agent-hitch 0.2.16 → 0.2.17

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 (361) hide show
  1. package/README.md +7 -28
  2. package/README.zh-CN.md +6 -26
  3. package/dist/scripts/benchmark-dataset-verification.js +118 -0
  4. package/dist/scripts/benchmark-dataset-verification.js.map +1 -0
  5. package/dist/scripts/canary-external-endpoint.js +56 -0
  6. package/dist/scripts/canary-external-endpoint.js.map +1 -0
  7. package/dist/scripts/canary-native-training.js +8 -6
  8. package/dist/scripts/canary-native-training.js.map +1 -1
  9. package/dist/scripts/canary-remote-training.js +8 -6
  10. package/dist/scripts/canary-remote-training.js.map +1 -1
  11. package/dist/scripts/check-architecture.js +1 -1
  12. package/dist/scripts/check-architecture.js.map +1 -1
  13. package/dist/src/adapters/providers/codex.js +5 -2
  14. package/dist/src/adapters/providers/codex.js.map +1 -1
  15. package/dist/src/adapters/providers/model-call.js +1 -2
  16. package/dist/src/adapters/providers/model-call.js.map +1 -1
  17. package/dist/src/adapters/providers/training-tool.js +2 -0
  18. package/dist/src/adapters/providers/training-tool.js.map +1 -1
  19. package/dist/src/backends/harbor/artifact-pins.js +42 -0
  20. package/dist/src/backends/harbor/artifact-pins.js.map +1 -0
  21. package/dist/src/backends/harbor/backend.js +16 -8
  22. package/dist/src/backends/harbor/backend.js.map +1 -1
  23. package/dist/src/backends/harbor/bound-credential-helper.js +24 -0
  24. package/dist/src/backends/harbor/bound-credential-helper.js.map +1 -0
  25. package/dist/src/backends/harbor/credential-snapshot.js +71 -0
  26. package/dist/src/backends/harbor/credential-snapshot.js.map +1 -0
  27. package/dist/src/backends/harbor/environment-config.js +25 -2
  28. package/dist/src/backends/harbor/environment-config.js.map +1 -1
  29. package/dist/src/backends/harbor/environment-files.js +143 -0
  30. package/dist/src/backends/harbor/environment-files.js.map +1 -0
  31. package/dist/src/backends/harbor/evaluation-binding.js +160 -0
  32. package/dist/src/backends/harbor/evaluation-binding.js.map +1 -0
  33. package/dist/src/backends/harbor/host-credential-helper.js +5 -0
  34. package/dist/src/backends/harbor/host-credential-helper.js.map +1 -1
  35. package/dist/src/backends/harbor/index.js +2 -0
  36. package/dist/src/backends/harbor/index.js.map +1 -1
  37. package/dist/src/backends/harbor/model-proxy-config.js +3 -18
  38. package/dist/src/backends/harbor/model-proxy-config.js.map +1 -1
  39. package/dist/src/backends/harbor/portable-regrade.js +15 -3
  40. package/dist/src/backends/harbor/portable-regrade.js.map +1 -1
  41. package/dist/src/backends/harbor/python-runtime.js +71 -0
  42. package/dist/src/backends/harbor/python-runtime.js.map +1 -0
  43. package/dist/src/backends/harbor/tools.js +6 -4
  44. package/dist/src/backends/harbor/tools.js.map +1 -1
  45. package/dist/src/backends/index.js +3 -0
  46. package/dist/src/backends/index.js.map +1 -1
  47. package/dist/src/cli/arguments.js +30 -30
  48. package/dist/src/cli/arguments.js.map +1 -1
  49. package/dist/src/cli/commands/capabilities.js +7 -3
  50. package/dist/src/cli/commands/capabilities.js.map +1 -1
  51. package/dist/src/cli/commands/daemon.js +0 -17
  52. package/dist/src/cli/commands/daemon.js.map +1 -1
  53. package/dist/src/cli/commands/eval-binding.js +32 -0
  54. package/dist/src/cli/commands/eval-binding.js.map +1 -0
  55. package/dist/src/cli/commands/eval.js +21 -15
  56. package/dist/src/cli/commands/eval.js.map +1 -1
  57. package/dist/src/cli/commands/models.js +32 -45
  58. package/dist/src/cli/commands/models.js.map +1 -1
  59. package/dist/src/cli/commands/run.js +0 -35
  60. package/dist/src/cli/commands/run.js.map +1 -1
  61. package/dist/src/cli/main.js +0 -4
  62. package/dist/src/cli/main.js.map +1 -1
  63. package/dist/src/cli/output.js +5 -16
  64. package/dist/src/cli/output.js.map +1 -1
  65. package/dist/src/control-plane/eval-records.js +3 -1
  66. package/dist/src/control-plane/eval-records.js.map +1 -1
  67. package/dist/src/control-plane/eval-recovery.js +4 -1
  68. package/dist/src/control-plane/eval-recovery.js.map +1 -1
  69. package/dist/src/control-plane/eval-scheduler.js +16 -8
  70. package/dist/src/control-plane/eval-scheduler.js.map +1 -1
  71. package/dist/src/control-plane/index.js +0 -2
  72. package/dist/src/control-plane/index.js.map +1 -1
  73. package/dist/src/control-plane/model-capture-planning.js +3 -4
  74. package/dist/src/control-plane/model-capture-planning.js.map +1 -1
  75. package/dist/src/control-plane/remote-model-routes.js +4 -24
  76. package/dist/src/control-plane/remote-model-routes.js.map +1 -1
  77. package/dist/src/control-plane/remote-offer-contract.js +2 -2
  78. package/dist/src/control-plane/remote-offer-contract.js.map +1 -1
  79. package/dist/src/control-plane/remote-result-transport.js +7 -8
  80. package/dist/src/control-plane/remote-result-transport.js.map +1 -1
  81. package/dist/src/control-plane/remote-work-coordinator.js +2 -2
  82. package/dist/src/control-plane/remote-work-coordinator.js.map +1 -1
  83. package/dist/src/control-plane/remote-work-recovery.js +2 -2
  84. package/dist/src/control-plane/remote-work-recovery.js.map +1 -1
  85. package/dist/src/control-plane/remote-worker-protocol.js +1 -1
  86. package/dist/src/control-plane/remote-worker-protocol.js.map +1 -1
  87. package/dist/src/control-plane/remote-worker-runner.js +21 -6
  88. package/dist/src/control-plane/remote-worker-runner.js.map +1 -1
  89. package/dist/src/control-plane/remote-workers.js +2 -2
  90. package/dist/src/control-plane/remote-workers.js.map +1 -1
  91. package/dist/src/control-plane/rerun-recovery.js +1 -1
  92. package/dist/src/control-plane/rerun-recovery.js.map +1 -1
  93. package/dist/src/control-plane/rerun-scheduler.js +21 -14
  94. package/dist/src/control-plane/rerun-scheduler.js.map +1 -1
  95. package/dist/src/control-plane/rerun-source.js +5 -2
  96. package/dist/src/control-plane/rerun-source.js.map +1 -1
  97. package/dist/src/control-plane/rerun-submission.js +7 -5
  98. package/dist/src/control-plane/rerun-submission.js.map +1 -1
  99. package/dist/src/control-plane/synthetic-result.js +3 -2
  100. package/dist/src/control-plane/synthetic-result.js.map +1 -1
  101. package/dist/src/control-plane/worker-model-relay.js +1 -1
  102. package/dist/src/control-plane/worker-model-relay.js.map +1 -1
  103. package/dist/src/controller-runtime/hash.js +6 -0
  104. package/dist/src/controller-runtime/hash.js.map +1 -1
  105. package/dist/src/daemon/client.js +12 -35
  106. package/dist/src/daemon/client.js.map +1 -1
  107. package/dist/src/daemon/evaluation-binding.js +33 -0
  108. package/dist/src/daemon/evaluation-binding.js.map +1 -0
  109. package/dist/src/daemon/evaluation-submission.js +27 -0
  110. package/dist/src/daemon/evaluation-submission.js.map +1 -0
  111. package/dist/src/daemon/index.js +1 -0
  112. package/dist/src/daemon/index.js.map +1 -1
  113. package/dist/src/daemon/json-stream.js +98 -0
  114. package/dist/src/daemon/json-stream.js.map +1 -0
  115. package/dist/src/daemon/scheduler.js +1 -4
  116. package/dist/src/daemon/scheduler.js.map +1 -1
  117. package/dist/src/daemon/server.js +14 -54
  118. package/dist/src/daemon/server.js.map +1 -1
  119. package/dist/src/domain/endpoints.js +2 -0
  120. package/dist/src/domain/endpoints.js.map +1 -0
  121. package/dist/src/domain/evaluation-binding.js +5 -0
  122. package/dist/src/domain/evaluation-binding.js.map +1 -0
  123. package/dist/src/domain/index.js +2 -0
  124. package/dist/src/domain/index.js.map +1 -1
  125. package/dist/src/domain/runs.js.map +1 -1
  126. package/dist/src/evals/benchmark-adapter-manifest.js +95 -32
  127. package/dist/src/evals/benchmark-adapter-manifest.js.map +1 -1
  128. package/dist/src/evals/benchmark-verification.js +106 -0
  129. package/dist/src/evals/benchmark-verification.js.map +1 -0
  130. package/dist/src/evals/collect-only-rerun.js +3 -2
  131. package/dist/src/evals/collect-only-rerun.js.map +1 -1
  132. package/dist/src/evals/directory.js +9 -1
  133. package/dist/src/evals/directory.js.map +1 -1
  134. package/dist/src/evals/endpoint-attempts.js +112 -0
  135. package/dist/src/evals/endpoint-attempts.js.map +1 -0
  136. package/dist/src/evals/endpoint-usage.js +8 -0
  137. package/dist/src/evals/endpoint-usage.js.map +1 -0
  138. package/dist/src/evals/environment-binding.js +9 -0
  139. package/dist/src/evals/environment-binding.js.map +1 -0
  140. package/dist/src/evals/eval-artifact-planning.js +10 -1
  141. package/dist/src/evals/eval-artifact-planning.js.map +1 -1
  142. package/dist/src/evals/eval-finalization.js +2 -3
  143. package/dist/src/evals/eval-finalization.js.map +1 -1
  144. package/dist/src/evals/eval-logical-plan.js +2 -0
  145. package/dist/src/evals/eval-logical-plan.js.map +1 -1
  146. package/dist/src/evals/eval-result-builder.js +4 -0
  147. package/dist/src/evals/eval-result-builder.js.map +1 -1
  148. package/dist/src/evals/eval-state.js +22 -0
  149. package/dist/src/evals/eval-state.js.map +1 -0
  150. package/dist/src/evals/execution-plan.js +6 -2
  151. package/dist/src/evals/execution-plan.js.map +1 -1
  152. package/dist/src/evals/index.js +5 -1
  153. package/dist/src/evals/index.js.map +1 -1
  154. package/dist/src/evals/infrastructure-retry.js +2 -0
  155. package/dist/src/evals/infrastructure-retry.js.map +1 -1
  156. package/dist/src/evals/model-capture-runtime.js +9 -13
  157. package/dist/src/evals/model-capture-runtime.js.map +1 -1
  158. package/dist/src/evals/native-phase-evidence.js +8 -2
  159. package/dist/src/evals/native-phase-evidence.js.map +1 -1
  160. package/dist/src/evals/planned-execution.js +7 -2
  161. package/dist/src/evals/planned-execution.js.map +1 -1
  162. package/dist/src/evals/planned-progress-publisher.js +18 -5
  163. package/dist/src/evals/planned-progress-publisher.js.map +1 -1
  164. package/dist/src/evals/planned-retry-execution.js +1 -0
  165. package/dist/src/evals/planned-retry-execution.js.map +1 -1
  166. package/dist/src/evals/preparation-rerun.js +0 -1
  167. package/dist/src/evals/preparation-rerun.js.map +1 -1
  168. package/dist/src/evals/progress.js +92 -7
  169. package/dist/src/evals/progress.js.map +1 -1
  170. package/dist/src/evals/records.js +12 -2
  171. package/dist/src/evals/records.js.map +1 -1
  172. package/dist/src/evals/recovery.js +2 -0
  173. package/dist/src/evals/recovery.js.map +1 -1
  174. package/dist/src/evals/remote-rerun-execution.js +1 -3
  175. package/dist/src/evals/remote-rerun-execution.js.map +1 -1
  176. package/dist/src/evals/remote-verifier-rerun.js +1 -1
  177. package/dist/src/evals/remote-verifier-rerun.js.map +1 -1
  178. package/dist/src/evals/request.js +14 -45
  179. package/dist/src/evals/request.js.map +1 -1
  180. package/dist/src/evals/rerun-observation.js +31 -0
  181. package/dist/src/evals/rerun-observation.js.map +1 -0
  182. package/dist/src/evals/rerun.js +23 -28
  183. package/dist/src/evals/rerun.js.map +1 -1
  184. package/dist/src/evals/service.js +45 -42
  185. package/dist/src/evals/service.js.map +1 -1
  186. package/dist/src/evals/shared-runtime.js +125 -0
  187. package/dist/src/evals/shared-runtime.js.map +1 -0
  188. package/dist/src/evals/trial-import.js +10 -2
  189. package/dist/src/evals/trial-import.js.map +1 -1
  190. package/dist/src/evals/trial-publication-recovery.js +1 -1
  191. package/dist/src/evals/trial-publication-recovery.js.map +1 -1
  192. package/dist/src/evals/verifier-artifacts.js +1 -0
  193. package/dist/src/evals/verifier-artifacts.js.map +1 -1
  194. package/dist/src/evals/verifier-only-rerun.js +1 -1
  195. package/dist/src/evals/verifier-only-rerun.js.map +1 -1
  196. package/dist/src/evals/verifier-rerun-result.js +3 -2
  197. package/dist/src/evals/verifier-rerun-result.js.map +1 -1
  198. package/dist/src/evals/verifier-score-artifacts.js +9 -1
  199. package/dist/src/evals/verifier-score-artifacts.js.map +1 -1
  200. package/dist/src/foundation/config.js +0 -5
  201. package/dist/src/foundation/config.js.map +1 -1
  202. package/dist/src/inference/index.js +3 -22
  203. package/dist/src/inference/index.js.map +1 -1
  204. package/dist/src/inference/lock.js +44 -130
  205. package/dist/src/inference/lock.js.map +1 -1
  206. package/dist/src/model-access/endpoint-contract.js +127 -0
  207. package/dist/src/model-access/endpoint-contract.js.map +1 -0
  208. package/dist/src/model-access/endpoint-credentials.js +54 -0
  209. package/dist/src/model-access/endpoint-credentials.js.map +1 -0
  210. package/dist/src/model-access/endpoint-probe-stream.js +66 -0
  211. package/dist/src/model-access/endpoint-probe-stream.js.map +1 -0
  212. package/dist/src/model-access/endpoint-verification.js +76 -0
  213. package/dist/src/model-access/endpoint-verification.js.map +1 -0
  214. package/dist/src/model-access/endpoints.js +105 -0
  215. package/dist/src/model-access/endpoints.js.map +1 -0
  216. package/dist/src/model-access/index.js +5 -1
  217. package/dist/src/model-access/index.js.map +1 -1
  218. package/dist/src/model-access/proxy.js +57 -14
  219. package/dist/src/model-access/proxy.js.map +1 -1
  220. package/dist/src/model-access/remote-model.js +21 -29
  221. package/dist/src/model-access/remote-model.js.map +1 -1
  222. package/dist/src/model-access/usage.js +30 -0
  223. package/dist/src/model-access/usage.js.map +1 -0
  224. package/dist/src/runs/endpoint-runtime.js +40 -0
  225. package/dist/src/runs/endpoint-runtime.js.map +1 -0
  226. package/dist/src/runs/executor.js +28 -42
  227. package/dist/src/runs/executor.js.map +1 -1
  228. package/dist/src/runs/local-inference-environment.js +5 -53
  229. package/dist/src/runs/local-inference-environment.js.map +1 -1
  230. package/dist/src/runs/manifest.js +1 -0
  231. package/dist/src/runs/manifest.js.map +1 -1
  232. package/dist/src/runs/outcome.js +3 -10
  233. package/dist/src/runs/outcome.js.map +1 -1
  234. package/dist/src/runs/queued.js +3 -2
  235. package/dist/src/runs/queued.js.map +1 -1
  236. package/dist/src/runs/records.js +3 -0
  237. package/dist/src/runs/records.js.map +1 -1
  238. package/dist/src/runs/request.js +10 -38
  239. package/dist/src/runs/request.js.map +1 -1
  240. package/dist/src/workers/remote-harbor-work-spec.js +8 -10
  241. package/dist/src/workers/remote-harbor-work-spec.js.map +1 -1
  242. package/dist/src/workers/remote-harbor-worker.js +3 -0
  243. package/dist/src/workers/remote-harbor-worker.js.map +1 -1
  244. package/docs/evidence/eval-preparation-2026-10-09.json +393 -0
  245. package/docs/evidence/eval-preparation-2026-10-09.md +178 -0
  246. package/docs/local-model-inference-spec.zh-CN.md +4 -663
  247. package/docs/resource-storage.zh-CN.md +116 -0
  248. package/docs/schemas/endpoint-evaluation-attempt.schema.json +29 -0
  249. package/docs/schemas/eval-artifact-pin.schema.json +14 -0
  250. package/docs/schemas/eval-execution-plan.schema.json +486 -105
  251. package/docs/schemas/eval-progress.schema.json +1 -1
  252. package/docs/schemas/eval-request.schema.json +149 -18
  253. package/docs/schemas/eval-rerun-submission.schema.json +8 -1
  254. package/docs/schemas/eval-result.schema.json +310 -56
  255. package/docs/schemas/evaluation-binding.schema.json +15 -0
  256. package/docs/schemas/external-endpoint-binding.schema.json +153 -0
  257. package/docs/schemas/external-endpoint-config.schema.json +93 -0
  258. package/docs/schemas/inference-execution-evidence.schema.json +253 -3
  259. package/docs/schemas/inference-lock-v3.schema.json +6 -0
  260. package/docs/schemas/inference-lock.schema.json +842 -341
  261. package/docs/schemas/ordered-eval-control.schema.json +7 -1
  262. package/docs/schemas/ready-endpoint-binding.schema.json +201 -0
  263. package/docs/schemas/run-manifest.schema.json +6 -0
  264. package/docs/schemas/run-request.schema.json +122 -20
  265. package/examples/external-inference/.env.example +16 -0
  266. package/examples/external-inference/README.md +203 -0
  267. package/examples/external-inference/accept.mjs +297 -0
  268. package/examples/external-inference/compose.gateway.yaml +119 -0
  269. package/examples/external-inference/compose.single.yaml +64 -0
  270. package/examples/external-inference/compose.tp2.yaml +15 -0
  271. package/examples/external-inference/deployment.example.json +58 -0
  272. package/examples/external-inference/endpoint.chat-completions.example.json +13 -0
  273. package/examples/external-inference/endpoint.responses.example.json +13 -0
  274. package/integrations/harbor/hitch_build_inputs.py +213 -0
  275. package/integrations/harbor/hitch_build_process.py +126 -0
  276. package/integrations/harbor/hitch_compose_preparation.py +274 -0
  277. package/integrations/harbor/hitch_harbor_agent.py +25 -39
  278. package/integrations/harbor/hitch_harbor_environment.py +96 -3
  279. package/integrations/harbor/hitch_image_cache.py +357 -0
  280. package/integrations/harbor/hitch_python_runtime.py +303 -0
  281. package/integrations/harbor/hitch_shared_runtime.py +123 -0
  282. package/integrations/model-call/cli.js +11 -10
  283. package/integrations/training-tool/cli.js +5 -3
  284. package/node_modules/smol-toml/LICENSE +24 -0
  285. package/node_modules/smol-toml/README.md +418 -0
  286. package/node_modules/smol-toml/dist/date.d.ts +41 -0
  287. package/node_modules/smol-toml/dist/date.js +127 -0
  288. package/node_modules/smol-toml/dist/error.d.ts +38 -0
  289. package/node_modules/smol-toml/dist/error.js +63 -0
  290. package/node_modules/smol-toml/dist/extract.js +69 -0
  291. package/node_modules/smol-toml/dist/index.cjs +734 -0
  292. package/node_modules/smol-toml/dist/index.d.ts +43 -0
  293. package/node_modules/smol-toml/dist/index.js +33 -0
  294. package/node_modules/smol-toml/dist/parse.d.ts +36 -0
  295. package/node_modules/smol-toml/dist/parse.js +149 -0
  296. package/node_modules/smol-toml/dist/primitive.js +238 -0
  297. package/node_modules/smol-toml/dist/stringify.d.ts +31 -0
  298. package/node_modules/smol-toml/dist/stringify.js +181 -0
  299. package/node_modules/smol-toml/dist/struct.js +179 -0
  300. package/node_modules/smol-toml/dist/util.d.ts +38 -0
  301. package/node_modules/smol-toml/dist/util.js +89 -0
  302. package/node_modules/smol-toml/package.json +68 -0
  303. package/package.json +15 -4
  304. package/dist/src/cli/commands/local.js +0 -130
  305. package/dist/src/cli/commands/local.js.map +0 -1
  306. package/dist/src/cli/commands/model-node.js +0 -26
  307. package/dist/src/cli/commands/model-node.js.map +0 -1
  308. package/dist/src/control-plane/inference-manager.js +0 -357
  309. package/dist/src/control-plane/inference-manager.js.map +0 -1
  310. package/dist/src/control-plane/managed-gateway-receipt.js +0 -26
  311. package/dist/src/control-plane/managed-gateway-receipt.js.map +0 -1
  312. package/dist/src/control-plane/managed-service-recovery.js +0 -158
  313. package/dist/src/control-plane/managed-service-recovery.js.map +0 -1
  314. package/dist/src/daemon/inference-prepare.js +0 -47
  315. package/dist/src/daemon/inference-prepare.js.map +0 -1
  316. package/dist/src/inference/daemon-capacity.js +0 -41
  317. package/dist/src/inference/daemon-capacity.js.map +0 -1
  318. package/dist/src/inference/device-reservation.js +0 -58
  319. package/dist/src/inference/device-reservation.js.map +0 -1
  320. package/dist/src/inference/doctor.js +0 -95
  321. package/dist/src/inference/doctor.js.map +0 -1
  322. package/dist/src/inference/generation-recovery.js +0 -83
  323. package/dist/src/inference/generation-recovery.js.map +0 -1
  324. package/dist/src/inference/managed-launcher.js +0 -35
  325. package/dist/src/inference/managed-launcher.js.map +0 -1
  326. package/dist/src/inference/materialize.js +0 -66
  327. package/dist/src/inference/materialize.js.map +0 -1
  328. package/dist/src/inference/model-location.js +0 -34
  329. package/dist/src/inference/model-location.js.map +0 -1
  330. package/dist/src/inference/node-client.js +0 -167
  331. package/dist/src/inference/node-client.js.map +0 -1
  332. package/dist/src/inference/node-registry.js +0 -79
  333. package/dist/src/inference/node-registry.js.map +0 -1
  334. package/dist/src/inference/observation.js +0 -79
  335. package/dist/src/inference/observation.js.map +0 -1
  336. package/dist/src/inference/preflight.js +0 -132
  337. package/dist/src/inference/preflight.js.map +0 -1
  338. package/dist/src/inference/process-sglang.js +0 -169
  339. package/dist/src/inference/process-sglang.js.map +0 -1
  340. package/dist/src/inference/protocol-probe.js +0 -29
  341. package/dist/src/inference/protocol-probe.js.map +0 -1
  342. package/dist/src/inference/remote-model-store.js +0 -24
  343. package/dist/src/inference/remote-model-store.js.map +0 -1
  344. package/dist/src/inference/runtime-catalog.js +0 -41
  345. package/dist/src/inference/runtime-catalog.js.map +0 -1
  346. package/dist/src/inference/runtime-store.js +0 -93
  347. package/dist/src/inference/runtime-store.js.map +0 -1
  348. package/dist/src/inference/service-attachment.js +0 -46
  349. package/dist/src/inference/service-attachment.js.map +0 -1
  350. package/dist/src/inference/service-inspection.js +0 -38
  351. package/dist/src/inference/service-inspection.js.map +0 -1
  352. package/dist/src/inference/sglang.js +0 -279
  353. package/dist/src/inference/sglang.js.map +0 -1
  354. package/dist/src/inference/supervisor.js +0 -460
  355. package/dist/src/inference/supervisor.js.map +0 -1
  356. package/dist/src/model-access/local-gateway-request.js +0 -139
  357. package/dist/src/model-access/local-gateway-request.js.map +0 -1
  358. package/dist/src/model-access/local-gateway.js +0 -323
  359. package/dist/src/model-access/local-gateway.js.map +0 -1
  360. package/dist/src/runs/local-inference-run.js +0 -53
  361. package/dist/src/runs/local-inference-run.js.map +0 -1
@@ -23,7 +23,7 @@
23
23
  "eval_id": { "type": "string", "pattern": "^eval_[a-f0-9]{32}$" },
24
24
  "benchmark_id": { "type": "string", "minLength": 1 },
25
25
  "benchmark_revision": { "type": "string", "minLength": 1 },
26
- "status": { "const": "running" },
26
+ "status": { "enum": ["running", "succeeded", "failed", "cancelled"] },
27
27
  "generation": { "type": "integer", "minimum": 0 },
28
28
  "planned_tasks": { "type": ["integer", "null"], "minimum": 0 },
29
29
  "planned_trials": { "type": ["integer", "null"], "minimum": 0 },
@@ -4,24 +4,155 @@
4
4
  "title": "Hitch eval request",
5
5
  "type": "object",
6
6
  "additionalProperties": false,
7
- "required": ["backend", "dataset", "harness_ref"],
7
+ "required": [
8
+ "backend",
9
+ "dataset",
10
+ "harness_ref"
11
+ ],
8
12
  "properties": {
9
- "schema_version": { "const": "1" },
10
- "backend": { "const": "harbor" },
11
- "dataset": { "type": "string", "minLength": 1 },
12
- "benchmark_id": { "type": "string", "minLength": 1 },
13
- "benchmark_revision": { "type": "string", "minLength": 1 },
14
- "harness_ref": { "type": "string", "minLength": 1 },
15
- "model": { "type": "string" },
16
- "attempts": { "type": "integer", "minimum": 1 },
17
- "max_concurrent": { "type": "integer", "minimum": 1 },
18
- "infrastructure_retries": { "type": "integer", "minimum": 0 },
19
- "infrastructure_retry_backoff_ms": { "type": "number", "minimum": 0 },
20
- "timeout_ms": { "type": "number", "minimum": 0 },
21
- "setup_timeout_ms": { "type": "number", "minimum": 0 },
22
- "agent_args": { "type": "array", "items": { "type": "string" } },
23
- "pass_env": { "type": "array", "items": { "type": "string" }, "uniqueItems": true },
24
- "local_inference": { "$ref": "local-inference-selection.schema.json" },
25
- "training_binding": { "$ref": "training-external-binding.schema.json" }
13
+ "schema_version": {
14
+ "const": "1"
15
+ },
16
+ "backend": {
17
+ "const": "harbor"
18
+ },
19
+ "dataset": {
20
+ "type": "string",
21
+ "minLength": 1
22
+ },
23
+ "benchmark_id": {
24
+ "type": "string",
25
+ "minLength": 1
26
+ },
27
+ "benchmark_revision": {
28
+ "type": "string",
29
+ "minLength": 1
30
+ },
31
+ "harness_ref": {
32
+ "type": "string",
33
+ "minLength": 1
34
+ },
35
+ "model": {
36
+ "type": "string"
37
+ },
38
+ "attempts": {
39
+ "type": "integer",
40
+ "minimum": 1
41
+ },
42
+ "max_concurrent": {
43
+ "type": "integer",
44
+ "minimum": 1
45
+ },
46
+ "infrastructure_retries": {
47
+ "type": "integer",
48
+ "minimum": 0
49
+ },
50
+ "infrastructure_retry_backoff_ms": {
51
+ "type": "number",
52
+ "minimum": 0
53
+ },
54
+ "timeout_ms": {
55
+ "type": "number",
56
+ "minimum": 0
57
+ },
58
+ "setup_timeout_ms": {
59
+ "type": "number",
60
+ "minimum": 0
61
+ },
62
+ "agent_args": {
63
+ "type": "array",
64
+ "items": {
65
+ "type": "string"
66
+ }
67
+ },
68
+ "pass_env": {
69
+ "type": "array",
70
+ "items": {
71
+ "type": "string"
72
+ },
73
+ "uniqueItems": true
74
+ },
75
+ "training_binding": {
76
+ "$ref": "training-external-binding.schema.json"
77
+ },
78
+ "environment_binding": {
79
+ "type": "string",
80
+ "pattern": "^hmac-sha256:[a-f0-9]{64}$"
81
+ },
82
+ "artifact_pins": {
83
+ "type": "array",
84
+ "minItems": 1,
85
+ "maxItems": 2,
86
+ "uniqueItems": true,
87
+ "items": {
88
+ "$ref": "eval-artifact-pin.schema.json"
89
+ },
90
+ "allOf": [
91
+ {
92
+ "contains": {
93
+ "type": "object",
94
+ "required": [
95
+ "platform"
96
+ ],
97
+ "properties": {
98
+ "platform": {
99
+ "const": "linux-x64"
100
+ }
101
+ }
102
+ },
103
+ "minContains": 0,
104
+ "maxContains": 1
105
+ },
106
+ {
107
+ "contains": {
108
+ "type": "object",
109
+ "required": [
110
+ "platform"
111
+ ],
112
+ "properties": {
113
+ "platform": {
114
+ "const": "linux-arm64"
115
+ }
116
+ }
117
+ },
118
+ "minContains": 0,
119
+ "maxContains": 1
120
+ }
121
+ ]
122
+ },
123
+ "endpoint_binding": {
124
+ "$ref": "external-endpoint-binding.schema.json"
125
+ },
126
+ "model_binding": {
127
+ "$ref": "ready-endpoint-binding.schema.json"
128
+ },
129
+ "generation": {
130
+ "type": "object",
131
+ "additionalProperties": false,
132
+ "properties": {
133
+ "max_output_tokens": {
134
+ "type": "integer",
135
+ "minimum": 1
136
+ },
137
+ "temperature": {
138
+ "type": "number"
139
+ },
140
+ "top_p": {
141
+ "type": "number"
142
+ },
143
+ "top_k": {
144
+ "type": "number"
145
+ },
146
+ "min_p": {
147
+ "type": "number"
148
+ },
149
+ "repetition_penalty": {
150
+ "type": "number"
151
+ },
152
+ "seed": {
153
+ "type": "number"
154
+ }
155
+ }
156
+ }
26
157
  }
27
158
  }
@@ -10,12 +10,19 @@
10
10
  "rerun_id": { "type": "string", "pattern": "^rerun_[a-f0-9]{32}$" },
11
11
  "eval_id": { "type": "string", "pattern": "^eval_[a-f0-9]{32}$" },
12
12
  "rerun_type": { "$ref": "#/$defs/rerunType" },
13
+ "endpoint_attempt": { "$ref": "endpoint-evaluation-attempt.schema.json" },
13
14
  "verifier_runtime_id": { "type": "string", "pattern": "^sha256:[a-f0-9]{64}$" },
14
15
  "semantics": { "$ref": "#/$defs/semantics" },
15
16
  "selector": { "$ref": "#/$defs/selector" },
16
17
  "submitted_at": { "type": "string", "format": "date-time" }
17
18
  },
18
- "allOf": [{ "if": { "required": ["verifier_runtime_id"] }, "then": { "properties": { "rerun_type": { "const": "verifier-only" } } } }],
19
+ "allOf": [
20
+ { "if": { "required": ["verifier_runtime_id"] }, "then": { "properties": { "rerun_type": { "const": "verifier-only" } } } },
21
+ { "if": { "required": ["endpoint_attempt"] }, "then": { "properties": {
22
+ "rerun_type": { "const": "candidate-restart" },
23
+ "selector": { "type": "object", "additionalProperties": false, "required": ["mode"], "properties": { "mode": { "const": "invalid" } } }
24
+ } } }
25
+ ],
19
26
  "$defs": {
20
27
  "rerunType": {
21
28
  "enum": ["candidate-restart", "candidate-resume", "trajectory-replay", "verifier-only", "collect-only"]
@@ -4,42 +4,148 @@
4
4
  "title": "Hitch eval result",
5
5
  "type": "object",
6
6
  "additionalProperties": true,
7
- "required": ["schema_version", "eval_id", "benchmark_id", "benchmark_revision", "status", "exit_code", "trials", "started_at", "completed_at"],
7
+ "required": [
8
+ "schema_version",
9
+ "eval_id",
10
+ "benchmark_id",
11
+ "benchmark_revision",
12
+ "status",
13
+ "exit_code",
14
+ "trials",
15
+ "started_at",
16
+ "completed_at"
17
+ ],
8
18
  "properties": {
9
- "schema_version": { "const": "1" },
10
- "eval_id": { "type": "string", "pattern": "^eval_[a-f0-9]{32}$" },
11
- "status": { "enum": ["succeeded", "failed", "cancelled"] },
12
- "exit_code": { "type": "integer", "minimum": 0 },
19
+ "schema_version": {
20
+ "const": "1"
21
+ },
22
+ "eval_id": {
23
+ "type": "string",
24
+ "pattern": "^eval_[a-f0-9]{32}$"
25
+ },
26
+ "status": {
27
+ "enum": [
28
+ "succeeded",
29
+ "failed",
30
+ "cancelled"
31
+ ]
32
+ },
33
+ "exit_code": {
34
+ "type": "integer",
35
+ "minimum": 0
36
+ },
13
37
  "backend": {
14
38
  "type": "object",
15
- "required": ["name", "executable", "identity", "config_path", "process_exit_code"],
39
+ "required": [
40
+ "name",
41
+ "executable",
42
+ "identity",
43
+ "config_path",
44
+ "process_exit_code"
45
+ ],
16
46
  "properties": {
17
- "name": { "const": "harbor" },
18
- "executable": { "type": "string", "minLength": 1 },
19
- "version": { "type": ["string", "null"] },
20
- "identity": { "type": "string", "pattern": "^sha256:[a-f0-9]{64}$" },
21
- "config_path": { "type": "string", "minLength": 1 },
22
- "result_path": { "type": ["string", "null"] },
23
- "process_exit_code": { "type": ["integer", "null"] }
47
+ "name": {
48
+ "const": "harbor"
49
+ },
50
+ "executable": {
51
+ "type": "string",
52
+ "minLength": 1
53
+ },
54
+ "version": {
55
+ "type": [
56
+ "string",
57
+ "null"
58
+ ]
59
+ },
60
+ "identity": {
61
+ "type": "string",
62
+ "pattern": "^sha256:[a-f0-9]{64}$"
63
+ },
64
+ "config_path": {
65
+ "type": "string",
66
+ "minLength": 1
67
+ },
68
+ "result_path": {
69
+ "type": [
70
+ "string",
71
+ "null"
72
+ ]
73
+ },
74
+ "process_exit_code": {
75
+ "type": [
76
+ "integer",
77
+ "null"
78
+ ]
79
+ }
24
80
  },
25
81
  "additionalProperties": true
26
82
  },
27
- "summary": { "type": ["object", "null"] },
28
- "benchmark_id": { "type": "string", "minLength": 1 },
29
- "benchmark_revision": { "type": "string", "minLength": 1 },
83
+ "summary": {
84
+ "type": [
85
+ "object",
86
+ "null"
87
+ ]
88
+ },
89
+ "benchmark_id": {
90
+ "type": "string",
91
+ "minLength": 1
92
+ },
93
+ "benchmark_revision": {
94
+ "type": "string",
95
+ "minLength": 1
96
+ },
30
97
  "prepared_artifact": {
31
98
  "type": "object",
32
- "required": ["artifact_id", "artifact_integrity", "entrypoint_integrity", "harness_id", "revision_identity", "platform", "source_type"],
99
+ "required": [
100
+ "artifact_id",
101
+ "artifact_integrity",
102
+ "entrypoint_integrity",
103
+ "harness_id",
104
+ "revision_identity",
105
+ "platform",
106
+ "source_type"
107
+ ],
33
108
  "properties": {
34
- "artifact_id": { "type": "string", "pattern": "^sha256:[0-9a-f]{64}$" },
35
- "artifact_integrity": { "type": "string", "pattern": "^sha256:[0-9a-f]{64}$" },
36
- "entrypoint_integrity": { "type": "string", "pattern": "^sha256:[0-9a-f]{64}$" },
37
- "harness_id": { "type": "string", "pattern": "^[a-z][a-z0-9-]*$" },
38
- "revision_identity": { "type": "string", "pattern": "^sha256:[0-9a-f]{64}$" },
39
- "platform": { "type": "string", "pattern": "^(?:linux|darwin|win32)-[a-z0-9_]+$" },
40
- "node_version": { "type": "string", "pattern": "^v[0-9]+\\.[0-9]+\\.[0-9]+$" },
41
- "source_type": { "enum": ["npm", "git"] },
42
- "storage": { "enum": ["host-artifact-store-v1", "harbor-artifact-cache-v2"] }
109
+ "artifact_id": {
110
+ "type": "string",
111
+ "pattern": "^sha256:[0-9a-f]{64}$"
112
+ },
113
+ "artifact_integrity": {
114
+ "type": "string",
115
+ "pattern": "^sha256:[0-9a-f]{64}$"
116
+ },
117
+ "entrypoint_integrity": {
118
+ "type": "string",
119
+ "pattern": "^sha256:[0-9a-f]{64}$"
120
+ },
121
+ "harness_id": {
122
+ "type": "string",
123
+ "pattern": "^[a-z][a-z0-9-]*$"
124
+ },
125
+ "revision_identity": {
126
+ "type": "string",
127
+ "pattern": "^sha256:[0-9a-f]{64}$"
128
+ },
129
+ "platform": {
130
+ "type": "string",
131
+ "pattern": "^(?:linux|darwin|win32)-[a-z0-9_]+$"
132
+ },
133
+ "node_version": {
134
+ "type": "string",
135
+ "pattern": "^v[0-9]+\\.[0-9]+\\.[0-9]+$"
136
+ },
137
+ "source_type": {
138
+ "enum": [
139
+ "npm",
140
+ "git"
141
+ ]
142
+ },
143
+ "storage": {
144
+ "enum": [
145
+ "host-artifact-store-v1",
146
+ "harbor-artifact-cache-v2"
147
+ ]
148
+ }
43
149
  },
44
150
  "additionalProperties": false
45
151
  },
@@ -48,25 +154,91 @@
48
154
  "minItems": 1,
49
155
  "items": {
50
156
  "type": "object",
51
- "required": ["artifact_id", "artifact_integrity", "entrypoint_integrity", "harness_id", "revision_identity", "platform", "node_version", "source_type", "task_ids", "runtime_contract"],
157
+ "required": [
158
+ "artifact_id",
159
+ "artifact_integrity",
160
+ "entrypoint_integrity",
161
+ "harness_id",
162
+ "revision_identity",
163
+ "platform",
164
+ "node_version",
165
+ "source_type",
166
+ "task_ids",
167
+ "runtime_contract"
168
+ ],
52
169
  "properties": {
53
- "artifact_id": { "type": "string", "pattern": "^sha256:[0-9a-f]{64}$" },
54
- "artifact_integrity": { "type": "string", "pattern": "^sha256:[0-9a-f]{64}$" },
55
- "entrypoint_integrity": { "type": "string", "pattern": "^sha256:[0-9a-f]{64}$" },
56
- "harness_id": { "type": "string", "pattern": "^[a-z][a-z0-9-]*$" },
57
- "revision_identity": { "type": "string", "pattern": "^sha256:[0-9a-f]{64}$" },
58
- "platform": { "type": "string", "pattern": "^(?:linux|darwin|win32)-[a-z0-9_]+$" },
59
- "node_version": { "type": "string", "pattern": "^v[0-9]+\\.[0-9]+\\.[0-9]+$" },
60
- "source_type": { "enum": ["npm", "git"] },
61
- "storage": { "enum": ["host-artifact-store-v1", "harbor-artifact-cache-v2"] },
62
- "task_ids": { "type": "array", "items": { "type": "string", "minLength": 1 }, "uniqueItems": true },
170
+ "artifact_id": {
171
+ "type": "string",
172
+ "pattern": "^sha256:[0-9a-f]{64}$"
173
+ },
174
+ "artifact_integrity": {
175
+ "type": "string",
176
+ "pattern": "^sha256:[0-9a-f]{64}$"
177
+ },
178
+ "entrypoint_integrity": {
179
+ "type": "string",
180
+ "pattern": "^sha256:[0-9a-f]{64}$"
181
+ },
182
+ "harness_id": {
183
+ "type": "string",
184
+ "pattern": "^[a-z][a-z0-9-]*$"
185
+ },
186
+ "revision_identity": {
187
+ "type": "string",
188
+ "pattern": "^sha256:[0-9a-f]{64}$"
189
+ },
190
+ "platform": {
191
+ "type": "string",
192
+ "pattern": "^(?:linux|darwin|win32)-[a-z0-9_]+$"
193
+ },
194
+ "node_version": {
195
+ "type": "string",
196
+ "pattern": "^v[0-9]+\\.[0-9]+\\.[0-9]+$"
197
+ },
198
+ "source_type": {
199
+ "enum": [
200
+ "npm",
201
+ "git"
202
+ ]
203
+ },
204
+ "storage": {
205
+ "enum": [
206
+ "host-artifact-store-v1",
207
+ "harbor-artifact-cache-v2"
208
+ ]
209
+ },
210
+ "task_ids": {
211
+ "type": "array",
212
+ "items": {
213
+ "type": "string",
214
+ "minLength": 1
215
+ },
216
+ "uniqueItems": true
217
+ },
63
218
  "runtime_contract": {
64
219
  "type": "object",
65
- "required": ["docker_platform", "artifact_platform", "node_version"],
220
+ "required": [
221
+ "docker_platform",
222
+ "artifact_platform",
223
+ "node_version"
224
+ ],
66
225
  "properties": {
67
- "docker_platform": { "enum": ["linux/amd64", "linux/arm64"] },
68
- "artifact_platform": { "enum": ["linux-x64", "linux-arm64"] },
69
- "node_version": { "type": "string", "pattern": "^v[0-9]+\\.[0-9]+\\.[0-9]+$" }
226
+ "docker_platform": {
227
+ "enum": [
228
+ "linux/amd64",
229
+ "linux/arm64"
230
+ ]
231
+ },
232
+ "artifact_platform": {
233
+ "enum": [
234
+ "linux-x64",
235
+ "linux-arm64"
236
+ ]
237
+ },
238
+ "node_version": {
239
+ "type": "string",
240
+ "pattern": "^v[0-9]+\\.[0-9]+\\.[0-9]+$"
241
+ }
70
242
  },
71
243
  "additionalProperties": false
72
244
  }
@@ -76,33 +248,115 @@
76
248
  },
77
249
  "trials": {
78
250
  "type": "array",
79
- "items": { "$ref": "eval-trial-reference.schema.json" }
251
+ "items": {
252
+ "$ref": "eval-trial-reference.schema.json"
253
+ }
80
254
  },
81
255
  "local_source_transport": {
82
256
  "type": "object",
83
- "required": ["kind", "resolution_identity", "commit", "tree", "payload_sha256", "payload_bytes"],
257
+ "required": [
258
+ "kind",
259
+ "resolution_identity",
260
+ "commit",
261
+ "tree",
262
+ "payload_sha256",
263
+ "payload_bytes"
264
+ ],
84
265
  "properties": {
85
- "kind": { "const": "local-git-commit" },
86
- "resolution_identity": { "type": "string", "pattern": "^sha256:[0-9a-f]{64}$" },
87
- "commit": { "type": "string", "pattern": "^(?:[0-9a-f]{40}|[0-9a-f]{64})$" },
88
- "tree": { "type": "string", "pattern": "^(?:[0-9a-f]{40}|[0-9a-f]{64})$" },
89
- "payload_sha256": { "type": "string", "pattern": "^sha256:[0-9a-f]{64}$" },
90
- "payload_bytes": { "type": "integer", "minimum": 0 },
91
- "object_count": { "type": "integer", "minimum": 0 },
92
- "file_count": { "type": "integer", "minimum": 0 }
266
+ "kind": {
267
+ "const": "local-git-commit"
268
+ },
269
+ "resolution_identity": {
270
+ "type": "string",
271
+ "pattern": "^sha256:[0-9a-f]{64}$"
272
+ },
273
+ "commit": {
274
+ "type": "string",
275
+ "pattern": "^(?:[0-9a-f]{40}|[0-9a-f]{64})$"
276
+ },
277
+ "tree": {
278
+ "type": "string",
279
+ "pattern": "^(?:[0-9a-f]{40}|[0-9a-f]{64})$"
280
+ },
281
+ "payload_sha256": {
282
+ "type": "string",
283
+ "pattern": "^sha256:[0-9a-f]{64}$"
284
+ },
285
+ "payload_bytes": {
286
+ "type": "integer",
287
+ "minimum": 0
288
+ },
289
+ "object_count": {
290
+ "type": "integer",
291
+ "minimum": 0
292
+ },
293
+ "file_count": {
294
+ "type": "integer",
295
+ "minimum": 0
296
+ }
93
297
  },
94
298
  "additionalProperties": false
95
299
  },
96
300
  "error": {
97
301
  "type": "object",
98
- "required": ["code", "message"],
302
+ "required": [
303
+ "code",
304
+ "message"
305
+ ],
99
306
  "properties": {
100
- "code": { "type": "string", "minLength": 1 },
101
- "message": { "type": "string" }
307
+ "code": {
308
+ "type": "string",
309
+ "minLength": 1
310
+ },
311
+ "message": {
312
+ "type": "string"
313
+ }
102
314
  },
103
315
  "additionalProperties": true
104
316
  },
105
- "started_at": { "type": "string", "format": "date-time" },
106
- "completed_at": { "type": "string", "format": "date-time" }
317
+ "started_at": {
318
+ "type": "string",
319
+ "format": "date-time"
320
+ },
321
+ "completed_at": {
322
+ "type": "string",
323
+ "format": "date-time"
324
+ },
325
+ "endpoint_binding": {
326
+ "$ref": "external-endpoint-binding.schema.json"
327
+ },
328
+ "usage": {
329
+ "type": "object",
330
+ "properties": {
331
+ "requests": {
332
+ "type": "integer",
333
+ "minimum": 0
334
+ },
335
+ "input_tokens": {
336
+ "type": "number",
337
+ "minimum": 0
338
+ },
339
+ "output_tokens": {
340
+ "type": "number",
341
+ "minimum": 0
342
+ },
343
+ "total_tokens": {
344
+ "type": "number",
345
+ "minimum": 0
346
+ },
347
+ "source": {
348
+ "const": "model-interactions-v1"
349
+ },
350
+ "completeness": {
351
+ "enum": [
352
+ "complete",
353
+ "partial"
354
+ ]
355
+ }
356
+ }
357
+ },
358
+ "task_completed": {
359
+ "type": "boolean"
360
+ }
107
361
  }
108
362
  }
@@ -0,0 +1,15 @@
1
+ {
2
+ "$schema": "https://json-schema.org/draft/2020-12/schema",
3
+ "$id": "https://agent-hitch.local/schemas/evaluation-binding.schema.json",
4
+ "title": "Hitch resident evaluation environment binding",
5
+ "type": "object",
6
+ "additionalProperties": false,
7
+ "required": ["schema_version", "binding_id", "model", "pass_env", "expires_at_ms"],
8
+ "properties": {
9
+ "schema_version": { "const": "1" },
10
+ "binding_id": { "type": "string", "pattern": "^hmac-sha256:[a-f0-9]{64}$" },
11
+ "model": { "type": "string", "minLength": 1, "maxLength": 1024, "pattern": "^[^\\u0000\\r\\n]+$" },
12
+ "pass_env": { "type": "array", "maxItems": 128, "uniqueItems": true, "items": { "type": "string", "pattern": "^[A-Za-z_][A-Za-z0-9_]*$" } },
13
+ "expires_at_ms": { "type": ["integer", "null"], "minimum": 0 }
14
+ }
15
+ }