right) => workSchedulingPriority((right.items[0] ?? right.retries[0]?.item))
50
+ - workSchedulingPriority((left.items[0] ?? left.retries[0]?.item))
51
+ || Buffer.compare(Buffer.from(left.taskId), Buffer.from(right.taskId)));
52
+ await Promise.all(taskEntries.map(async ({ taskId, items, retries: pendingRetries }) => {
53
+ for (const pending of pendingRetries) {
54
+ if (stopDispatch || options.signal?.aborted)
55
+ break;
56
+ try {
57
+ const environmentImages = await loadTrialEnvironmentImages({
58
+ taskId, uses: pending.item.image_refs ?? [],
59
+ ...(options.environmentImageManifestLoader ? { loader: options.environmentImageManifestLoader } : {}),
60
+ });
61
+ const retries = await runPlannedInfrastructureRetriesFromSource(options, publisher, semaphore, metrics, pending.item, {
62
+ attempt: pending.item.logical_attempt, workId: pending.item.work_id, tasks: [taskId], refs: [pending.trigger],
63
+ ...(environmentImages ? { environmentImages } : {}),
64
+ }, pending.decision.retry_index);
65
+ infrastructureRetryRuns.push(...retries);
66
+ }
67
+ catch (error) {
68
+ executionFailure ??= error;
69
+ stopDispatch = true;
70
+ }
71
+ }
72
+ for (const item of items) {
73
+ if (stopDispatch || options.signal?.aborted)
74
+ break;
75
+ let release;
76
+ let completed;
77
+ let initialStarted = false;
78
+ try {
79
+ options.sink.emit({
80
+ type: "eval.work.ready", work_id: item.work_id, slot_id: item.slots[0], task_id: taskId,
81
+ attempt: item.logical_attempt, priority: workSchedulingPriority(item), policy: item.scheduling?.policy ?? "fifo-compat",
82
+ });
83
+ let parentAllocationId;
84
+ if (options.admission) {
85
+ const admissionStartedAt = Date.now();
86
+ const permit = await options.admission.acquire({
87
+ evalId: options.evalId,
88
+ workItem: item,
89
+ maxParallelism: options.plan.max_parallelism,
90
+ ...(options.signal ? { signal: options.signal } : {}),
91
+ });
92
+ parentAllocationId = permit.allocationId;
93
+ release = permit.release;
94
+ const waitMs = Date.now() - admissionStartedAt;
95
+ metrics.addBlocked("resource", waitMs);
96
+ if (waitMs > 0)
97
+ options.sink.emit({
98
+ type: "eval.resource.blocked",
99
+ work_id: item.work_id,
100
+ duration_ms: waitMs,
101
+ reason: "resource-or-collision-admission",
102
+ });
103
+ options.sink.emit({
104
+ type: "eval.work-item.admitted",
105
+ work_id: item.work_id,
106
+ allocation_id: permit.allocationId,
107
+ collision_keys: permit.collisionKeys,
108
+ reservation: item.reservation,
109
+ });
110
+ }
111
+ else {
112
+ release = await semaphore?.acquire(options.signal, workSchedulingPriority(item));
113
+ }
114
+ if (stopDispatch || options.signal?.aborted)
115
+ break;
116
+ metrics.startWork(item.work_id, "initial");
117
+ initialStarted = true;
118
+ completed = await executeWorkItem(options, publisher, item, metrics, parentAllocationId);
119
+ results.push(completed);
120
+ if (completed.durationMs !== undefined) {
121
+ await recordTaskDuration({
122
+ root: options.root,
123
+ benchmarkId: options.request.benchmark_id,
124
+ benchmarkRevision: options.request.benchmark_revision,
125
+ taskId,
126
+ provider: item.provider,
127
+ model: options.request.model,
128
+ durationMs: completed.durationMs,
129
+ retryableInfrastructureFailure: completed.refs.some((ref) => physicalRetryAllowed(classifyTrialFailure(ref))),
130
+ }).catch((error) => options.sink.emit({
131
+ type: "eval.work-duration.recording-failed",
132
+ work_id: item.work_id,
133
+ task_id: taskId,
134
+ code: error.code || "duration_stats_write_failed",
135
+ }));
136
+ }
137
+ const failed = completed.run.backend.process_exit_code !== 0 || completed.run.rawResult === null
138
+ || Boolean(options.localTransport && localSourceBackendFailure(completed.run.rawResult));
139
+ options.sink.emit({
140
+ type: "eval.work-item.completed",
141
+ work_id: item.work_id,
142
+ lease_id: completed.leaseId,
143
+ task_id: taskId,
144
+ attempt: item.logical_attempt,
145
+ process_exit_code: completed.run.backend.process_exit_code,
146
+ result_available: completed.run.rawResult !== null,
147
+ });
148
+ options.sink.emit({
149
+ type: "eval.work.completed",
150
+ work_id: item.work_id,
151
+ lease_id: completed.leaseId,
152
+ task_id: taskId,
153
+ attempt: item.logical_attempt,
154
+ process_exit_code: completed.run.backend.process_exit_code,
155
+ result_available: completed.run.rawResult !== null,
156
+ });
157
+ if (failed)
158
+ stopDispatch = true;
159
+ }
160
+ catch (error) {
161
+ options.sink.emit({
162
+ type: "eval.work.lost",
163
+ work_id: item.work_id,
164
+ task_id: taskId,
165
+ attempt: item.logical_attempt,
166
+ code: error.code || "work_execution_failed",
167
+ });
168
+ executionFailure ??= error;
169
+ stopDispatch = true;
170
+ break;
171
+ }
172
+ finally {
173
+ if (initialStarted)
174
+ metrics.finishWork(item.work_id);
175
+ release?.();
176
+ }
177
+ if (!completed || stopDispatch || options.signal?.aborted)
178
+ break;
179
+ try {
180
+ const retries = await runPlannedInfrastructureRetries(options, publisher, semaphore, metrics, item, completed);
181
+ infrastructureRetryRuns.push(...retries);
182
+ }
183
+ catch (error) {
184
+ options.sink.emit({
185
+ type: "eval.work.lost",
186
+ work_id: item.work_id,
187
+ task_id: taskId,
188
+ attempt: item.logical_attempt,
189
+ execution_kind: "physical-infrastructure-retry",
190
+ code: error.code || "infrastructure_retry_failed",
191
+ });
192
+ executionFailure ??= error;
193
+ stopDispatch = true;
194
+ break;
195
+ }
196
+ }
197
+ }));
198
+ await publisher.settle();
199
+ if (executionFailure !== undefined)
200
+ throw executionFailure;
201
+ results.sort((left, right) => workOrder(options.plan, left.workId) - workOrder(options.plan, right.workId));
202
+ infrastructureRetryRuns.sort((left, right) => left.tasks.join("\0").localeCompare(right.tasks.join("\0"))
203
+ || left.attempt - right.attempt || left.retry - right.retry);
204
+ const schedulerSummary = metrics.summary({
205
+ maxParallelism: options.plan.max_parallelism,
206
+ verifierSkipped: publisher.current().trials.filter((trial) => trial.invalid_reason === "candidate_evidence_unavailable").length,
207
+ prioritized: options.plan.work_items.some((item) => item.scheduling?.policy === "critical-path-lpt-v1"),
208
+ });
209
+ options.sink.emit({ type: "eval.scheduler.summary", ...schedulerSummary });
210
+ return { progress: publisher.current(), backendRuns: results, infrastructureRetryRuns, schedulerSummary };
211
+ }
212
+ async function executeWorkItem(options, publisher, item, metrics, parentAllocationId) {
213
+ const startedAt = Date.now();
214
+ if (options.remoteWorkExecutor) {
215
+ const completed = await executeRemotePlannedWorkItem({ options, item, publish: (ref) => publisher.publish(ref, item.work_id) });
216
+ return { ...completed, durationMs: Math.max(1, Date.now() - startedAt) };
217
+ }
218
+ const lease = await createExecutionLease({
219
+ evalDirectory: options.evalDirectory,
220
+ evalId: options.evalId,
221
+ workId: item.work_id,
222
+ worker: parentAllocationId ? { ...options.worker, parentAllocationId } : options.worker,
223
+ reservation: item.reservation,
224
+ ttlMs: DEFAULT_EXECUTION_LEASE_TTL_MS,
225
+ });
226
+ const epoch = lease.current().epoch;
227
+ const providerProcess = { recorded: false };
228
+ options.sink.emit({
229
+ type: "lease.offered",
230
+ work_id: item.work_id,
231
+ lease_id: lease.leaseId,
232
+ lease_epoch: epoch,
233
+ worker_id: options.worker.workerId,
234
+ });
235
+ await lease.markRunning(epoch);
236
+ options.sink.emit({
237
+ type: "lease.accepted",
238
+ work_id: item.work_id,
239
+ lease_id: lease.leaseId,
240
+ lease_epoch: epoch,
241
+ worker_id: options.worker.workerId,
242
+ });
243
+ await options.onWorkItemState?.(item.work_id, lease.leaseId, "running");
244
+ options.sink.emit({
245
+ type: "eval.work.leased",
246
+ work_id: item.work_id,
247
+ lease_id: lease.leaseId,
248
+ lease_epoch: lease.current().epoch,
249
+ reservation: item.reservation,
250
+ });
251
+ options.sink.emit({
252
+ type: "eval.work-item.started",
253
+ work_id: item.work_id,
254
+ lease_id: lease.leaseId,
255
+ lease_epoch: lease.current().epoch,
256
+ task_id: item.task_ids[0],
257
+ attempt: item.logical_attempt,
258
+ });
259
+ options.sink.emit({
260
+ type: "eval.work.started",
261
+ work_id: item.work_id,
262
+ lease_id: lease.leaseId,
263
+ lease_epoch: lease.current().epoch,
264
+ task_id: item.task_ids[0],
265
+ attempt: item.logical_attempt,
266
+ });
267
+ let heartbeatFailure;
268
+ let heartbeatTail = Promise.resolve();
269
+ const heartbeatTimer = setInterval(() => {
270
+ if (heartbeatFailure !== undefined)
271
+ return;
272
+ heartbeatTail = heartbeatTail.then(async () => {
273
+ const renewed = await lease.heartbeat(epoch);
274
+ options.sink.emit({
275
+ type: "lease.renewed",
276
+ work_id: item.work_id,
277
+ lease_id: lease.leaseId,
278
+ lease_epoch: renewed.epoch,
279
+ heartbeat_at: renewed.heartbeat_at,
280
+ expires_at: renewed.expires_at,
281
+ });
282
+ }).catch((error) => { heartbeatFailure ??= error; });
283
+ }, DEFAULT_EXECUTION_LEASE_HEARTBEAT_MS);
284
+ heartbeatTimer.unref();
285
+ try {
286
+ const completed = await executeLeasedWorkItem(options, publisher, item, lease.current(), providerProcess, metrics);
287
+ await heartbeatTail;
288
+ if (heartbeatFailure !== undefined)
289
+ throw heartbeatFailure;
290
+ return { ...completed, leaseId: lease.leaseId, durationMs: Math.max(1, Date.now() - startedAt) };
291
+ }
292
+ finally {
293
+ clearInterval(heartbeatTimer);
294
+ await heartbeatTail;
295
+ if (providerProcess.recorded)
296
+ await releaseLocalDockerProcessRecord({ root: options.root, leaseId: lease.leaseId, epoch });
297
+ const released = await lease.release(epoch);
298
+ options.sink.emit({
299
+ type: "lease.released",
300
+ work_id: item.work_id,
301
+ lease_id: lease.leaseId,
302
+ lease_epoch: released.epoch,
303
+ worker_id: options.worker.workerId,
304
+ });
305
+ options.sink.emit({
306
+ type: "eval.work-item.lease-released",
307
+ work_id: item.work_id,
308
+ lease_id: lease.leaseId,
309
+ lease_epoch: released.epoch,
310
+ state: released.state,
311
+ });
312
+ if (options.dockerResourceReaper) {
313
+ options.sink.emit({ type: "sandbox.cleanup.started", work_id: item.work_id, lease_id: lease.leaseId });
314
+ try {
315
+ const report = await options.dockerResourceReaper({
316
+ root: options.root,
317
+ leaseIds: [lease.leaseId],
318
+ env: options.env,
319
+ });
320
+ options.sink.emit({ type: "eval.work-item.resources-reaped", work_id: item.work_id, lease_id: lease.leaseId, scanned: report.scanned, deleted: report.deleted.length, issues: report.issues.length });
321
+ options.sink.emit({ type: "sandbox.cleanup.completed", work_id: item.work_id, lease_id: lease.leaseId, scanned: report.scanned, deleted: report.deleted.length, residual_resources: report.issues.length });
322
+ }
323
+ catch (error) {
324
+ options.sink.emit({ type: "eval.work-item.reaper-failed", work_id: item.work_id, lease_id: lease.leaseId, code: error.code || "docker_reaper_failed" });
325
+ options.sink.emit({ type: "sandbox.cleanup.failed", work_id: item.work_id, lease_id: lease.leaseId, code: error.code || "docker_reaper_failed" });
326
+ }
327
+ }
328
+ await options.onWorkItemState?.(item.work_id, lease.leaseId, "terminal");
329
+ }
330
+ }
331
+ async function executeLeasedWorkItem(options, publisher, item, lease, providerProcess, metrics) {
332
+ const logicalAttempt = item.logical_attempt;
333
+ const taskId = item.task_ids[0];
334
+ const environmentImages = await loadTrialEnvironmentImages({ taskId, uses: item.image_refs ?? [], ...(options.environmentImageManifestLoader ? { loader: options.environmentImageManifestLoader } : {}) });
335
+ const runtimeResources = runtimeResourcesForTask(options.plan, taskId, item.reservation);
336
+ const ownership = dockerResourceOwnership(options.root, lease, taskId);
337
+ const resourceObserver = startDockerResourceObserver({ ownership, workerId: options.worker.workerId, collisionDomainId: options.worker.collisionDomainId, reservation: item.reservation, mainLimits: runtimeResources.mainLimits, sidecarLimits: runtimeResources.sidecarLimits, env: options.env, ...(options.signal ? { signal: options.signal } : {}) });
338
+ const backendDirectory = path.join(options.evalDirectory, "harbor", "work-items", item.work_id, `epoch-${String(lease.epoch).padStart(6, "0")}`);
339
+ const harborJobDirectory = path.join(backendDirectory, "job");
340
+ const prebuiltImage = prebuiltTaskImage(environmentImages);
341
+ const refs = [];
342
+ const publish = async (ref) => {
343
+ if (ref.attempt !== logicalAttempt || ref.task_id !== taskId) {
344
+ throw new HitchError(`Harbor work item returned an unselected trial: ${ref.task_id}#${ref.attempt}`, {
345
+ code: "eval_work_item_trial_mismatch",
346
+ exitCode: 12,
347
+ });
348
+ }
349
+ const existing = refs.find((current) => current.trial_id === ref.trial_id);
350
+ if (existing) {
351
+ if (JSON.stringify(existing) !== JSON.stringify(ref))
352
+ throw new Error(`work item trial identity changed: ${ref.trial_id}`);
353
+ return;
354
+ }
355
+ refs.push(ref);
356
+ if (ref.invalid_reason === "candidate_evidence_unavailable")
357
+ options.sink.emit({
358
+ type: "eval.verifier.skipped", work_id: item.work_id, trial_id: ref.trial_id, task_id: ref.task_id,
359
+ reason: "candidate_evidence_unavailable", candidate_executes: true, verifier_executes: false,
360
+ });
361
+ await publisher.publish(ref, item.work_id);
362
+ };
363
+ const run = await runHarborBackend({
364
+ evalId: options.evalId,
365
+ evalDirectory: options.evalDirectory,
366
+ backendDirectory,
367
+ logicalAttempt,
368
+ taskNames: [taskId],
369
+ request: { ...options.request, attempts: 1, max_concurrent: 1 },
370
+ root: options.root,
371
+ resolvedRevision: options.resolvedRevision,
372
+ runtimeDirectory: options.controllerRuntime.directory,
373
+ runtimeId: options.controllerRuntime.runtime_id,
374
+ preparedArtifact: preparedArtifactForWorkItem(options, item),
375
+ ...(options.interactionCaptureExporter ? { modelProxy: options.interactionCaptureExporter.route } : {}),
376
+ executionResources: runtimeResources.mainLimits,
377
+ ...(Object.keys(runtimeResources.sidecarLimits).length > 0 ? { dockerServiceLimits: runtimeResources.sidecarLimits } : {}),
378
+ dockerOwnership: ownership,
379
+ resolvedImages: resolvedImageMapping(item.image_refs ?? []),
380
+ ...(prebuiltImage ? { prebuiltTaskImage: prebuiltImage } : {}),
381
+ env: options.env,
382
+ ...(options.harborExecutable !== undefined ? { harborExecutable: options.harborExecutable } : {}),
383
+ ...(options.signal ? { signal: options.signal } : {}),
384
+ ...(options.trialBundleGraceMs === undefined ? {} : { trialBundleGraceMs: options.trialBundleGraceMs }),
385
+ ...(options.worker.provider === "local-docker" && process.platform !== "win32" ? {
386
+ recoverableProcess: true,
387
+ onProcessStarted: (pid) => recordLocalDockerProcessStart({
388
+ root: options.root,
389
+ workerId: options.worker.workerId,
390
+ evalDirectory: options.evalDirectory,
391
+ lease,
392
+ backendDirectory,
393
+ pid,
394
+ }).then(() => { providerProcess.recorded = true; }),
395
+ onProcessExited: (result) => recordLocalDockerProcessExit({
396
+ root: options.root,
397
+ leaseId: lease.lease_id,
398
+ epoch: lease.epoch,
399
+ ...result,
400
+ }).then(() => undefined),
401
+ } : {}),
402
+ emit: (event) => options.sink.emit({ ...event, work_id: item.work_id, task_id: taskId }),
403
+ onTrialSettled: async (trial, context) => {
404
+ try {
405
+ const ref = await importEvalTrialRun({
406
+ root: options.root,
407
+ evalId: options.evalId,
408
+ evalDirectory: options.evalDirectory,
409
+ harborJobDirectory,
410
+ expectedAttempt: logicalAttempt,
411
+ request: options.request,
412
+ resolvedRevision: options.resolvedRevision,
413
+ benchmarkId: options.request.benchmark_id,
414
+ benchmarkRevision: options.request.benchmark_revision,
415
+ runtimeId: options.controllerRuntime.runtime_id,
416
+ env: options.env, ...(options.signal ? { signal: options.signal } : {}),
417
+ ...(options.plan.model_capture ? { modelCapturePlan: options.plan.model_capture } : {}),
418
+ ...(options.interactionCaptureExporter ? {
419
+ interactionCaptureExporter: options.interactionCaptureExporter,
420
+ } : {}),
421
+ executionEvidence: verifyTrialEnvironmentImageExecution(await resourceObserver.capture(), environmentImages),
422
+ ...(environmentImages ? { environmentImages } : {}),
423
+ requireCompleteMarker: true,
424
+ allowMissingBundleDiagnostic: context.bundleWaitExpired,
425
+ }, trial, refs.length, refs);
426
+ await publish(ref);
427
+ return true;
428
+ }
429
+ catch (error) {
430
+ if (error instanceof TrialBundlePendingError)
431
+ return false;
432
+ throw error;
433
+ }
434
+ },
435
+ }).finally(async () => { await resourceObserver.stop(); });
436
+ const collectionStartedAt = Date.now();
437
+ for (const event of await harborPhaseTimingEvents(harborJobDirectory, run.rawResult)) {
438
+ options.sink.emit(event);
439
+ if (event.type === "eval.verifier.completed" && typeof event.duration_ms === "number")
440
+ metrics.addVerifier(event.duration_ms);
441
+ }
442
+ const terminalRefs = await importEvalTrialRuns({
443
+ root: options.root,
444
+ evalId: options.evalId,
445
+ evalDirectory: options.evalDirectory,
446
+ harborJobDirectory,
447
+ expectedAttempt: logicalAttempt,
448
+ request: options.request,
449
+ resolvedRevision: options.resolvedRevision,
450
+ benchmarkId: options.request.benchmark_id,
451
+ benchmarkRevision: options.request.benchmark_revision,
452
+ runtimeId: options.controllerRuntime.runtime_id,
453
+ env: options.env, ...(options.signal ? { signal: options.signal } : {}),
454
+ ...(options.plan.model_capture ? { modelCapturePlan: options.plan.model_capture } : {}),
455
+ ...(options.interactionCaptureExporter ? {
456
+ interactionCaptureExporter: options.interactionCaptureExporter,
457
+ } : {}),
458
+ executionEvidence: verifyTrialEnvironmentImageExecution(await resourceObserver.capture(), environmentImages),
459
+ ...(environmentImages ? { environmentImages } : {}),
460
+ rawResult: run.rawResult,
461
+ }, refs);
462
+ for (const ref of terminalRefs)
463
+ await publish(ref);
464
+ options.sink.emit({ type: "eval.collection.completed", work_id: item.work_id, duration_ms: Date.now() - collectionStartedAt });
465
+ if (run.rawResult !== null)
466
+ assertBackendTrialSet(run.rawResult, refs);
467
+ return { attempt: logicalAttempt, workId: item.work_id, tasks: [taskId], refs, run, ...(environmentImages ? { environmentImages } : {}) };
468
+ }
469
+ //# sourceMappingURL=planned-execution.js.map
@@ -0,0 +1 @@
1
+ {"version":3,"file":"planned-execution.js","sourceRoot":"","sources":["../../../src/evals/planned-execution.ts"],"names":[],"mappings":"AAAA,OAAO,EAAE,iBAAiB,EAAE,MAAM,iCAAiC,CAAC;AACpE,OAAO,IAAI,MAAM,WAAW,CAAC;AAE7B,OAAO,EAAE,gBAAgB,EAAE,MAAM,sBAAsB,CAAC;AAIxD,OAAO,EAAE,UAAU,EAAY,MAAM,wBAAwB,CAAC;AAE9D,OAAO,EAAE,oCAAoC,EAAE,8BAA8B,EAAE,oBAAoB,EAAE,MAAM,uBAAuB,CAAC;AAEnI,OAAO,EAAE,4BAA4B,EAAE,6BAA6B,EAAE,+BAA+B,EAAE,MAAM,4BAA4B,CAAC;AAC1I,OAAO,EAAE,uBAAuB,EAAE,MAAM,uBAAuB,CAAC;AAEhE,OAAO,EAAE,qBAAqB,EAAE,yBAAyB,EAAE,MAAM,qBAAqB,CAAC;AACvF,OAAO,EAAE,kBAAkB,EAAE,mBAAmB,EAAE,uBAAuB,EAAE,MAAM,mBAAmB,CAAC;AAErG,OAAO,EAAE,4BAA4B,EAAE,MAAM,+BAA+B,CAAC;AAC7E,OAAO,EAAE,uBAAuB,EAAE,MAAM,+BAA+B,CAAC;AACxE,OAAO,EAAE,2BAA2B,EAAE,MAAM,+BAA+B,CAAC;AAC5E,OAAO,EAAE,oBAAoB,EAAE,MAAM,iCAAiC,CAAC;AACvE,OAAO,EAAE,0BAA0B,EAAE,iBAAiB,EAAE,oCAAoC,EAAE,MAAM,iCAAiC,CAAC;AAEtI,OAAO,EAAE,iBAAiB,EAAE,kBAAkB,EAAE,SAAS,EAAE,sBAAsB,EAAE,MAAM,gCAAgC,CAAC;AAG1H,OAAO,EAAE,uBAAuB,EAAE,MAAM,2BAA2B,CAAC;AACpE,OAAO,EAAE,2BAA2B,EAAE,MAAM,0BAA0B,CAAC;AACvE,OAAO,EAAE,kBAAkB,EAAE,MAAM,yBAAyB,CAAC;AAC7D,OAAO,EAAE,oBAAoB,EAAE,oBAAoB,EAAE,MAAM,yBAAyB,CAAC;AACrF,OAAO,EAAE,oBAAoB,EAAE,MAAM,wBAAwB,CAAC;AAC9D,OAAO,EAAE,gBAAgB,EAAE,uCAAuC,EAAE,+BAA+B,EAAE,yCAAyC,EAAE,MAAM,8BAA8B,CAAC;AA6BrL,MAAM,CAAC,KAAK,UAAU,yBAAyB,CAAC,OAAoC;IAMlF,kBAAkB,CAAC,OAAO,CAAC,IAAI,CAAC,CAAC;IACjC,MAAM,SAAS,GAAG,IAAI,iBAAiB,CAAC,OAAO,CAAC,QAAQ,EAAE,OAAO,CAAC,CAAC;IACnE,MAAM,SAAS,GAAG,OAAO,CAAC,SAAS,CAAC,CAAC,CAAC,SAAS,CAAC,CAAC,CAAC,IAAI,iBAAiB,CAAC,OAAO,CAAC,IAAI,CAAC,eAAe,CAAC,CAAC;IACtG,MAAM,OAAO,GAAG,IAAI,oBAAoB,EAAE,CAAC;IAC3C,MAAM,uCAAuC,CAAC,OAAO,CAAC,CAAC;IACvD,MAAM,oBAAoB,GAAG,MAAM,gBAAgB,CAAC,OAAO,CAAC,CAAC;IAC7D,MAAM,MAAM,GAAG,IAAI,GAAG,EAA+B,CAAC;IACtD,KAAK,MAAM,IAAI,IAAI,OAAO,CAAC,IAAI,CAAC,UAAU,EAAE,CAAC;QAC3C,MAAM,MAAM,GAAG,IAAI,CAAC,QAAQ,CAAC,CAAC,CAAW,CAAC;QAC1C,IAAI,OAAO,CAAC,QAAQ,CAAC,MAAM,CAAC,IAAI,CAAC,CAAC,KAAK,EAAE,EAAE,CAAC,KAAK,CAAC,OAAO,KAAK,MAAM,IAAI,KAAK,CAAC,OAAO,KAAK,IAAI,CAAC,eAAe,CAAC;YAAE,SAAS;QAC1H,MAAM,KAAK,GAAG,MAAM,CAAC,GAAG,CAAC,MAAM,CAAC,IAAI,EAAE,CAAC;QACvC,KAAK,CAAC,IAAI,CAAC,IAAI,CAAC,CAAC;QACjB,MAAM,CAAC,GAAG,CAAC,MAAM,EAAE,KAAK,CAAC,CAAC;IAC5B,CAAC;IACD,KAAK,MAAM,KAAK,IAAI,MAAM,CAAC,MAAM,EAAE;QAAE,KAAK,CAAC,IAAI,CAAC,CAAC,IAAI,EAAE,KAAK,EAAE,EAAE,CAAE,IAAI,CAAC,eAA0B,GAAI,KAAK,CAAC,eAA0B,CAAC,CAAC;IACvI,MAAM,OAAO,GAAwB,EAAE,CAAC;IACxC,MAAM,uBAAuB,GAA6B,EAAE,CAAC;IAC7D,IAAI,YAAY,GAAG,KAAK,CAAC;IACzB,IAAI,gBAAyB,CAAC;IAC9B,MAAM,OAAO,GAAG,CAAC,GAAG,IAAI,GAAG,CAAC,CAAC,GAAG,MAAM,CAAC,IAAI,EAAE,EAAE,GAAG,oBAAoB,CAAC,IAAI,EAAE,CAAC,CAAC,CAAC,CAAC;IACjF,MAAM,WAAW,GAAG,OAAO,CAAC,GAAG,CAAC,CAAC,MAAM,EAAE,EAAE,CAAC,CAAC;QAC3C,MAAM;QACN,KAAK,EAAE,MAAM,CAAC,GAAG,CAAC,MAAM,CAAC,IAAI,EAAE;QAC/B,OAAO,EAAE,oBAAoB,CAAC,GAAG,CAAC,MAAM,CAAC,IAAI,EAAE;KAChD,CAAC,CAAC,CAAC,IAAI,CAAC,CAAC,IAAI,EAAE,KAAK,EAAE,EAAE,CAAC,sBAAsB,CAAC,CAAC,KAAK,CAAC,KAAK,CAAC,CAAC,CAAC,IAAI,KAAK,CAAC,OAAO,CAAC,CAAC,CAAC,EAAE,IAAI,CAAsB,CAAC;UAC7G,sBAAsB,CAAC,CAAC,IAAI,CAAC,KAAK,CAAC,CAAC,CAAC,IAAI,IAAI,CAAC,OAAO,CAAC,CAAC,CAAC,EAAE,IAAI,CAAsB,CAAC;WACpF,MAAM,CAAC,OAAO,CAAC,MAAM,CAAC,IAAI,CAAC,IAAI,CAAC,MAAM,CAAC,EAAE,MAAM,CAAC,IAAI,CAAC,KAAK,CAAC,MAAM,CAAC,CAAC,CAAC,CAAC;IAC1E,MAAM,OAAO,CAAC,GAAG,CAAC,WAAW,CAAC,GAAG,CAAC,KAAK,EAAE,EAAE,MAAM,EAAE,KAAK,EAAE,OAAO,EAAE,cAAc,EAAE,EAAE,EAAE;QACrF,KAAK,MAAM,OAAO,IAAI,cAAc,EAAE,CAAC;YACrC,IAAI,YAAY,IAAI,OAAO,CAAC,MAAM,EAAE,OAAO;gBAAE,MAAM;YACnD,IAAI,CAAC;gBACH,MAAM,iBAAiB,GAAG,MAAM,0BAA0B,CAAC;oBACzD,MAAM,EAAE,IAAI,EAAE,OAAO,CAAC,IAAI,CAAC,UAAU,IAAI,EAAE;oBAC3C,GAAG,CAAC,OAAO,CAAC,8BAA8B,CAAC,CAAC,CAAC,EAAE,MAAM,EAAE,OAAO,CAAC,8BAA8B,EAAE,CAAC,CAAC,CAAC,EAAE,CAAC;iBACtG,CAAC,CAAC;gBACH,MAAM,OAAO,GAAG,MAAM,yCAAyC,CAAC,OAAO,EAAE,SAAS,EAAE,SAAS,EAAE,OAAO,EAAE,OAAO,CAAC,IAAI,EAAE;oBACpH,OAAO,EAAE,OAAO,CAAC,IAAI,CAAC,eAAyB,EAAE,MAAM,EAAE,OAAO,CAAC,IAAI,CAAC,OAAO,EAAE,KAAK,EAAE,CAAC,MAAM,CAAC,EAAE,IAAI,EAAE,CAAC,OAAO,CAAC,OAAO,CAAC;oBACvH,GAAG,CAAC,iBAAiB,CAAC,CAAC,CAAC,EAAE,iBAAiB,EAAE,CAAC,CAAC,CAAC,EAAE,CAAC;iBACpD,EAAE,OAAO,CAAC,QAAQ,CAAC,WAAW,CAAC,CAAC;gBACjC,uBAAuB,CAAC,IAAI,CAAC,GAAG,OAAO,CAAC,CAAC;YAC3C,CAAC;YAAC,OAAO,KAAK,EAAE,CAAC;gBACf,gBAAgB,KAAK,KAAK,CAAC;gBAC3B,YAAY,GAAG,IAAI,CAAC;YACtB,CAAC;QACH,CAAC;QACD,KAAK,MAAM,IAAI,IAAI,KAAK,EAAE,CAAC;YACzB,IAAI,YAAY,IAAI,OAAO,CAAC,MAAM,EAAE,OAAO;gBAAE,MAAM;YACnD,IAAI,OAAiC,CAAC;YACtC,IAAI,SAAwC,CAAC;YAC7C,IAAI,cAAc,GAAG,KAAK,CAAC;YAC3B,IAAI,CAAC;gBACH,OAAO,CAAC,IAAI,CAAC,IAAI,CAAC;oBAChB,IAAI,EAAE,iBAAiB,EAAE,OAAO,EAAE,IAAI,CAAC,OAAO,EAAE,OAAO,EAAE,IAAI,CAAC,KAAK,CAAC,CAAC,CAAC,EAAE,OAAO,EAAE,MAAM;oBACvF,OAAO,EAAE,IAAI,CAAC,eAAe,EAAE,QAAQ,EAAE,sBAAsB,CAAC,IAAI,CAAC,EAAE,MAAM,EAAE,IAAI,CAAC,UAAU,EAAE,MAAM,IAAI,aAAa;iBACxH,CAAC,CAAC;gBACH,IAAI,kBAAsC,CAAC;gBAC3C,IAAI,OAAO,CAAC,SAAS,EAAE,CAAC;oBACtB,MAAM,kBAAkB,GAAG,IAAI,CAAC,GAAG,EAAE,CAAC;oBACtC,MAAM,MAAM,GAAG,MAAM,OAAO,CAAC,SAAS,CAAC,OAAO,CAAC;wBAC7C,MAAM,EAAE,OAAO,CAAC,MAAM;wBACtB,QAAQ,EAAE,IAAI;wBACd,cAAc,EAAE,OAAO,CAAC,IAAI,CAAC,eAAe;wBAC5C,GAAG,CAAC,OAAO,CAAC,MAAM,CAAC,CAAC,CAAC,EAAE,MAAM,EAAE,OAAO,CAAC,MAAM,EAAE,CAAC,CAAC,CAAC,EAAE,CAAC;qBACtD,CAAC,CAAC;oBACH,kBAAkB,GAAG,MAAM,CAAC,YAAY,CAAC;oBACzC,OAAO,GAAG,MAAM,CAAC,OAAO,CAAC;oBACzB,MAAM,MAAM,GAAG,IAAI,CAAC,GAAG,EAAE,GAAG,kBAAkB,CAAC;oBAC/C,OAAO,CAAC,UAAU,CAAC,UAAU,EAAE,MAAM,CAAC,CAAC;oBACvC,IAAI,MAAM,GAAG,CAAC;wBAAE,OAAO,CAAC,IAAI,CAAC,IAAI,CAAC;4BAChC,IAAI,EAAE,uBAAuB;4BAC7B,OAAO,EAAE,IAAI,CAAC,OAAO;4BACrB,WAAW,EAAE,MAAM;4BACnB,MAAM,EAAE,iCAAiC;yBAC1C,CAAC,CAAC;oBACH,OAAO,CAAC,IAAI,CAAC,IAAI,CAAC;wBAChB,IAAI,EAAE,yBAAyB;wBAC/B,OAAO,EAAE,IAAI,CAAC,OAAO;wBACrB,aAAa,EAAE,MAAM,CAAC,YAAY;wBAClC,cAAc,EAAE,MAAM,CAAC,aAAa;wBACpC,WAAW,EAAE,IAAI,CAAC,WAAW;qBAC9B,CAAC,CAAC;gBACL,CAAC;qBAAM,CAAC;oBACN,OAAO,GAAG,MAAM,SAAS,EAAE,OAAO,CAAC,OAAO,CAAC,MAAM,EAAE,sBAAsB,CAAC,IAAI,CAAC,CAAC,CAAC;gBACnF,CAAC;gBACD,IAAI,YAAY,IAAI,OAAO,CAAC,MAAM,EAAE,OAAO;oBAAE,MAAM;gBACnD,OAAO,CAAC,SAAS,CAAC,IAAI,CAAC,OAAO,EAAE,SAAS,CAAC,CAAC;gBAC3C,cAAc,GAAG,IAAI,CAAC;gBACtB,SAAS,GAAG,MAAM,eAAe,CAAC,OAAO,EAAE,SAAS,EAAE,IAAI,EAAE,OAAO,EAAE,kBAAkB,CAAC,CAAC;gBACzF,OAAO,CAAC,IAAI,CAAC,SAAS,CAAC,CAAC;gBACxB,IAAI,SAAS,CAAC,UAAU,KAAK,SAAS,EAAE,CAAC;oBACvC,MAAM,kBAAkB,CAAC;wBACvB,IAAI,EAAE,OAAO,CAAC,IAAI;wBAClB,WAAW,EAAE,OAAO,CAAC,OAAO,CAAC,YAAY;wBACzC,iBAAiB,EAAE,OAAO,CAAC,OAAO,CAAC,kBAAkB;wBACrD,MAAM;wBACN,QAAQ,EAAE,IAAI,CAAC,QAAQ;wBACvB,KAAK,EAAE,OAAO,CAAC,OAAO,CAAC,KAAK;wBAC5B,UAAU,EAAE,SAAS,CAAC,UAAU;wBAChC,8BAA8B,EAAE,SAAS,CAAC,IAAI,CAAC,IAAI,CAAC,CAAC,GAAG,EAAE,EAAE,CAAC,oBAAoB,CAAC,oBAAoB,CAAC,GAAG,CAAC,CAAC,CAAC;qBAC9G,CAAC,CAAC,KAAK,CAAC,CAAC,KAAK,EAAE,EAAE,CAAC,OAAO,CAAC,IAAI,CAAC,IAAI,CAAC;wBACpC,IAAI,EAAE,qCAAqC;wBAC3C,OAAO,EAAE,IAAI,CAAC,OAAO;wBACrB,OAAO,EAAE,MAAM;wBACf,IAAI,EAAG,KAA2B,CAAC,IAAI,IAAI,6BAA6B;qBACzE,CAAC,CAAC,CAAC;gBACN,CAAC;gBACD,MAAM,MAAM,GAAG,SAAS,CAAC,GAAG,CAAC,OAAO,CAAC,iBAAiB,KAAK,CAAC,IAAI,SAAS,CAAC,GAAG,CAAC,SAAS,KAAK,IAAI;uBAC3F,OAAO,CAAC,OAAO,CAAC,cAAc,IAAI,yBAAyB,CAAC,SAAS,CAAC,GAAG,CAAC,SAAS,CAAC,CAAC,CAAC;gBAC3F,OAAO,CAAC,IAAI,CAAC,IAAI,CAAC;oBAChB,IAAI,EAAE,0BAA0B;oBAChC,OAAO,EAAE,IAAI,CAAC,OAAO;oBACrB,QAAQ,EAAE,SAAS,CAAC,OAAO;oBAC3B,OAAO,EAAE,MAAM;oBACf,OAAO,EAAE,IAAI,CAAC,eAAe;oBAC7B,iBAAiB,EAAE,SAAS,CAAC,GAAG,CAAC,OAAO,CAAC,iBAAiB;oBAC1D,gBAAgB,EAAE,SAAS,CAAC,GAAG,CAAC,SAAS,KAAK,IAAI;iBACnD,CAAC,CAAC;gBACH,OAAO,CAAC,IAAI,CAAC,IAAI,CAAC;oBAChB,IAAI,EAAE,qBAAqB;oBAC3B,OAAO,EAAE,IAAI,CAAC,OAAO;oBACrB,QAAQ,EAAE,SAAS,CAAC,OAAO;oBAC3B,OAAO,EAAE,MAAM;oBACf,OAAO,EAAE,IAAI,CAAC,eAAe;oBAC7B,iBAAiB,EAAE,SAAS,CAAC,GAAG,CAAC,OAAO,CAAC,iBAAiB;oBAC1D,gBAAgB,EAAE,SAAS,CAAC,GAAG,CAAC,SAAS,KAAK,IAAI;iBACnD,CAAC,CAAC;gBACH,IAAI,MAAM;oBAAE,YAAY,GAAG,IAAI,CAAC;YAClC,CAAC;YAAC,OAAO,KAAK,EAAE,CAAC;gBACf,OAAO,CAAC,IAAI,CAAC,IAAI,CAAC;oBAChB,IAAI,EAAE,gBAAgB;oBACtB,OAAO,EAAE,IAAI,CAAC,OAAO;oBACrB,OAAO,EAAE,MAAM;oBACf,OAAO,EAAE,IAAI,CAAC,eAAe;oBAC7B,IAAI,EAAG,KAA2B,CAAC,IAAI,IAAI,uBAAuB;iBACnE,CAAC,CAAC;gBACH,gBAAgB,KAAK,KAAK,CAAC;gBAC3B,YAAY,GAAG,IAAI,CAAC;gBACpB,MAAM;YACR,CAAC;oBAAS,CAAC;gBACT,IAAI,cAAc;oBAAE,OAAO,CAAC,UAAU,CAAC,IAAI,CAAC,OAAO,CAAC,CAAC;gBACrD,OAAO,EAAE,EAAE,CAAC;YACd,CAAC;YACD,IAAI,CAAC,SAAS,IAAI,YAAY,IAAI,OAAO,CAAC,MAAM,EAAE,OAAO;gBAAE,MAAM;YACjE,IAAI,CAAC;gBACH,MAAM,OAAO,GAAG,MAAM,+BAA+B,CAAC,OAAO,EAAE,SAAS,EAAE,SAAS,EAAE,OAAO,EAAE,IAAI,EAAE,SAAS,CAAC,CAAC;gBAC/G,uBAAuB,CAAC,IAAI,CAAC,GAAG,OAAO,CAAC,CAAC;YAC3C,CAAC;YAAC,OAAO,KAAK,EAAE,CAAC;gBACf,OAAO,CAAC,IAAI,CAAC,IAAI,CAAC;oBAChB,IAAI,EAAE,gBAAgB;oBACtB,OAAO,EAAE,IAAI,CAAC,OAAO;oBACrB,OAAO,EAAE,MAAM;oBACf,OAAO,EAAE,IAAI,CAAC,eAAe;oBAC7B,cAAc,EAAE,+BAA+B;oBAC/C,IAAI,EAAG,KAA2B,CAAC,IAAI,IAAI,6BAA6B;iBACzE,CAAC,CAAC;gBACH,gBAAgB,KAAK,KAAK,CAAC;gBAC3B,YAAY,GAAG,IAAI,CAAC;gBACpB,MAAM;YACR,CAAC;QACH,CAAC;IACH,CAAC,CAAC,CAAC,CAAC;IACJ,MAAM,SAAS,CAAC,MAAM,EAAE,CAAC;IACzB,IAAI,gBAAgB,KAAK,SAAS;QAAE,MAAM,gBAAgB,CAAC;IAC3D,OAAO,CAAC,IAAI,CAAC,CAAC,IAAI,EAAE,KAAK,EAAE,EAAE,CAAC,SAAS,CAAC,OAAO,CAAC,IAAI,EAAE,IAAI,CAAC,MAAM,CAAC,GAAG,SAAS,CAAC,OAAO,CAAC,IAAI,EAAE,KAAK,CAAC,MAAM,CAAC,CAAC,CAAC;IAC5G,uBAAuB,CAAC,IAAI,CAAC,CAAC,IAAI,EAAE,KAAK,EAAE,EAAE,CAAC,IAAI,CAAC,KAAK,CAAC,IAAI,CAAC,IAAI,CAAC,CAAC,aAAa,CAAC,KAAK,CAAC,KAAK,CAAC,IAAI,CAAC,IAAI,CAAC,CAAC;WACpG,IAAI,CAAC,OAAO,GAAG,KAAK,CAAC,OAAO,IAAI,IAAI,CAAC,KAAK,GAAG,KAAK,CAAC,KAAK,CAAC,CAAC;IAC/D,MAAM,gBAAgB,GAAG,OAAO,CAAC,OAAO,CAAC;QACvC,cAAc,EAAE,OAAO,CAAC,IAAI,CAAC,eAAe;QAC5C,eAAe,EAAE,SAAS,CAAC,OAAO,EAAE,CAAC,MAAM,CAAC,MAAM,CAAC,CAAC,KAAK,EAAE,EAAE,CAAC,KAAK,CAAC,cAAc,KAAK,gCAAgC,CAAC,CAAC,MAAM;QAC/H,WAAW,EAAE,OAAO,CAAC,IAAI,CAAC,UAAU,CAAC,IAAI,CAAC,CAAC,IAAI,EAAE,EAAE,CAAC,IAAI,CAAC,UAAU,EAAE,MAAM,KAAK,sBAAsB,CAAC;KACxG,CAAC,CAAC;IACH,OAAO,CAAC,IAAI,CAAC,IAAI,CAAC,EAAE,IAAI,EAAE,wBAAwB,EAAE,GAAG,gBAAgB,EAAE,CAAC,CAAC;IAC3E,OAAO,EAAE,QAAQ,EAAE,SAAS,CAAC,OAAO,EAAE,EAAE,WAAW,EAAE,OAAO,EAAE,uBAAuB,EAAE,gBAAgB,EAAE,CAAC;AAC5G,CAAC;AAED,KAAK,UAAU,eAAe,CAC5B,OAAoC,EACpC,SAA4B,EAC5B,IAAuB,EACvB,OAA6B,EAC7B,kBAA2B;IAE3B,MAAM,SAAS,GAAG,IAAI,CAAC,GAAG,EAAE,CAAC;IAC7B,IAAI,OAAO,CAAC,kBAAkB,EAAE,CAAC;QAC/B,MAAM,SAAS,GAAG,MAAM,4BAA4B,CAAC,EAAE,OAAO,EAAE,IAAI,EAAE,OAAO,EAAE,CAAC,GAAG,EAAE,EAAE,CAAC,SAAS,CAAC,OAAO,CAAC,GAAG,EAAE,IAAI,CAAC,OAAO,CAAC,EAAE,CAAC,CAAC;QAChI,OAAO,EAAE,GAAG,SAAS,EAAE,UAAU,EAAE,IAAI,CAAC,GAAG,CAAC,CAAC,EAAE,IAAI,CAAC,GAAG,EAAE,GAAG,SAAS,CAAC,EAAE,CAAC;IAC3E,CAAC;IACD,MAAM,KAAK,GAAG,MAAM,oBAAoB,CAAC;QACvC,aAAa,EAAE,OAAO,CAAC,aAAa;QACpC,MAAM,EAAE,OAAO,CAAC,MAAM;QACtB,MAAM,EAAE,IAAI,CAAC,OAAO;QACpB,MAAM,EAAE,kBAAkB,CAAC,CAAC,CAAC,EAAE,GAAG,OAAO,CAAC,MAAM,EAAE,kBAAkB,EAAE,CAAC,CAAC,CAAC,OAAO,CAAC,MAAM;QACvF,WAAW,EAAE,IAAI,CAAC,WAAW;QAC7B,KAAK,EAAE,8BAA8B;KACtC,CAAC,CAAC;IACH,MAAM,KAAK,GAAG,KAAK,CAAC,OAAO,EAAE,CAAC,KAAK,CAAC;IACpC,MAAM,eAAe,GAAG,EAAE,QAAQ,EAAE,KAAK,EAAE,CAAC;IAC5C,OAAO,CAAC,IAAI,CAAC,IAAI,CAAC;QAChB,IAAI,EAAE,eAAe;QACrB,OAAO,EAAE,IAAI,CAAC,OAAO;QACrB,QAAQ,EAAE,KAAK,CAAC,OAAO;QACvB,WAAW,EAAE,KAAK;QAClB,SAAS,EAAE,OAAO,CAAC,MAAM,CAAC,QAAQ;KACnC,CAAC,CAAC;IACH,MAAM,KAAK,CAAC,WAAW,CAAC,KAAK,CAAC,CAAC;IAC/B,OAAO,CAAC,IAAI,CAAC,IAAI,CAAC;QAChB,IAAI,EAAE,gBAAgB;QACtB,OAAO,EAAE,IAAI,CAAC,OAAO;QACrB,QAAQ,EAAE,KAAK,CAAC,OAAO;QACvB,WAAW,EAAE,KAAK;QAClB,SAAS,EAAE,OAAO,CAAC,MAAM,CAAC,QAAQ;KACnC,CAAC,CAAC;IACH,MAAM,OAAO,CAAC,eAAe,EAAE,CAAC,IAAI,CAAC,OAAO,EAAE,KAAK,CAAC,OAAO,EAAE,SAAS,CAAC,CAAC;IACxE,OAAO,CAAC,IAAI,CAAC,IAAI,CAAC;QAChB,IAAI,EAAE,kBAAkB;QACxB,OAAO,EAAE,IAAI,CAAC,OAAO;QACrB,QAAQ,EAAE,KAAK,CAAC,OAAO;QACvB,WAAW,EAAE,KAAK,CAAC,OAAO,EAAE,CAAC,KAAK;QAClC,WAAW,EAAE,IAAI,CAAC,WAAW;KAC9B,CAAC,CAAC;IACH,OAAO,CAAC,IAAI,CAAC,IAAI,CAAC;QAChB,IAAI,EAAE,wBAAwB;QAC9B,OAAO,EAAE,IAAI,CAAC,OAAO;QACrB,QAAQ,EAAE,KAAK,CAAC,OAAO;QACvB,WAAW,EAAE,KAAK,CAAC,OAAO,EAAE,CAAC,KAAK;QAClC,OAAO,EAAE,IAAI,CAAC,QAAQ,CAAC,CAAC,CAAC;QACzB,OAAO,EAAE,IAAI,CAAC,eAAe;KAC9B,CAAC,CAAC;IACH,OAAO,CAAC,IAAI,CAAC,IAAI,CAAC;QAChB,IAAI,EAAE,mBAAmB;QACzB,OAAO,EAAE,IAAI,CAAC,OAAO;QACrB,QAAQ,EAAE,KAAK,CAAC,OAAO;QACvB,WAAW,EAAE,KAAK,CAAC,OAAO,EAAE,CAAC,KAAK;QAClC,OAAO,EAAE,IAAI,CAAC,QAAQ,CAAC,CAAC,CAAC;QACzB,OAAO,EAAE,IAAI,CAAC,eAAe;KAC9B,CAAC,CAAC;IACH,IAAI,gBAAyB,CAAC;IAC9B,IAAI,aAAa,GAAkB,OAAO,CAAC,OAAO,EAAE,CAAC;IACrD,MAAM,cAAc,GAAG,WAAW,CAAC,GAAG,EAAE;QACtC,IAAI,gBAAgB,KAAK,SAAS;YAAE,OAAO;QAC3C,aAAa,GAAG,aAAa,CAAC,IAAI,CAAC,KAAK,IAAI,EAAE;YAC5C,MAAM,OAAO,GAAG,MAAM,KAAK,CAAC,SAAS,CAAC,KAAK,CAAC,CAAC;YAC7C,OAAO,CAAC,IAAI,CAAC,IAAI,CAAC;gBAChB,IAAI,EAAE,eAAe;gBACrB,OAAO,EAAE,IAAI,CAAC,OAAO;gBACrB,QAAQ,EAAE,KAAK,CAAC,OAAO;gBACvB,WAAW,EAAE,OAAO,CAAC,KAAK;gBAC1B,YAAY,EAAE,OAAO,CAAC,YAAY;gBAClC,UAAU,EAAE,OAAO,CAAC,UAAU;aAC/B,CAAC,CAAC;QACL,CAAC,CAAC,CAAC,KAAK,CAAC,CAAC,KAAK,EAAE,EAAE,GAAG,gBAAgB,KAAK,KAAK,CAAC,CAAC,CAAC,CAAC,CAAC;IACvD,CAAC,EAAE,oCAAoC,CAAC,CAAC;IACzC,cAAc,CAAC,KAAK,EAAE,CAAC;IACvB,IAAI,CAAC;QACH,MAAM,SAAS,GAAG,MAAM,qBAAqB,CAAC,OAAO,EAAE,SAAS,EAAE,IAAI,EAAE,KAAK,CAAC,OAAO,EAAE,EAAE,eAAe,EAAE,OAAO,CAAC,CAAC;QACnH,MAAM,aAAa,CAAC;QACpB,IAAI,gBAAgB,KAAK,SAAS;YAAE,MAAM,gBAAgB,CAAC;QAC3D,OAAO,EAAE,GAAG,SAAS,EAAE,OAAO,EAAE,KAAK,CAAC,OAAO,EAAE,UAAU,EAAE,IAAI,CAAC,GAAG,CAAC,CAAC,EAAE,IAAI,CAAC,GAAG,EAAE,GAAG,SAAS,CAAC,EAAE,CAAC;IACnG,CAAC;YAAS,CAAC;QACT,aAAa,CAAC,cAAc,CAAC,CAAC;QAC9B,MAAM,aAAa,CAAC;QACpB,IAAI,eAAe,CAAC,QAAQ;YAAE,MAAM,+BAA+B,CAAC,EAAE,IAAI,EAAE,OAAO,CAAC,IAAI,EAAE,OAAO,EAAE,KAAK,CAAC,OAAO,EAAE,KAAK,EAAE,CAAC,CAAC;QAC3H,MAAM,QAAQ,GAAG,MAAM,KAAK,CAAC,OAAO,CAAC,KAAK,CAAC,CAAC;QAC5C,OAAO,CAAC,IAAI,CAAC,IAAI,CAAC;YAChB,IAAI,EAAE,gBAAgB;YACtB,OAAO,EAAE,IAAI,CAAC,OAAO;YACrB,QAAQ,EAAE,KAAK,CAAC,OAAO;YACvB,WAAW,EAAE,QAAQ,CAAC,KAAK;YAC3B,SAAS,EAAE,OAAO,CAAC,MAAM,CAAC,QAAQ;SACnC,CAAC,CAAC;QACH,OAAO,CAAC,IAAI,CAAC,IAAI,CAAC;YAChB,IAAI,EAAE,+BAA+B;YACrC,OAAO,EAAE,IAAI,CAAC,OAAO;YACrB,QAAQ,EAAE,KAAK,CAAC,OAAO;YACvB,WAAW,EAAE,QAAQ,CAAC,KAAK;YAC3B,KAAK,EAAE,QAAQ,CAAC,KAAK;SACtB,CAAC,CAAC;QACH,IAAI,OAAO,CAAC,oBAAoB,EAAE,CAAC;YACjC,OAAO,CAAC,IAAI,CAAC,IAAI,CAAC,EAAE,IAAI,EAAE,yBAAyB,EAAE,OAAO,EAAE,IAAI,CAAC,OAAO,EAAE,QAAQ,EAAE,KAAK,CAAC,OAAO,EAAE,CAAC,CAAC;YACvG,IAAI,CAAC;gBACH,MAAM,MAAM,GAAG,MAAM,OAAO,CAAC,oBAAoB,CAAC;oBAChD,IAAI,EAAE,OAAO,CAAC,IAAI;oBAClB,QAAQ,EAAE,CAAC,KAAK,CAAC,OAAO,CAAC;oBACzB,GAAG,EAAE,OAAO,CAAC,GAAG;iBACjB,CAAC,CAAC;gBACH,OAAO,CAAC,IAAI,CAAC,IAAI,CAAC,EAAE,IAAI,EAAE,iCAAiC,EAAE,OAAO,EAAE,IAAI,CAAC,OAAO,EAAE,QAAQ,EAAE,KAAK,CAAC,OAAO,EAAE,OAAO,EAAE,MAAM,CAAC,OAAO,EAAE,OAAO,EAAE,MAAM,CAAC,OAAO,CAAC,MAAM,EAAE,MAAM,EAAE,MAAM,CAAC,MAAM,CAAC,MAAM,EAAE,CAAC,CAAC;gBACtM,OAAO,CAAC,IAAI,CAAC,IAAI,CAAC,EAAE,IAAI,EAAE,2BAA2B,EAAE,OAAO,EAAE,IAAI,CAAC,OAAO,EAAE,QAAQ,EAAE,KAAK,CAAC,OAAO,EAAE,OAAO,EAAE,MAAM,CAAC,OAAO,EAAE,OAAO,EAAE,MAAM,CAAC,OAAO,CAAC,MAAM,EAAE,kBAAkB,EAAE,MAAM,CAAC,MAAM,CAAC,MAAM,EAAE,CAAC,CAAC;YAC9M,CAAC;YAAC,OAAO,KAAK,EAAE,CAAC;gBACf,OAAO,CAAC,IAAI,CAAC,IAAI,CAAC,EAAE,IAAI,EAAE,8BAA8B,EAAE,OAAO,EAAE,IAAI,CAAC,OAAO,EAAE,QAAQ,EAAE,KAAK,CAAC,OAAO,EAAE,IAAI,EAAG,KAA2B,CAAC,IAAI,IAAI,sBAAsB,EAAE,CAAC,CAAC;gBAC/K,OAAO,CAAC,IAAI,CAAC,IAAI,CAAC,EAAE,IAAI,EAAE,wBAAwB,EAAE,OAAO,EAAE,IAAI,CAAC,OAAO,EAAE,QAAQ,EAAE,KAAK,CAAC,OAAO,EAAE,IAAI,EAAG,KAA2B,CAAC,IAAI,IAAI,sBAAsB,EAAE,CAAC,CAAC;YAC3K,CAAC;QACH,CAAC;QACD,MAAM,OAAO,CAAC,eAAe,EAAE,CAAC,IAAI,CAAC,OAAO,EAAE,KAAK,CAAC,OAAO,EAAE,UAAU,CAAC,CAAC;IAC3E,CAAC;AACH,CAAC;AAED,KAAK,UAAU,qBAAqB,CAClC,OAAoC,EACpC,SAA4B,EAC5B,IAAuB,EACvB,KAAuB,EACvB,eAAsC,EACtC,OAA6B;IAE7B,MAAM,cAAc,GAAG,IAAI,CAAC,eAAyB,CAAC;IACtD,MAAM,MAAM,GAAG,IAAI,CAAC,QAAQ,CAAC,CAAC,CAAW,CAAC;IAC1C,MAAM,iBAAiB,GAAG,MAAM,0BAA0B,CAAC,EAAE,MAAM,EAAE,IAAI,EAAE,IAAI,CAAC,UAAU,IAAI,EAAE,EAAE,GAAG,CAAC,OAAO,CAAC,8BAA8B,CAAC,CAAC,CAAC,EAAE,MAAM,EAAE,OAAO,CAAC,8BAA8B,EAAE,CAAC,CAAC,CAAC,EAAE,CAAC,EAAE,CAAC,CAAC;IAC3M,MAAM,gBAAgB,GAAG,uBAAuB,CAAC,OAAO,CAAC,IAAI,EAAE,MAAM,EAAE,IAAI,CAAC,WAAW,CAAC,CAAC;IACzF,MAAM,SAAS,GAAG,uBAAuB,CAAC,OAAO,CAAC,IAAI,EAAE,KAAK,EAAE,MAAM,CAAC,CAAC;IACvE,MAAM,gBAAgB,GAAG,2BAA2B,CAAC,EAAE,SAAS,EAAE,QAAQ,EAAE,OAAO,CAAC,MAAM,CAAC,QAAQ,EAAE,iBAAiB,EAAE,OAAO,CAAC,MAAM,CAAC,iBAAiB,EAAE,WAAW,EAAE,IAAI,CAAC,WAAW,EAAE,UAAU,EAAE,gBAAgB,CAAC,UAAU,EAAE,aAAa,EAAE,gBAAgB,CAAC,aAAa,EAAE,GAAG,EAAE,OAAO,CAAC,GAAG,EAAE,GAAG,CAAC,OAAO,CAAC,MAAM,CAAC,CAAC,CAAC,EAAE,MAAM,EAAE,OAAO,CAAC,MAAM,EAAE,CAAC,CAAC,CAAC,EAAE,CAAC,EAAE,CAAC,CAAC;IAC5V,MAAM,gBAAgB,GAAG,IAAI,CAAC,IAAI,CAAC,OAAO,CAAC,aAAa,EAAE,QAAQ,EAAE,YAAY,EAAE,IAAI,CAAC,OAAO,EAAE,SAAS,MAAM,CAAC,KAAK,CAAC,KAAK,CAAC,CAAC,QAAQ,CAAC,CAAC,EAAE,GAAG,CAAC,EAAE,CAAC,CAAC;IACjJ,MAAM,kBAAkB,GAAG,IAAI,CAAC,IAAI,CAAC,gBAAgB,EAAE,KAAK,CAAC,CAAC;IAC9D,MAAM,aAAa,GAAG,iBAAiB,CAAC,iBAAiB,CAAC,CAAC;IAC3D,MAAM,IAAI,GAAqB,EAAE,CAAC;IAClC,MAAM,OAAO,GAAG,KAAK,EAAE,GAAmB,EAAiB,EAAE;QAC3D,IAAI,GAAG,CAAC,OAAO,KAAK,cAAc,IAAI,GAAG,CAAC,OAAO,KAAK,MAAM,EAAE,CAAC;YAC7D,MAAM,IAAI,UAAU,CAAC,kDAAkD,GAAG,CAAC,OAAO,IAAI,GAAG,CAAC,OAAO,EAAE,EAAE;gBACnG,IAAI,EAAE,+BAA+B;gBACrC,QAAQ,EAAE,EAAE;aACb,CAAC,CAAC;QACL,CAAC;QACD,MAAM,QAAQ,GAAG,IAAI,CAAC,IAAI,CAAC,CAAC,OAAO,EAAE,EAAE,CAAC,OAAO,CAAC,QAAQ,KAAK,GAAG,CAAC,QAAQ,CAAC,CAAC;QAC3E,IAAI,QAAQ,EAAE,CAAC;YACb,IAAI,IAAI,CAAC,SAAS,CAAC,QAAQ,CAAC,KAAK,IAAI,CAAC,SAAS,CAAC,GAAG,CAAC;gBAAE,MAAM,IAAI,KAAK,CAAC,qCAAqC,GAAG,CAAC,QAAQ,EAAE,CAAC,CAAC;YAC3H,OAAO;QACT,CAAC;QACD,IAAI,CAAC,IAAI,CAAC,GAAG,CAAC,CAAC;QACf,IAAI,GAAG,CAAC,cAAc,KAAK,gCAAgC;YAAE,OAAO,CAAC,IAAI,CAAC,IAAI,CAAC;gBAC7E,IAAI,EAAE,uBAAuB,EAAE,OAAO,EAAE,IAAI,CAAC,OAAO,EAAE,QAAQ,EAAE,GAAG,CAAC,QAAQ,EAAE,OAAO,EAAE,GAAG,CAAC,OAAO;gBAClG,MAAM,EAAE,gCAAgC,EAAE,kBAAkB,EAAE,IAAI,EAAE,iBAAiB,EAAE,KAAK;aAC7F,CAAC,CAAC;QACH,MAAM,SAAS,CAAC,OAAO,CAAC,GAAG,EAAE,IAAI,CAAC,OAAO,CAAC,CAAC;IAC7C,CAAC,CAAC;IACF,MAAM,GAAG,GAAG,MAAM,gBAAgB,CAAC;QACjC,MAAM,EAAE,OAAO,CAAC,MAAM;QACtB,aAAa,EAAE,OAAO,CAAC,aAAa;QACpC,gBAAgB;QAChB,cAAc;QACd,SAAS,EAAE,CAAC,MAAM,CAAC;QACnB,OAAO,EAAE,EAAE,GAAG,OAAO,CAAC,OAAO,EAAE,QAAQ,EAAE,CAAC,EAAE,cAAc,EAAE,CAAC,EAAE;QAC/D,IAAI,EAAE,OAAO,CAAC,IAAI;QAClB,gBAAgB,EAAE,OAAO,CAAC,gBAAgB;QAC1C,gBAAgB,EAAE,OAAO,CAAC,iBAAiB,CAAC,SAAS;QACrD,SAAS,EAAE,OAAO,CAAC,iBAAiB,CAAC,UAAU;QAC/C,gBAAgB,EAAE,2BAA2B,CAAC,OAAO,EAAE,IAAI,CAAC;QAC5D,GAAG,CAAC,OAAO,CAAC,0BAA0B,CAAC,CAAC,CAAC,EAAE,UAAU,EAAE,OAAO,CAAC,0BAA0B,CAAC,KAAK,EAAE,CAAC,CAAC,CAAC,EAAE,CAAC;QACvG,kBAAkB,EAAE,gBAAgB,CAAC,UAAU;QAC/C,GAAG,CAAC,MAAM,CAAC,IAAI,CAAC,gBAAgB,CAAC,aAAa,CAAC,CAAC,MAAM,GAAG,CAAC,CAAC,CAAC,CAAC,EAAE,mBAAmB,EAAE,gBAAgB,CAAC,aAAa,EAAE,CAAC,CAAC,CAAC,EAAE,CAAC;QAC1H,eAAe,EAAE,SAAS;QAC1B,cAAc,EAAE,oBAAoB,CAAC,IAAI,CAAC,UAAU,IAAI,EAAE,CAAC;QAC3D,GAAG,CAAC,aAAa,CAAC,CAAC,CAAC,EAAE,iBAAiB,EAAE,aAAa,EAAE,CAAC,CAAC,CAAC,EAAE,CAAC;QAC9D,GAAG,EAAE,OAAO,CAAC,GAAG;QAChB,GAAG,CAAC,OAAO,CAAC,gBAAgB,KAAK,SAAS,CAAC,CAAC,CAAC,EAAE,gBAAgB,EAAE,OAAO,CAAC,gBAAgB,EAAE,CAAC,CAAC,CAAC,EAAE,CAAC;QACjG,GAAG,CAAC,OAAO,CAAC,MAAM,CAAC,CAAC,CAAC,EAAE,MAAM,EAAE,OAAO,CAAC,MAAM,EAAE,CAAC,CAAC,CAAC,EAAE,CAAC;QACrD,GAAG,CAAC,OAAO,CAAC,kBAAkB,KAAK,SAAS,CAAC,CAAC,CAAC,EAAE,CAAC,CAAC,CAAC,EAAE,kBAAkB,EAAE,OAAO,CAAC,kBAAkB,EAAE,CAAC;QACvG,GAAG,CAAC,OAAO,CAAC,MAAM,CAAC,QAAQ,KAAK,cAAc,IAAI,OAAO,CAAC,QAAQ,KAAK,OAAO,CAAC,CAAC,CAAC;YAC/E,kBAAkB,EAAE,IAAI;YACxB,gBAAgB,EAAE,CAAC,GAAW,EAAE,EAAE,CAAC,6BAA6B,CAAC;gBAC/D,IAAI,EAAE,OAAO,CAAC,IAAI;gBAClB,QAAQ,EAAE,OAAO,CAAC,MAAM,CAAC,QAAQ;gBACjC,aAAa,EAAE,OAAO,CAAC,aAAa;gBACpC,KAAK;gBACL,gBAAgB;gBAChB,GAAG;aACJ,CAAC,CAAC,IAAI,CAAC,GAAG,EAAE,GAAG,eAAe,CAAC,QAAQ,GAAG,IAAI,CAAC,CAAC,CAAC,CAAC;YACnD,eAAe,EAAE,CAAC,MAA8D,EAAE,EAAE,CAAC,4BAA4B,CAAC;gBAChH,IAAI,EAAE,OAAO,CAAC,IAAI;gBAClB,OAAO,EAAE,KAAK,CAAC,QAAQ;gBACvB,KAAK,EAAE,KAAK,CAAC,KAAK;gBAClB,GAAG,MAAM;aACV,CAAC,CAAC,IAAI,CAAC,GAAG,EAAE,CAAC,SAAS,CAAC;SACzB,CAAC,CAAC,CAAC,EAAE,CAAC;QACP,IAAI,EAAE,CAAC,KAAK,EAAE,EAAE,CAAC,OAAO,CAAC,IAAI,CAAC,IAAI,CAAC,EAAE,GAAG,KAAK,EAAE,OAAO,EAAE,IAAI,CAAC,OAAO,EAAE,OAAO,EAAE,MAAM,EAAE,CAAC;QACxF,cAAc,EAAE,KAAK,EAAE,KAAK,EAAE,OAAO,EAAoB,EAAE;YACzD,IAAI,CAAC;gBACH,MAAM,GAAG,GAAG,MAAM,kBAAkB,CAAC;oBACnC,IAAI,EAAE,OAAO,CAAC,IAAI;oBAClB,MAAM,EAAE,OAAO,CAAC,MAAM;oBACtB,aAAa,EAAE,OAAO,CAAC,aAAa;oBACpC,kBAAkB;oBAClB,eAAe,EAAE,cAAc;oBAC/B,OAAO,EAAE,OAAO,CAAC,OAAO;oBACxB,gBAAgB,EAAE,OAAO,CAAC,gBAAgB;oBAC1C,WAAW,EAAE,OAAO,CAAC,OAAO,CAAC,YAAY;oBACzC,iBAAiB,EAAE,OAAO,CAAC,OAAO,CAAC,kBAAkB;oBACrD,SAAS,EAAE,OAAO,CAAC,iBAAiB,CAAC,UAAU;oBAC/C,GAAG,EAAE,OAAO,CAAC,GAAG,EAAE,GAAG,CAAC,OAAO,CAAC,MAAM,CAAC,CAAC,CAAC,EAAE,MAAM,EAAE,OAAO,CAAC,MAAM,EAAE,CAAC,CAAC,CAAC,EAAE,CAAC;oBACvE,GAAG,CAAC,OAAO,CAAC,IAAI,CAAC,aAAa,CAAC,CAAC,CAAC,EAAE,gBAAgB,EAAE,OAAO,CAAC,IAAI,CAAC,aAAa,EAAE,CAAC,CAAC,CAAC,EAAE,CAAC;oBACvF,GAAG,CAAC,OAAO,CAAC,0BAA0B,CAAC,CAAC,CAAC;wBACvC,0BAA0B,EAAE,OAAO,CAAC,0BAA0B;qBAC/D,CAAC,CAAC,CAAC,EAAE,CAAC;oBACP,iBAAiB,EAAE,oCAAoC,CAAC,MAAM,gBAAgB,CAAC,OAAO,EAAE,EAAE,iBAAiB,CAAC;oBAC5G,GAAG,CAAC,iBAAiB,CAAC,CAAC,CAAC,EAAE,iBAAiB,EAAE,CAAC,CAAC,CAAC,EAAE,CAAC;oBACnD,qBAAqB,EAAE,IAAI;oBAC3B,4BAA4B,EAAE,OAAO,CAAC,iBAAiB;iBACxD,EAAE,KAAK,EAAE,IAAI,CAAC,MAAM,EAAE,IAAI,CAAC,CAAC;gBAC7B,MAAM,OAAO,CAAC,GAAG,CAAC,CAAC;gBACnB,OAAO,IAAI,CAAC;YACd,CAAC;YAAC,OAAO,KAAK,EAAE,CAAC;gBACf,IAAI,KAAK,YAAY,uBAAuB;oBAAE,OAAO,KAAK,CAAC;gBAC3D,MAAM,KAAK,CAAC;YACd,CAAC;QACH,CAAC;KACF,CAAC,CAAC,OAAO,CAAC,KAAK,IAAI,EAAE,GAAG,MAAM,gBAAgB,CAAC,IAAI,EAAE,CAAC,CAAC,CAAC,CAAC,CAAC;IAC3D,MAAM,mBAAmB,GAAG,IAAI,CAAC,GAAG,EAAE,CAAC;IACvC,KAAK,MAAM,KAAK,IAAI,MAAM,uBAAuB,CAAC,kBAAkB,EAAE,GAAG,CAAC,SAAS,CAAC,EAAE,CAAC;QACrF,OAAO,CAAC,IAAI,CAAC,IAAI,CAAC,KAAK,CAAC,CAAC;QACzB,IAAI,KAAK,CAAC,IAAI,KAAK,yBAAyB,IAAI,OAAO,KAAK,CAAC,WAAW,KAAK,QAAQ;YAAE,OAAO,CAAC,WAAW,CAAC,KAAK,CAAC,WAAW,CAAC,CAAC;IAChI,CAAC;IACD,MAAM,YAAY,GAAG,MAAM,mBAAmB,CAAC;QAC7C,IAAI,EAAE,OAAO,CAAC,IAAI;QAClB,MAAM,EAAE,OAAO,CAAC,MAAM;QACtB,aAAa,EAAE,OAAO,CAAC,aAAa;QACpC,kBAAkB;QAClB,eAAe,EAAE,cAAc;QAC/B,OAAO,EAAE,OAAO,CAAC,OAAO;QACxB,gBAAgB,EAAE,OAAO,CAAC,gBAAgB;QAC1C,WAAW,EAAE,OAAO,CAAC,OAAO,CAAC,YAAY;QACzC,iBAAiB,EAAE,OAAO,CAAC,OAAO,CAAC,kBAAkB;QACrD,SAAS,EAAE,OAAO,CAAC,iBAAiB,CAAC,UAAU;QAC/C,GAAG,EAAE,OAAO,CAAC,GAAG,EAAE,GAAG,CAAC,OAAO,CAAC,MAAM,CAAC,CAAC,CAAC,EAAE,MAAM,EAAE,OAAO,CAAC,MAAM,EAAE,CAAC,CAAC,CAAC,EAAE,CAAC;QACvE,GAAG,CAAC,OAAO,CAAC,IAAI,CAAC,aAAa,CAAC,CAAC,CAAC,EAAE,gBAAgB,EAAE,OAAO,CAAC,IAAI,CAAC,aAAa,EAAE,CAAC,CAAC,CAAC,EAAE,CAAC;QACvF,GAAG,CAAC,OAAO,CAAC,0BAA0B,CAAC,CAAC,CAAC;YACvC,0BAA0B,EAAE,OAAO,CAAC,0BAA0B;SAC/D,CAAC,CAAC,CAAC,EAAE,CAAC;QACP,iBAAiB,EAAE,oCAAoC,CAAC,MAAM,gBAAgB,CAAC,OAAO,EAAE,EAAE,iBAAiB,CAAC;QAC5G,GAAG,CAAC,iBAAiB,CAAC,CAAC,CAAC,EAAE,iBAAiB,EAAE,CAAC,CAAC,CAAC,EAAE,CAAC;QACnD,SAAS,EAAE,GAAG,CAAC,SAAS;KACzB,EAAE,IAAI,CAAC,CAAC;IACT,KAAK,MAAM,GAAG,IAAI,YAAY;QAAE,MAAM,OAAO,CAAC,GAAG,CAAC,CAAC;IACnD,OAAO,CAAC,IAAI,CAAC,IAAI,CAAC,EAAE,IAAI,EAAE,2BAA2B,EAAE,OAAO,EAAE,IAAI,CAAC,OAAO,EAAE,WAAW,EAAE,IAAI,CAAC,GAAG,EAAE,GAAG,mBAAmB,EAAE,CAAC,CAAC;IAC/H,IAAI,GAAG,CAAC,SAAS,KAAK,IAAI;QAAE,qBAAqB,CAAC,GAAG,CAAC,SAAS,EAAE,IAAI,CAAC,CAAC;IACvE,OAAO,EAAE,OAAO,EAAE,cAAc,EAAE,MAAM,EAAE,IAAI,CAAC,OAAO,EAAE,KAAK,EAAE,CAAC,MAAM,CAAC,EAAE,IAAI,EAAE,GAAG,EAAE,GAAG,CAAC,iBAAiB,CAAC,CAAC,CAAC,EAAE,iBAAiB,EAAE,CAAC,CAAC,CAAC,EAAE,CAAC,EAAE,CAAC;AAC5I,CAAC"}
@@ -0,0 +1,62 @@
1
+ import path from "node:path";
2
+ import { readJSON } from "../foundation/index.js";
3
+ import { mergeEvalProgressTrial, replaceInvalidEvalProgressTrial, writeEvalProgress } from "./progress.js";
4
+ import { validateEvalTrialReferences } from "./trial-reference-validation.js";
5
+ export class ProgressPublisher {
6
+ progress;
7
+ tail = Promise.resolve();
8
+ options;
9
+ constructor(progress, options) {
10
+ this.progress = progress;
11
+ this.options = options;
12
+ }
13
+ publish(ref, workId) {
14
+ return this.enqueue(ref, workId, "settle");
15
+ }
16
+ replaceInvalid(ref, workId) {
17
+ return this.enqueue(ref, workId, "replace-invalid");
18
+ }
19
+ enqueue(ref, workId, mode) {
20
+ const operation = this.tail.then(async () => {
21
+ const previous = this.progress.generation;
22
+ const next = mode === "settle"
23
+ ? mergeEvalProgressTrial(this.progress, ref)
24
+ : replaceInvalidEvalProgressTrial(this.progress, ref);
25
+ if (next.generation === previous)
26
+ return;
27
+ await validateEvalTrialReferences(this.options.root, this.options.evalId, [ref], {
28
+ benchmarkId: this.options.request.benchmark_id,
29
+ benchmarkRevision: this.options.request.benchmark_revision,
30
+ });
31
+ if (ref.run_group)
32
+ this.options.sink.emit({ type: "result.group.sealed", work_id: workId, run_group_id: ref.run_group.run_group_id, trial_id: ref.trial_id, group_digest: ref.run_group.digest });
33
+ else {
34
+ const bundle = await readJSON(path.join(this.options.root, "runs", ref.run_id, "bundle.index.json"));
35
+ this.options.sink.emit({ type: "result.bundle.sealed", work_id: workId, run_id: ref.run_id, trial_id: ref.trial_id, bundle_digest: bundle.bundle_digest });
36
+ }
37
+ this.progress = next;
38
+ await writeEvalProgress(this.options.evalDirectory, this.progress);
39
+ this.options.sink.emit({
40
+ type: mode === "settle" ? "eval.trial.published" : "eval.trial.replaced",
41
+ work_id: workId,
42
+ trial_id: ref.trial_id,
43
+ task_id: ref.task_id,
44
+ attempt: ref.attempt,
45
+ run_id: ref.run_id,
46
+ ...(ref.run_group ? { run_group_id: ref.run_group.run_group_id } : {}),
47
+ observation_status: ref.observation_status,
48
+ settled_trials: this.progress.trials.length,
49
+ generation: this.progress.generation,
50
+ });
51
+ });
52
+ this.tail = operation.catch(() => { });
53
+ return operation;
54
+ }
55
+ async settle() {
56
+ await this.tail;
57
+ }
58
+ current() {
59
+ return this.progress;
60
+ }
61
+ }
62
+ //# sourceMappingURL=planned-progress-publisher.js.map