@namzu/sdk 2.0.0 → 3.0.0

This diff represents the content of publicly available package versions that have been released to one of the supported registries. The information contained in this diff is provided for informational purposes only and reflects changes between package versions as they appear in their respective public registries.
Files changed (1503) hide show
  1. package/CHANGELOG.md +2265 -0
  2. package/README.md +49 -53
  3. package/dist/__tests__/default-path-conformance.test.d.ts +2 -0
  4. package/dist/__tests__/default-path-conformance.test.d.ts.map +1 -0
  5. package/dist/__tests__/default-path-conformance.test.js +262 -0
  6. package/dist/__tests__/default-path-conformance.test.js.map +1 -0
  7. package/dist/__tests__/shipped-defaults.test.d.ts +2 -0
  8. package/dist/__tests__/shipped-defaults.test.d.ts.map +1 -0
  9. package/dist/__tests__/shipped-defaults.test.js +97 -0
  10. package/dist/__tests__/shipped-defaults.test.js.map +1 -0
  11. package/dist/advisory/__tests__/budget-enforcement.test.d.ts +2 -0
  12. package/dist/advisory/__tests__/budget-enforcement.test.d.ts.map +1 -0
  13. package/dist/advisory/__tests__/budget-enforcement.test.js +159 -0
  14. package/dist/advisory/__tests__/budget-enforcement.test.js.map +1 -0
  15. package/dist/advisory/__tests__/structured-response.test.d.ts +2 -0
  16. package/dist/advisory/__tests__/structured-response.test.d.ts.map +1 -0
  17. package/dist/advisory/__tests__/structured-response.test.js +99 -0
  18. package/dist/advisory/__tests__/structured-response.test.js.map +1 -0
  19. package/dist/advisory/budget.d.ts +15 -0
  20. package/dist/advisory/budget.d.ts.map +1 -0
  21. package/dist/advisory/budget.js +21 -0
  22. package/dist/advisory/budget.js.map +1 -0
  23. package/dist/advisory/context.d.ts +2 -0
  24. package/dist/advisory/context.d.ts.map +1 -1
  25. package/dist/advisory/context.js +16 -0
  26. package/dist/advisory/context.js.map +1 -1
  27. package/dist/advisory/executor.d.ts +14 -11
  28. package/dist/advisory/executor.d.ts.map +1 -1
  29. package/dist/advisory/executor.js +36 -24
  30. package/dist/advisory/executor.js.map +1 -1
  31. package/dist/advisory/executor.test.d.ts +8 -6
  32. package/dist/advisory/executor.test.d.ts.map +1 -1
  33. package/dist/advisory/executor.test.js +12 -8
  34. package/dist/advisory/executor.test.js.map +1 -1
  35. package/dist/advisory/index.d.ts +2 -0
  36. package/dist/advisory/index.d.ts.map +1 -1
  37. package/dist/advisory/index.js +2 -0
  38. package/dist/advisory/index.js.map +1 -1
  39. package/dist/advisory/parse.d.ts +23 -0
  40. package/dist/advisory/parse.d.ts.map +1 -0
  41. package/dist/advisory/parse.js +56 -0
  42. package/dist/advisory/parse.js.map +1 -0
  43. package/dist/agents/AbstractAgent.d.ts +48 -0
  44. package/dist/agents/AbstractAgent.d.ts.map +1 -1
  45. package/dist/agents/AbstractAgent.js +77 -0
  46. package/dist/agents/AbstractAgent.js.map +1 -1
  47. package/dist/agents/PipelineAgent.d.ts +10 -0
  48. package/dist/agents/PipelineAgent.d.ts.map +1 -1
  49. package/dist/agents/PipelineAgent.js +12 -0
  50. package/dist/agents/PipelineAgent.js.map +1 -1
  51. package/dist/agents/ReactiveAgent.d.ts +10 -0
  52. package/dist/agents/ReactiveAgent.d.ts.map +1 -1
  53. package/dist/agents/ReactiveAgent.js +40 -0
  54. package/dist/agents/ReactiveAgent.js.map +1 -1
  55. package/dist/agents/RouterAgent.d.ts +10 -0
  56. package/dist/agents/RouterAgent.d.ts.map +1 -1
  57. package/dist/agents/RouterAgent.js +23 -2
  58. package/dist/agents/RouterAgent.js.map +1 -1
  59. package/dist/agents/SupervisorAgent.d.ts +11 -1
  60. package/dist/agents/SupervisorAgent.d.ts.map +1 -1
  61. package/dist/agents/SupervisorAgent.js +23 -1
  62. package/dist/agents/SupervisorAgent.js.map +1 -1
  63. package/dist/agents/__tests__/idempotency-key.test.d.ts +2 -0
  64. package/dist/agents/__tests__/idempotency-key.test.d.ts.map +1 -0
  65. package/dist/agents/__tests__/idempotency-key.test.js +149 -0
  66. package/dist/agents/__tests__/idempotency-key.test.js.map +1 -0
  67. package/dist/agents/__tests__/invocation-lock.test.d.ts +2 -0
  68. package/dist/agents/__tests__/invocation-lock.test.d.ts.map +1 -0
  69. package/dist/agents/__tests__/invocation-lock.test.js +126 -0
  70. package/dist/agents/__tests__/invocation-lock.test.js.map +1 -0
  71. package/dist/agents/__tests__/reachability.test.d.ts +2 -0
  72. package/dist/agents/__tests__/reachability.test.d.ts.map +1 -0
  73. package/dist/agents/__tests__/reachability.test.js +173 -0
  74. package/dist/agents/__tests__/reachability.test.js.map +1 -0
  75. package/dist/agents/__tests__/supervisor-ledger.test.js +1 -1
  76. package/dist/agents/__tests__/supervisor-ledger.test.js.map +1 -1
  77. package/dist/agents/index.d.ts +1 -1
  78. package/dist/agents/index.d.ts.map +1 -1
  79. package/dist/agents/lock.d.ts +0 -1
  80. package/dist/agents/lock.d.ts.map +1 -1
  81. package/dist/agents/lock.js.map +1 -1
  82. package/dist/bridge/a2a/late-events.test.d.ts +2 -0
  83. package/dist/bridge/a2a/late-events.test.d.ts.map +1 -0
  84. package/dist/bridge/a2a/late-events.test.js +70 -0
  85. package/dist/bridge/a2a/late-events.test.js.map +1 -0
  86. package/dist/bridge/a2a/mapper.d.ts.map +1 -1
  87. package/dist/bridge/a2a/mapper.js +55 -3
  88. package/dist/bridge/a2a/mapper.js.map +1 -1
  89. package/dist/bridge/a2a/mapper.test.js +1 -1
  90. package/dist/bridge/a2a/mapper.test.js.map +1 -1
  91. package/dist/bridge/a2a/message.d.ts.map +1 -1
  92. package/dist/bridge/a2a/message.js +5 -1
  93. package/dist/bridge/a2a/message.js.map +1 -1
  94. package/dist/bridge/sse/late-events.test.d.ts +2 -0
  95. package/dist/bridge/sse/late-events.test.d.ts.map +1 -0
  96. package/dist/bridge/sse/late-events.test.js +163 -0
  97. package/dist/bridge/sse/late-events.test.js.map +1 -0
  98. package/dist/bridge/sse/mapper.d.ts.map +1 -1
  99. package/dist/bridge/sse/mapper.js +100 -0
  100. package/dist/bridge/sse/mapper.js.map +1 -1
  101. package/dist/compaction/__tests__/cleared-result-recovery.test.d.ts +2 -0
  102. package/dist/compaction/__tests__/cleared-result-recovery.test.d.ts.map +1 -0
  103. package/dist/compaction/__tests__/cleared-result-recovery.test.js +101 -0
  104. package/dist/compaction/__tests__/cleared-result-recovery.test.js.map +1 -0
  105. package/dist/compaction/__tests__/context-window.test.d.ts +2 -0
  106. package/dist/compaction/__tests__/context-window.test.d.ts.map +1 -0
  107. package/dist/compaction/__tests__/context-window.test.js +62 -0
  108. package/dist/compaction/__tests__/context-window.test.js.map +1 -0
  109. package/dist/compaction/__tests__/dangling.test.js +16 -5
  110. package/dist/compaction/__tests__/dangling.test.js.map +1 -1
  111. package/dist/compaction/__tests__/lossiness.test.d.ts +2 -0
  112. package/dist/compaction/__tests__/lossiness.test.d.ts.map +1 -0
  113. package/dist/compaction/__tests__/lossiness.test.js +163 -0
  114. package/dist/compaction/__tests__/lossiness.test.js.map +1 -0
  115. package/dist/compaction/__tests__/state-across-resume.test.d.ts +2 -0
  116. package/dist/compaction/__tests__/state-across-resume.test.d.ts.map +1 -0
  117. package/dist/compaction/__tests__/state-across-resume.test.js +88 -0
  118. package/dist/compaction/__tests__/state-across-resume.test.js.map +1 -0
  119. package/dist/compaction/__tests__/tool-result-editing.test.d.ts +2 -0
  120. package/dist/compaction/__tests__/tool-result-editing.test.d.ts.map +1 -0
  121. package/dist/compaction/__tests__/tool-result-editing.test.js +184 -0
  122. package/dist/compaction/__tests__/tool-result-editing.test.js.map +1 -0
  123. package/dist/compaction/__tests__/trim-opens-on-user.test.d.ts +2 -0
  124. package/dist/compaction/__tests__/trim-opens-on-user.test.d.ts.map +1 -0
  125. package/dist/compaction/__tests__/trim-opens-on-user.test.js +109 -0
  126. package/dist/compaction/__tests__/trim-opens-on-user.test.js.map +1 -0
  127. package/dist/compaction/__tests__/verifier-model.test.d.ts +2 -0
  128. package/dist/compaction/__tests__/verifier-model.test.d.ts.map +1 -0
  129. package/dist/compaction/__tests__/verifier-model.test.js +58 -0
  130. package/dist/compaction/__tests__/verifier-model.test.js.map +1 -0
  131. package/dist/compaction/context-window.d.ts +44 -0
  132. package/dist/compaction/context-window.d.ts.map +1 -0
  133. package/dist/compaction/context-window.js +97 -0
  134. package/dist/compaction/context-window.js.map +1 -0
  135. package/dist/compaction/dangling.js +48 -1
  136. package/dist/compaction/dangling.js.map +1 -1
  137. package/dist/compaction/extractor.d.ts.map +1 -1
  138. package/dist/compaction/extractor.js +55 -1
  139. package/dist/compaction/extractor.js.map +1 -1
  140. package/dist/compaction/index.d.ts +5 -0
  141. package/dist/compaction/index.d.ts.map +1 -1
  142. package/dist/compaction/index.js +3 -0
  143. package/dist/compaction/index.js.map +1 -1
  144. package/dist/compaction/manager.d.ts +25 -0
  145. package/dist/compaction/manager.d.ts.map +1 -1
  146. package/dist/compaction/manager.js +42 -8
  147. package/dist/compaction/manager.js.map +1 -1
  148. package/dist/compaction/managers/structured.d.ts.map +1 -1
  149. package/dist/compaction/managers/structured.js +7 -3
  150. package/dist/compaction/managers/structured.js.map +1 -1
  151. package/dist/compaction/retention.d.ts +15 -0
  152. package/dist/compaction/retention.d.ts.map +1 -0
  153. package/dist/compaction/retention.js +89 -0
  154. package/dist/compaction/retention.js.map +1 -0
  155. package/dist/compaction/serializer.d.ts.map +1 -1
  156. package/dist/compaction/serializer.js +28 -11
  157. package/dist/compaction/serializer.js.map +1 -1
  158. package/dist/compaction/tool-result-editing.d.ts +54 -0
  159. package/dist/compaction/tool-result-editing.d.ts.map +1 -0
  160. package/dist/compaction/tool-result-editing.js +164 -0
  161. package/dist/compaction/tool-result-editing.js.map +1 -0
  162. package/dist/compaction/types.d.ts +10 -0
  163. package/dist/compaction/types.d.ts.map +1 -1
  164. package/dist/compaction/verifier.d.ts +20 -1
  165. package/dist/compaction/verifier.d.ts.map +1 -1
  166. package/dist/compaction/verifier.js +11 -2
  167. package/dist/compaction/verifier.js.map +1 -1
  168. package/dist/compaction/wire.d.ts +54 -0
  169. package/dist/compaction/wire.d.ts.map +1 -0
  170. package/dist/compaction/wire.js +45 -0
  171. package/dist/compaction/wire.js.map +1 -0
  172. package/dist/config/runtime.d.ts +161 -45
  173. package/dist/config/runtime.d.ts.map +1 -1
  174. package/dist/config/runtime.js +61 -19
  175. package/dist/config/runtime.js.map +1 -1
  176. package/dist/connector/builtins/http.d.ts.map +1 -1
  177. package/dist/connector/builtins/http.js +14 -1
  178. package/dist/connector/builtins/http.js.map +1 -1
  179. package/dist/connector/builtins/http.test.js +12 -0
  180. package/dist/connector/builtins/http.test.js.map +1 -1
  181. package/dist/connector/mcp/__tests__/discovery-policy.test.d.ts +2 -0
  182. package/dist/connector/mcp/__tests__/discovery-policy.test.d.ts.map +1 -0
  183. package/dist/connector/mcp/__tests__/discovery-policy.test.js +109 -0
  184. package/dist/connector/mcp/__tests__/discovery-policy.test.js.map +1 -0
  185. package/dist/connector/mcp/__tests__/list-pagination.test.d.ts +2 -0
  186. package/dist/connector/mcp/__tests__/list-pagination.test.d.ts.map +1 -0
  187. package/dist/connector/mcp/__tests__/list-pagination.test.js +81 -0
  188. package/dist/connector/mcp/__tests__/list-pagination.test.js.map +1 -0
  189. package/dist/connector/mcp/__tests__/policy.test.d.ts +2 -0
  190. package/dist/connector/mcp/__tests__/policy.test.d.ts.map +1 -0
  191. package/dist/connector/mcp/__tests__/policy.test.js +86 -0
  192. package/dist/connector/mcp/__tests__/policy.test.js.map +1 -0
  193. package/dist/connector/mcp/__tests__/protocol-negotiation.test.d.ts +2 -0
  194. package/dist/connector/mcp/__tests__/protocol-negotiation.test.d.ts.map +1 -0
  195. package/dist/connector/mcp/__tests__/protocol-negotiation.test.js +81 -0
  196. package/dist/connector/mcp/__tests__/protocol-negotiation.test.js.map +1 -0
  197. package/dist/connector/mcp/__tests__/schema-fidelity.test.d.ts +2 -0
  198. package/dist/connector/mcp/__tests__/schema-fidelity.test.d.ts.map +1 -0
  199. package/dist/connector/mcp/__tests__/schema-fidelity.test.js +206 -0
  200. package/dist/connector/mcp/__tests__/schema-fidelity.test.js.map +1 -0
  201. package/dist/connector/mcp/__tests__/structured-result.test.d.ts +2 -0
  202. package/dist/connector/mcp/__tests__/structured-result.test.d.ts.map +1 -0
  203. package/dist/connector/mcp/__tests__/structured-result.test.js +112 -0
  204. package/dist/connector/mcp/__tests__/structured-result.test.js.map +1 -0
  205. package/dist/connector/mcp/adapter.d.ts +14 -0
  206. package/dist/connector/mcp/adapter.d.ts.map +1 -1
  207. package/dist/connector/mcp/adapter.js +292 -18
  208. package/dist/connector/mcp/adapter.js.map +1 -1
  209. package/dist/connector/mcp/client.d.ts +38 -0
  210. package/dist/connector/mcp/client.d.ts.map +1 -1
  211. package/dist/connector/mcp/client.js +138 -12
  212. package/dist/connector/mcp/client.js.map +1 -1
  213. package/dist/connector/mcp/client.test.d.ts +2 -0
  214. package/dist/connector/mcp/client.test.d.ts.map +1 -0
  215. package/dist/connector/mcp/client.test.js +105 -0
  216. package/dist/connector/mcp/client.test.js.map +1 -0
  217. package/dist/connector/mcp/discovery.d.ts +32 -1
  218. package/dist/connector/mcp/discovery.d.ts.map +1 -1
  219. package/dist/connector/mcp/discovery.js +41 -4
  220. package/dist/connector/mcp/discovery.js.map +1 -1
  221. package/dist/connector/mcp/http-sse.d.ts.map +1 -1
  222. package/dist/connector/mcp/http-sse.js +15 -1
  223. package/dist/connector/mcp/http-sse.js.map +1 -1
  224. package/dist/connector/mcp/index.d.ts +3 -0
  225. package/dist/connector/mcp/index.d.ts.map +1 -1
  226. package/dist/connector/mcp/index.js +1 -0
  227. package/dist/connector/mcp/index.js.map +1 -1
  228. package/dist/connector/mcp/policy.d.ts +57 -0
  229. package/dist/connector/mcp/policy.d.ts.map +1 -0
  230. package/dist/connector/mcp/policy.js +84 -0
  231. package/dist/connector/mcp/policy.js.map +1 -0
  232. package/dist/connector/mcp/schema-refs.d.ts +26 -0
  233. package/dist/connector/mcp/schema-refs.d.ts.map +1 -0
  234. package/dist/connector/mcp/schema-refs.js +111 -0
  235. package/dist/connector/mcp/schema-refs.js.map +1 -0
  236. package/dist/constants/agent/index.d.ts.map +1 -1
  237. package/dist/constants/agent/index.js +3 -0
  238. package/dist/constants/agent/index.js.map +1 -1
  239. package/dist/constants/compaction/index.js +1 -1
  240. package/dist/constants/compaction/index.js.map +1 -1
  241. package/dist/constants/continuation.d.ts +4 -4
  242. package/dist/constants/continuation.js +4 -4
  243. package/dist/constants/mcp/index.d.ts +31 -0
  244. package/dist/constants/mcp/index.d.ts.map +1 -1
  245. package/dist/constants/mcp/index.js +31 -0
  246. package/dist/constants/mcp/index.js.map +1 -1
  247. package/dist/constants/plugin/index.d.ts +23 -2
  248. package/dist/constants/plugin/index.d.ts.map +1 -1
  249. package/dist/constants/plugin/index.js +23 -2
  250. package/dist/constants/plugin/index.js.map +1 -1
  251. package/dist/constants/sandbox/index.d.ts +2 -3
  252. package/dist/constants/sandbox/index.d.ts.map +1 -1
  253. package/dist/constants/sandbox/index.js +6 -5
  254. package/dist/constants/sandbox/index.js.map +1 -1
  255. package/dist/constants/telemetry/index.d.ts +7 -0
  256. package/dist/constants/telemetry/index.d.ts.map +1 -1
  257. package/dist/constants/telemetry/index.js +7 -0
  258. package/dist/constants/telemetry/index.js.map +1 -1
  259. package/dist/constants/tools/index.d.ts +16 -0
  260. package/dist/constants/tools/index.d.ts.map +1 -1
  261. package/dist/constants/tools/index.js +16 -0
  262. package/dist/constants/tools/index.js.map +1 -1
  263. package/dist/contracts/a2a.d.ts +94 -94
  264. package/dist/contracts/api.d.ts +40 -2
  265. package/dist/contracts/api.d.ts.map +1 -1
  266. package/dist/contracts/run-status.d.ts +17 -0
  267. package/dist/contracts/run-status.d.ts.map +1 -0
  268. package/dist/contracts/run-status.js +41 -0
  269. package/dist/contracts/run-status.js.map +1 -0
  270. package/dist/contracts/schemas.d.ts +6 -6
  271. package/dist/eval/__tests__/case-deadline.test.d.ts +2 -0
  272. package/dist/eval/__tests__/case-deadline.test.d.ts.map +1 -0
  273. package/dist/eval/__tests__/case-deadline.test.js +128 -0
  274. package/dist/eval/__tests__/case-deadline.test.js.map +1 -0
  275. package/dist/eval/__tests__/end-to-end.test.d.ts +2 -0
  276. package/dist/eval/__tests__/end-to-end.test.d.ts.map +1 -0
  277. package/dist/eval/__tests__/end-to-end.test.js +223 -0
  278. package/dist/eval/__tests__/end-to-end.test.js.map +1 -0
  279. package/dist/eval/__tests__/harness-honesty.test.d.ts +2 -0
  280. package/dist/eval/__tests__/harness-honesty.test.d.ts.map +1 -0
  281. package/dist/eval/__tests__/harness-honesty.test.js +85 -0
  282. package/dist/eval/__tests__/harness-honesty.test.js.map +1 -0
  283. package/dist/eval/__tests__/harness.test.d.ts +2 -0
  284. package/dist/eval/__tests__/harness.test.d.ts.map +1 -0
  285. package/dist/eval/__tests__/harness.test.js +207 -0
  286. package/dist/eval/__tests__/harness.test.js.map +1 -0
  287. package/dist/eval/__tests__/judge.test.d.ts +2 -0
  288. package/dist/eval/__tests__/judge.test.d.ts.map +1 -0
  289. package/dist/eval/__tests__/judge.test.js +255 -0
  290. package/dist/eval/__tests__/judge.test.js.map +1 -0
  291. package/dist/eval/__tests__/scorer-gates.test.d.ts +2 -0
  292. package/dist/eval/__tests__/scorer-gates.test.d.ts.map +1 -0
  293. package/dist/eval/__tests__/scorer-gates.test.js +136 -0
  294. package/dist/eval/__tests__/scorer-gates.test.js.map +1 -0
  295. package/dist/eval/experiment.d.ts +54 -0
  296. package/dist/eval/experiment.d.ts.map +1 -0
  297. package/dist/eval/experiment.js +263 -0
  298. package/dist/eval/experiment.js.map +1 -0
  299. package/dist/eval/from-run.d.ts +30 -0
  300. package/dist/eval/from-run.d.ts.map +1 -0
  301. package/dist/eval/from-run.js +46 -0
  302. package/dist/eval/from-run.js.map +1 -0
  303. package/dist/eval/index.d.ts +8 -0
  304. package/dist/eval/index.d.ts.map +1 -0
  305. package/dist/eval/index.js +5 -0
  306. package/dist/eval/index.js.map +1 -0
  307. package/dist/eval/judge.d.ts +63 -0
  308. package/dist/eval/judge.d.ts.map +1 -0
  309. package/dist/eval/judge.js +136 -0
  310. package/dist/eval/judge.js.map +1 -0
  311. package/dist/eval/scorers.d.ts +40 -0
  312. package/dist/eval/scorers.d.ts.map +1 -0
  313. package/dist/eval/scorers.js +151 -0
  314. package/dist/eval/scorers.js.map +1 -0
  315. package/dist/eval/types.d.ts +150 -0
  316. package/dist/eval/types.d.ts.map +1 -0
  317. package/dist/eval/types.js +2 -0
  318. package/dist/eval/types.js.map +1 -0
  319. package/dist/execution/local.d.ts.map +1 -1
  320. package/dist/execution/local.js +6 -1
  321. package/dist/execution/local.js.map +1 -1
  322. package/dist/execution/local.test.d.ts +2 -0
  323. package/dist/execution/local.test.d.ts.map +1 -0
  324. package/dist/execution/local.test.js +37 -0
  325. package/dist/execution/local.test.js.map +1 -0
  326. package/dist/gateway/__tests__/list-survives-eviction.test.d.ts +2 -0
  327. package/dist/gateway/__tests__/list-survives-eviction.test.d.ts.map +1 -0
  328. package/dist/gateway/__tests__/list-survives-eviction.test.js +102 -0
  329. package/dist/gateway/__tests__/list-survives-eviction.test.js.map +1 -0
  330. package/dist/gateway/local.d.ts +42 -2
  331. package/dist/gateway/local.d.ts.map +1 -1
  332. package/dist/gateway/local.js +110 -3
  333. package/dist/gateway/local.js.map +1 -1
  334. package/dist/manager/agent/__tests__/dispose-cancels.test.d.ts +2 -0
  335. package/dist/manager/agent/__tests__/dispose-cancels.test.d.ts.map +1 -0
  336. package/dist/manager/agent/__tests__/dispose-cancels.test.js +88 -0
  337. package/dist/manager/agent/__tests__/dispose-cancels.test.js.map +1 -0
  338. package/dist/manager/agent/__tests__/lifecycle.test.js +230 -0
  339. package/dist/manager/agent/__tests__/lifecycle.test.js.map +1 -1
  340. package/dist/manager/agent/lifecycle.d.ts +35 -0
  341. package/dist/manager/agent/lifecycle.d.ts.map +1 -1
  342. package/dist/manager/agent/lifecycle.js +117 -7
  343. package/dist/manager/agent/lifecycle.js.map +1 -1
  344. package/dist/manager/run/persistence.d.ts +90 -1
  345. package/dist/manager/run/persistence.d.ts.map +1 -1
  346. package/dist/manager/run/persistence.js +120 -0
  347. package/dist/manager/run/persistence.js.map +1 -1
  348. package/dist/plugin/__tests__/enable-contributions.test.js +3 -3
  349. package/dist/plugin/__tests__/enable-contributions.test.js.map +1 -1
  350. package/dist/plugin/__tests__/hook-ordering.test.d.ts +2 -0
  351. package/dist/plugin/__tests__/hook-ordering.test.d.ts.map +1 -0
  352. package/dist/plugin/__tests__/hook-ordering.test.js +227 -0
  353. package/dist/plugin/__tests__/hook-ordering.test.js.map +1 -0
  354. package/dist/plugin/__tests__/install-time-refusal.test.d.ts +2 -0
  355. package/dist/plugin/__tests__/install-time-refusal.test.d.ts.map +1 -0
  356. package/dist/plugin/__tests__/install-time-refusal.test.js +49 -0
  357. package/dist/plugin/__tests__/install-time-refusal.test.js.map +1 -0
  358. package/dist/plugin/__tests__/lifecycle.test.js +95 -68
  359. package/dist/plugin/__tests__/lifecycle.test.js.map +1 -1
  360. package/dist/plugin/lifecycle.d.ts +15 -1
  361. package/dist/plugin/lifecycle.d.ts.map +1 -1
  362. package/dist/plugin/lifecycle.js +67 -20
  363. package/dist/plugin/lifecycle.js.map +1 -1
  364. package/dist/plugin/loader.d.ts +15 -0
  365. package/dist/plugin/loader.d.ts.map +1 -1
  366. package/dist/plugin/loader.js +33 -1
  367. package/dist/plugin/loader.js.map +1 -1
  368. package/dist/probe/registry.d.ts.map +1 -1
  369. package/dist/probe/registry.js +19 -1
  370. package/dist/probe/registry.js.map +1 -1
  371. package/dist/probe/registry.test.d.ts +9 -3
  372. package/dist/probe/registry.test.d.ts.map +1 -1
  373. package/dist/probe/registry.test.js +44 -6
  374. package/dist/probe/registry.test.js.map +1 -1
  375. package/dist/provider/__tests__/mock-script.test.d.ts +2 -0
  376. package/dist/provider/__tests__/mock-script.test.d.ts.map +1 -0
  377. package/dist/provider/__tests__/mock-script.test.js +151 -0
  378. package/dist/provider/__tests__/mock-script.test.js.map +1 -0
  379. package/dist/provider/__tests__/retry-visibility.test.d.ts +2 -0
  380. package/dist/provider/__tests__/retry-visibility.test.d.ts.map +1 -0
  381. package/dist/provider/__tests__/retry-visibility.test.js +137 -0
  382. package/dist/provider/__tests__/retry-visibility.test.js.map +1 -0
  383. package/dist/provider/__tests__/retry.test.d.ts +2 -0
  384. package/dist/provider/__tests__/retry.test.d.ts.map +1 -0
  385. package/dist/provider/__tests__/retry.test.js +259 -0
  386. package/dist/provider/__tests__/retry.test.js.map +1 -0
  387. package/dist/provider/capabilities.d.ts +1 -0
  388. package/dist/provider/capabilities.d.ts.map +1 -1
  389. package/dist/provider/capabilities.js +2 -0
  390. package/dist/provider/capabilities.js.map +1 -1
  391. package/dist/provider/collect.d.ts +1 -1
  392. package/dist/provider/collect.js +1 -1
  393. package/dist/provider/collect.test.d.ts +1 -1
  394. package/dist/provider/collect.test.js +1 -1
  395. package/dist/provider/index.d.ts +3 -0
  396. package/dist/provider/index.d.ts.map +1 -1
  397. package/dist/provider/index.js +2 -0
  398. package/dist/provider/index.js.map +1 -1
  399. package/dist/provider/mock-register.d.ts.map +1 -1
  400. package/dist/provider/mock-register.js +5 -2
  401. package/dist/provider/mock-register.js.map +1 -1
  402. package/dist/provider/mock.d.ts +35 -5
  403. package/dist/provider/mock.d.ts.map +1 -1
  404. package/dist/provider/mock.js +104 -32
  405. package/dist/provider/mock.js.map +1 -1
  406. package/dist/provider/retry.d.ts +42 -0
  407. package/dist/provider/retry.d.ts.map +1 -0
  408. package/dist/provider/retry.js +161 -0
  409. package/dist/provider/retry.js.map +1 -0
  410. package/dist/public-runtime.d.ts +26 -7
  411. package/dist/public-runtime.d.ts.map +1 -1
  412. package/dist/public-runtime.js +50 -7
  413. package/dist/public-runtime.js.map +1 -1
  414. package/dist/public-types.d.ts +5 -1
  415. package/dist/public-types.d.ts.map +1 -1
  416. package/dist/rag/__tests__/retrieval-math.test.d.ts +2 -0
  417. package/dist/rag/__tests__/retrieval-math.test.d.ts.map +1 -0
  418. package/dist/rag/__tests__/retrieval-math.test.js +46 -0
  419. package/dist/rag/__tests__/retrieval-math.test.js.map +1 -0
  420. package/dist/rag/chunking.d.ts +12 -0
  421. package/dist/rag/chunking.d.ts.map +1 -1
  422. package/dist/rag/chunking.js +32 -1
  423. package/dist/rag/chunking.js.map +1 -1
  424. package/dist/rag/embedding.d.ts +4 -4
  425. package/dist/rag/embedding.d.ts.map +1 -1
  426. package/dist/rag/embedding.js +23 -3
  427. package/dist/rag/embedding.js.map +1 -1
  428. package/dist/rag/embedding.test.d.ts +1 -1
  429. package/dist/rag/embedding.test.js +59 -10
  430. package/dist/rag/embedding.test.js.map +1 -1
  431. package/dist/rag/index.d.ts +1 -1
  432. package/dist/rag/index.d.ts.map +1 -1
  433. package/dist/rag/index.js +1 -1
  434. package/dist/rag/index.js.map +1 -1
  435. package/dist/rag/retriever.d.ts +0 -1
  436. package/dist/rag/retriever.d.ts.map +1 -1
  437. package/dist/rag/retriever.js +75 -22
  438. package/dist/rag/retriever.js.map +1 -1
  439. package/dist/rag/vector-store.test.d.ts +1 -1
  440. package/dist/rag/vector-store.test.js +2 -2
  441. package/dist/rag/vector-store.test.js.map +1 -1
  442. package/dist/registry/ManagedRegistry.test.d.ts +1 -1
  443. package/dist/registry/ManagedRegistry.test.js +1 -1
  444. package/dist/registry/Registry.test.d.ts +1 -1
  445. package/dist/registry/Registry.test.js +1 -1
  446. package/dist/registry/connector/definitions.test.d.ts +1 -1
  447. package/dist/registry/connector/definitions.test.js +1 -1
  448. package/dist/registry/tool/__tests__/schema.test.d.ts +2 -0
  449. package/dist/registry/tool/__tests__/schema.test.d.ts.map +1 -0
  450. package/dist/registry/tool/__tests__/schema.test.js +176 -0
  451. package/dist/registry/tool/__tests__/schema.test.js.map +1 -0
  452. package/dist/registry/tool/__tests__/tool-name-validation.test.d.ts +2 -0
  453. package/dist/registry/tool/__tests__/tool-name-validation.test.d.ts.map +1 -0
  454. package/dist/registry/tool/__tests__/tool-name-validation.test.js +70 -0
  455. package/dist/registry/tool/__tests__/tool-name-validation.test.js.map +1 -0
  456. package/dist/registry/tool/execute.d.ts +37 -0
  457. package/dist/registry/tool/execute.d.ts.map +1 -1
  458. package/dist/registry/tool/execute.js +91 -21
  459. package/dist/registry/tool/execute.js.map +1 -1
  460. package/dist/registry/tool/schema.d.ts +26 -0
  461. package/dist/registry/tool/schema.d.ts.map +1 -0
  462. package/dist/registry/tool/schema.js +77 -0
  463. package/dist/registry/tool/schema.js.map +1 -0
  464. package/dist/registry/toolset/catalog.d.ts.map +1 -1
  465. package/dist/registry/toolset/catalog.js +0 -1
  466. package/dist/registry/toolset/catalog.js.map +1 -1
  467. package/dist/registry/toolset/catalog.test.js +0 -1
  468. package/dist/registry/toolset/catalog.test.js.map +1 -1
  469. package/dist/run/reporter.d.ts.map +1 -1
  470. package/dist/run/reporter.js +76 -0
  471. package/dist/run/reporter.js.map +1 -1
  472. package/dist/runtime/bidi/__tests__/session.test.d.ts +2 -0
  473. package/dist/runtime/bidi/__tests__/session.test.d.ts.map +1 -0
  474. package/dist/runtime/bidi/__tests__/session.test.js +201 -0
  475. package/dist/runtime/bidi/__tests__/session.test.js.map +1 -0
  476. package/dist/runtime/bidi/index.d.ts +5 -0
  477. package/dist/runtime/bidi/index.d.ts.map +1 -0
  478. package/dist/runtime/bidi/index.js +3 -0
  479. package/dist/runtime/bidi/index.js.map +1 -0
  480. package/dist/runtime/bidi/mock.d.ts +34 -0
  481. package/dist/runtime/bidi/mock.d.ts.map +1 -0
  482. package/dist/runtime/bidi/mock.js +60 -0
  483. package/dist/runtime/bidi/mock.js.map +1 -0
  484. package/dist/runtime/bidi/session.d.ts +44 -0
  485. package/dist/runtime/bidi/session.d.ts.map +1 -0
  486. package/dist/runtime/bidi/session.js +154 -0
  487. package/dist/runtime/bidi/session.js.map +1 -0
  488. package/dist/runtime/decision/parser.d.ts.map +1 -1
  489. package/dist/runtime/decision/parser.js +8 -1
  490. package/dist/runtime/decision/parser.js.map +1 -1
  491. package/dist/runtime/decision/parser.test.d.ts +2 -0
  492. package/dist/runtime/decision/parser.test.d.ts.map +1 -0
  493. package/dist/runtime/decision/parser.test.js +37 -0
  494. package/dist/runtime/decision/parser.test.js.map +1 -0
  495. package/dist/runtime/query/__tests__/answer-review.test.d.ts +2 -0
  496. package/dist/runtime/query/__tests__/answer-review.test.d.ts.map +1 -0
  497. package/dist/runtime/query/__tests__/answer-review.test.js +141 -0
  498. package/dist/runtime/query/__tests__/answer-review.test.js.map +1 -0
  499. package/dist/runtime/query/__tests__/capability-negotiation.test.js +24 -111
  500. package/dist/runtime/query/__tests__/capability-negotiation.test.js.map +1 -1
  501. package/dist/runtime/query/__tests__/checkpoint-store.test.js +12 -44
  502. package/dist/runtime/query/__tests__/checkpoint-store.test.js.map +1 -1
  503. package/dist/runtime/query/__tests__/checkpoint.test.js +2 -10
  504. package/dist/runtime/query/__tests__/checkpoint.test.js.map +1 -1
  505. package/dist/runtime/query/__tests__/citations.test.d.ts +2 -0
  506. package/dist/runtime/query/__tests__/citations.test.d.ts.map +1 -0
  507. package/dist/runtime/query/__tests__/citations.test.js +85 -0
  508. package/dist/runtime/query/__tests__/citations.test.js.map +1 -0
  509. package/dist/runtime/query/__tests__/content-budget.test.d.ts +2 -0
  510. package/dist/runtime/query/__tests__/content-budget.test.d.ts.map +1 -0
  511. package/dist/runtime/query/__tests__/content-budget.test.js +69 -0
  512. package/dist/runtime/query/__tests__/content-budget.test.js.map +1 -0
  513. package/dist/runtime/query/__tests__/context.test.js +3 -2
  514. package/dist/runtime/query/__tests__/context.test.js.map +1 -1
  515. package/dist/runtime/query/__tests__/deferred-tools.test.js +21 -130
  516. package/dist/runtime/query/__tests__/deferred-tools.test.js.map +1 -1
  517. package/dist/runtime/query/__tests__/delegation-resume.test.d.ts +2 -0
  518. package/dist/runtime/query/__tests__/delegation-resume.test.d.ts.map +1 -0
  519. package/dist/runtime/query/__tests__/delegation-resume.test.js +111 -0
  520. package/dist/runtime/query/__tests__/delegation-resume.test.js.map +1 -0
  521. package/dist/runtime/query/__tests__/document-attachments.test.d.ts +2 -0
  522. package/dist/runtime/query/__tests__/document-attachments.test.d.ts.map +1 -0
  523. package/dist/runtime/query/__tests__/document-attachments.test.js +92 -0
  524. package/dist/runtime/query/__tests__/document-attachments.test.js.map +1 -0
  525. package/dist/runtime/query/__tests__/durable-park-and-trace.test.d.ts +2 -0
  526. package/dist/runtime/query/__tests__/durable-park-and-trace.test.d.ts.map +1 -0
  527. package/dist/runtime/query/__tests__/durable-park-and-trace.test.js +194 -0
  528. package/dist/runtime/query/__tests__/durable-park-and-trace.test.js.map +1 -0
  529. package/dist/runtime/query/__tests__/durable-question-park.test.d.ts +2 -0
  530. package/dist/runtime/query/__tests__/durable-question-park.test.d.ts.map +1 -0
  531. package/dist/runtime/query/__tests__/durable-question-park.test.js +253 -0
  532. package/dist/runtime/query/__tests__/durable-question-park.test.js.map +1 -0
  533. package/dist/runtime/query/__tests__/durable-run-state.test.d.ts +2 -0
  534. package/dist/runtime/query/__tests__/durable-run-state.test.d.ts.map +1 -0
  535. package/dist/runtime/query/__tests__/durable-run-state.test.js +158 -0
  536. package/dist/runtime/query/__tests__/durable-run-state.test.js.map +1 -0
  537. package/dist/runtime/query/__tests__/executor-plugin-hooks.test.js +24 -20
  538. package/dist/runtime/query/__tests__/executor-plugin-hooks.test.js.map +1 -1
  539. package/dist/runtime/query/__tests__/finish-reason-vs-tool-calls.test.d.ts +2 -0
  540. package/dist/runtime/query/__tests__/finish-reason-vs-tool-calls.test.d.ts.map +1 -0
  541. package/dist/runtime/query/__tests__/finish-reason-vs-tool-calls.test.js +187 -0
  542. package/dist/runtime/query/__tests__/finish-reason-vs-tool-calls.test.js.map +1 -0
  543. package/dist/runtime/query/__tests__/grant-cannot-overrule-the-gate.test.d.ts +2 -0
  544. package/dist/runtime/query/__tests__/grant-cannot-overrule-the-gate.test.d.ts.map +1 -0
  545. package/dist/runtime/query/__tests__/grant-cannot-overrule-the-gate.test.js +95 -0
  546. package/dist/runtime/query/__tests__/grant-cannot-overrule-the-gate.test.js.map +1 -0
  547. package/dist/runtime/query/__tests__/guardrails-e2e.test.d.ts +2 -0
  548. package/dist/runtime/query/__tests__/guardrails-e2e.test.d.ts.map +1 -0
  549. package/dist/runtime/query/__tests__/guardrails-e2e.test.js +132 -0
  550. package/dist/runtime/query/__tests__/guardrails-e2e.test.js.map +1 -0
  551. package/dist/runtime/query/__tests__/guardrails.test.d.ts +2 -0
  552. package/dist/runtime/query/__tests__/guardrails.test.d.ts.map +1 -0
  553. package/dist/runtime/query/__tests__/guardrails.test.js +163 -0
  554. package/dist/runtime/query/__tests__/guardrails.test.js.map +1 -0
  555. package/dist/runtime/query/__tests__/hook-model-visibility.test.d.ts +2 -0
  556. package/dist/runtime/query/__tests__/hook-model-visibility.test.d.ts.map +1 -0
  557. package/dist/runtime/query/__tests__/hook-model-visibility.test.js +119 -0
  558. package/dist/runtime/query/__tests__/hook-model-visibility.test.js.map +1 -0
  559. package/dist/runtime/query/__tests__/long-document-flow.test.js +9 -2
  560. package/dist/runtime/query/__tests__/long-document-flow.test.js.map +1 -1
  561. package/dist/runtime/query/__tests__/memory-promotion.test.d.ts +2 -0
  562. package/dist/runtime/query/__tests__/memory-promotion.test.d.ts.map +1 -0
  563. package/dist/runtime/query/__tests__/memory-promotion.test.js +119 -0
  564. package/dist/runtime/query/__tests__/memory-promotion.test.js.map +1 -0
  565. package/dist/runtime/query/__tests__/orphan-task-turn-end.test.js +5 -27
  566. package/dist/runtime/query/__tests__/orphan-task-turn-end.test.js.map +1 -1
  567. package/dist/runtime/query/__tests__/overflow-relief.test.d.ts +2 -0
  568. package/dist/runtime/query/__tests__/overflow-relief.test.d.ts.map +1 -0
  569. package/dist/runtime/query/__tests__/overflow-relief.test.js +119 -0
  570. package/dist/runtime/query/__tests__/overflow-relief.test.js.map +1 -0
  571. package/dist/runtime/query/__tests__/partial-batch-resume.test.d.ts +2 -0
  572. package/dist/runtime/query/__tests__/partial-batch-resume.test.d.ts.map +1 -0
  573. package/dist/runtime/query/__tests__/partial-batch-resume.test.js +205 -0
  574. package/dist/runtime/query/__tests__/partial-batch-resume.test.js.map +1 -0
  575. package/dist/runtime/query/__tests__/plugin-hooks.test.js +0 -1
  576. package/dist/runtime/query/__tests__/plugin-hooks.test.js.map +1 -1
  577. package/dist/runtime/query/__tests__/prepare-step-chain.test.d.ts +2 -0
  578. package/dist/runtime/query/__tests__/prepare-step-chain.test.d.ts.map +1 -0
  579. package/dist/runtime/query/__tests__/prepare-step-chain.test.js +104 -0
  580. package/dist/runtime/query/__tests__/prepare-step-chain.test.js.map +1 -0
  581. package/dist/runtime/query/__tests__/prepare-step.test.d.ts +2 -0
  582. package/dist/runtime/query/__tests__/prepare-step.test.d.ts.map +1 -0
  583. package/dist/runtime/query/__tests__/prepare-step.test.js +194 -0
  584. package/dist/runtime/query/__tests__/prepare-step.test.js.map +1 -0
  585. package/dist/runtime/query/__tests__/probe-veto-is-error.test.d.ts +2 -0
  586. package/dist/runtime/query/__tests__/probe-veto-is-error.test.d.ts.map +1 -0
  587. package/dist/runtime/query/__tests__/probe-veto-is-error.test.js +111 -0
  588. package/dist/runtime/query/__tests__/probe-veto-is-error.test.js.map +1 -0
  589. package/dist/runtime/query/__tests__/prompt.test.js +1 -1
  590. package/dist/runtime/query/__tests__/prompt.test.js.map +1 -1
  591. package/dist/runtime/query/__tests__/provider-fault-pauses.test.d.ts +2 -0
  592. package/dist/runtime/query/__tests__/provider-fault-pauses.test.d.ts.map +1 -0
  593. package/dist/runtime/query/__tests__/provider-fault-pauses.test.js +86 -0
  594. package/dist/runtime/query/__tests__/provider-fault-pauses.test.js.map +1 -0
  595. package/dist/runtime/query/__tests__/reasoning-stream.test.d.ts +2 -0
  596. package/dist/runtime/query/__tests__/reasoning-stream.test.d.ts.map +1 -0
  597. package/dist/runtime/query/__tests__/reasoning-stream.test.js +146 -0
  598. package/dist/runtime/query/__tests__/reasoning-stream.test.js.map +1 -0
  599. package/dist/runtime/query/__tests__/recoverable-settle.test.d.ts +2 -0
  600. package/dist/runtime/query/__tests__/recoverable-settle.test.d.ts.map +1 -0
  601. package/dist/runtime/query/__tests__/recoverable-settle.test.js +119 -0
  602. package/dist/runtime/query/__tests__/recoverable-settle.test.js.map +1 -0
  603. package/dist/runtime/query/__tests__/resume-budget.test.d.ts +2 -0
  604. package/dist/runtime/query/__tests__/resume-budget.test.d.ts.map +1 -0
  605. package/dist/runtime/query/__tests__/resume-budget.test.js +130 -0
  606. package/dist/runtime/query/__tests__/resume-budget.test.js.map +1 -0
  607. package/dist/runtime/query/__tests__/resume-pending.test.d.ts +2 -0
  608. package/dist/runtime/query/__tests__/resume-pending.test.d.ts.map +1 -0
  609. package/dist/runtime/query/__tests__/resume-pending.test.js +226 -0
  610. package/dist/runtime/query/__tests__/resume-pending.test.js.map +1 -0
  611. package/dist/runtime/query/__tests__/rich-content-budget-wiring.test.d.ts +2 -0
  612. package/dist/runtime/query/__tests__/rich-content-budget-wiring.test.d.ts.map +1 -0
  613. package/dist/runtime/query/__tests__/rich-content-budget-wiring.test.js +89 -0
  614. package/dist/runtime/query/__tests__/rich-content-budget-wiring.test.js.map +1 -0
  615. package/dist/runtime/query/__tests__/stop-condition.test.d.ts +2 -0
  616. package/dist/runtime/query/__tests__/stop-condition.test.d.ts.map +1 -0
  617. package/dist/runtime/query/__tests__/stop-condition.test.js +205 -0
  618. package/dist/runtime/query/__tests__/stop-condition.test.js.map +1 -0
  619. package/dist/runtime/query/__tests__/structured-output.test.d.ts +2 -0
  620. package/dist/runtime/query/__tests__/structured-output.test.d.ts.map +1 -0
  621. package/dist/runtime/query/__tests__/structured-output.test.js +219 -0
  622. package/dist/runtime/query/__tests__/structured-output.test.js.map +1 -0
  623. package/dist/runtime/query/__tests__/structured-run-failure.test.d.ts +2 -0
  624. package/dist/runtime/query/__tests__/structured-run-failure.test.d.ts.map +1 -0
  625. package/dist/runtime/query/__tests__/structured-run-failure.test.js +117 -0
  626. package/dist/runtime/query/__tests__/structured-run-failure.test.js.map +1 -0
  627. package/dist/runtime/query/__tests__/terminal-tool.test.d.ts +2 -0
  628. package/dist/runtime/query/__tests__/terminal-tool.test.d.ts.map +1 -0
  629. package/dist/runtime/query/__tests__/terminal-tool.test.js +119 -0
  630. package/dist/runtime/query/__tests__/terminal-tool.test.js.map +1 -0
  631. package/dist/runtime/query/__tests__/tool-grants.test.d.ts +2 -0
  632. package/dist/runtime/query/__tests__/tool-grants.test.d.ts.map +1 -0
  633. package/dist/runtime/query/__tests__/tool-grants.test.js +85 -0
  634. package/dist/runtime/query/__tests__/tool-grants.test.js.map +1 -0
  635. package/dist/runtime/query/__tests__/tool-output-budget.test.d.ts +2 -0
  636. package/dist/runtime/query/__tests__/tool-output-budget.test.d.ts.map +1 -0
  637. package/dist/runtime/query/__tests__/tool-output-budget.test.js +0 -0
  638. package/dist/runtime/query/__tests__/tool-output-budget.test.js.map +1 -0
  639. package/dist/runtime/query/__tests__/tool-pause.test.d.ts +2 -0
  640. package/dist/runtime/query/__tests__/tool-pause.test.d.ts.map +1 -0
  641. package/dist/runtime/query/__tests__/tool-pause.test.js +165 -0
  642. package/dist/runtime/query/__tests__/tool-pause.test.js.map +1 -0
  643. package/dist/runtime/query/__tests__/tool-progress.test.d.ts +2 -0
  644. package/dist/runtime/query/__tests__/tool-progress.test.d.ts.map +1 -0
  645. package/dist/runtime/query/__tests__/tool-progress.test.js +122 -0
  646. package/dist/runtime/query/__tests__/tool-progress.test.js.map +1 -0
  647. package/dist/runtime/query/__tests__/tool-repair-retry.test.d.ts +2 -0
  648. package/dist/runtime/query/__tests__/tool-repair-retry.test.d.ts.map +1 -0
  649. package/dist/runtime/query/__tests__/tool-repair-retry.test.js +196 -0
  650. package/dist/runtime/query/__tests__/tool-repair-retry.test.js.map +1 -0
  651. package/dist/runtime/query/__tests__/tool-result-wire.test.d.ts +2 -0
  652. package/dist/runtime/query/__tests__/tool-result-wire.test.d.ts.map +1 -0
  653. package/dist/runtime/query/__tests__/tool-result-wire.test.js +106 -0
  654. package/dist/runtime/query/__tests__/tool-result-wire.test.js.map +1 -0
  655. package/dist/runtime/query/__tests__/tool-timeout.test.d.ts +2 -0
  656. package/dist/runtime/query/__tests__/tool-timeout.test.d.ts.map +1 -0
  657. package/dist/runtime/query/__tests__/tool-timeout.test.js +184 -0
  658. package/dist/runtime/query/__tests__/tool-timeout.test.js.map +1 -0
  659. package/dist/runtime/query/__tests__/trace-parenting.test.d.ts +2 -0
  660. package/dist/runtime/query/__tests__/trace-parenting.test.d.ts.map +1 -0
  661. package/dist/runtime/query/__tests__/trace-parenting.test.js +111 -0
  662. package/dist/runtime/query/__tests__/trace-parenting.test.js.map +1 -0
  663. package/dist/runtime/query/__tests__/wiring-seams.test.d.ts +2 -0
  664. package/dist/runtime/query/__tests__/wiring-seams.test.d.ts.map +1 -0
  665. package/dist/runtime/query/__tests__/wiring-seams.test.js +237 -0
  666. package/dist/runtime/query/__tests__/wiring-seams.test.js.map +1 -0
  667. package/dist/runtime/query/checkpoint.d.ts +124 -6
  668. package/dist/runtime/query/checkpoint.d.ts.map +1 -1
  669. package/dist/runtime/query/checkpoint.js +207 -6
  670. package/dist/runtime/query/checkpoint.js.map +1 -1
  671. package/dist/runtime/query/events.js +2 -2
  672. package/dist/runtime/query/events.js.map +1 -1
  673. package/dist/runtime/query/executor.d.ts +224 -7
  674. package/dist/runtime/query/executor.d.ts.map +1 -1
  675. package/dist/runtime/query/executor.js +662 -35
  676. package/dist/runtime/query/executor.js.map +1 -1
  677. package/dist/runtime/query/guard.d.ts +18 -0
  678. package/dist/runtime/query/guard.d.ts.map +1 -1
  679. package/dist/runtime/query/guard.js +16 -1
  680. package/dist/runtime/query/guard.js.map +1 -1
  681. package/dist/runtime/query/guardrail-presets.d.ts +36 -0
  682. package/dist/runtime/query/guardrail-presets.d.ts.map +1 -0
  683. package/dist/runtime/query/guardrail-presets.js +96 -0
  684. package/dist/runtime/query/guardrail-presets.js.map +1 -0
  685. package/dist/runtime/query/guardrails.d.ts +35 -0
  686. package/dist/runtime/query/guardrails.d.ts.map +1 -0
  687. package/dist/runtime/query/guardrails.js +102 -0
  688. package/dist/runtime/query/guardrails.js.map +1 -0
  689. package/dist/runtime/query/index.d.ts +213 -2
  690. package/dist/runtime/query/index.d.ts.map +1 -1
  691. package/dist/runtime/query/index.js +458 -10
  692. package/dist/runtime/query/index.js.map +1 -1
  693. package/dist/runtime/query/iteration/__tests__/cancel-settles-turn.test.d.ts +2 -0
  694. package/dist/runtime/query/iteration/__tests__/cancel-settles-turn.test.d.ts.map +1 -0
  695. package/dist/runtime/query/iteration/__tests__/cancel-settles-turn.test.js +131 -0
  696. package/dist/runtime/query/iteration/__tests__/cancel-settles-turn.test.js.map +1 -0
  697. package/dist/runtime/query/iteration/index.d.ts +76 -1
  698. package/dist/runtime/query/iteration/index.d.ts.map +1 -1
  699. package/dist/runtime/query/iteration/index.js +577 -34
  700. package/dist/runtime/query/iteration/index.js.map +1 -1
  701. package/dist/runtime/query/iteration/phases/advisory.d.ts.map +1 -1
  702. package/dist/runtime/query/iteration/phases/advisory.js +14 -2
  703. package/dist/runtime/query/iteration/phases/advisory.js.map +1 -1
  704. package/dist/runtime/query/iteration/phases/advisory.test.d.ts +1 -1
  705. package/dist/runtime/query/iteration/phases/advisory.test.js +4 -1
  706. package/dist/runtime/query/iteration/phases/advisory.test.js.map +1 -1
  707. package/dist/runtime/query/iteration/phases/checkpoint.js +2 -2
  708. package/dist/runtime/query/iteration/phases/checkpoint.js.map +1 -1
  709. package/dist/runtime/query/iteration/phases/compaction-safe-cut.test.js +1 -1
  710. package/dist/runtime/query/iteration/phases/compaction-safe-cut.test.js.map +1 -1
  711. package/dist/runtime/query/iteration/phases/compaction.d.ts +22 -1
  712. package/dist/runtime/query/iteration/phases/compaction.d.ts.map +1 -1
  713. package/dist/runtime/query/iteration/phases/compaction.js +297 -64
  714. package/dist/runtime/query/iteration/phases/compaction.js.map +1 -1
  715. package/dist/runtime/query/iteration/phases/context-measurement.test.d.ts +21 -0
  716. package/dist/runtime/query/iteration/phases/context-measurement.test.d.ts.map +1 -0
  717. package/dist/runtime/query/iteration/phases/context-measurement.test.js +161 -0
  718. package/dist/runtime/query/iteration/phases/context-measurement.test.js.map +1 -0
  719. package/dist/runtime/query/iteration/phases/context.d.ts +62 -2
  720. package/dist/runtime/query/iteration/phases/context.d.ts.map +1 -1
  721. package/dist/runtime/query/iteration/phases/context.js +87 -0
  722. package/dist/runtime/query/iteration/phases/context.js.map +1 -1
  723. package/dist/runtime/query/iteration/phases/hitl-answer-question.test.js +3 -1
  724. package/dist/runtime/query/iteration/phases/hitl-answer-question.test.js.map +1 -1
  725. package/dist/runtime/query/iteration/phases/overflow-escalation.test.d.ts +15 -0
  726. package/dist/runtime/query/iteration/phases/overflow-escalation.test.d.ts.map +1 -0
  727. package/dist/runtime/query/iteration/phases/overflow-escalation.test.js +130 -0
  728. package/dist/runtime/query/iteration/phases/overflow-escalation.test.js.map +1 -0
  729. package/dist/runtime/query/iteration/phases/plan.d.ts.map +1 -1
  730. package/dist/runtime/query/iteration/phases/plan.js +8 -2
  731. package/dist/runtime/query/iteration/phases/plan.js.map +1 -1
  732. package/dist/runtime/query/iteration/phases/retained-messages.test.d.ts +2 -0
  733. package/dist/runtime/query/iteration/phases/retained-messages.test.d.ts.map +1 -0
  734. package/dist/runtime/query/iteration/phases/retained-messages.test.js +180 -0
  735. package/dist/runtime/query/iteration/phases/retained-messages.test.js.map +1 -0
  736. package/dist/runtime/query/iteration/phases/tool-review-denials.test.d.ts +2 -0
  737. package/dist/runtime/query/iteration/phases/tool-review-denials.test.d.ts.map +1 -0
  738. package/dist/runtime/query/iteration/phases/tool-review-denials.test.js +224 -0
  739. package/dist/runtime/query/iteration/phases/tool-review-denials.test.js.map +1 -0
  740. package/dist/runtime/query/iteration/phases/tool-review.d.ts +29 -1
  741. package/dist/runtime/query/iteration/phases/tool-review.d.ts.map +1 -1
  742. package/dist/runtime/query/iteration/phases/tool-review.js +114 -44
  743. package/dist/runtime/query/iteration/phases/tool-review.js.map +1 -1
  744. package/dist/runtime/query/iteration/phases/working-memory-compaction.test.js +20 -9
  745. package/dist/runtime/query/iteration/phases/working-memory-compaction.test.js.map +1 -1
  746. package/dist/runtime/query/iteration/stream-turn.d.ts +2 -26
  747. package/dist/runtime/query/iteration/stream-turn.d.ts.map +1 -1
  748. package/dist/runtime/query/iteration/stream-turn.js +254 -16
  749. package/dist/runtime/query/iteration/stream-turn.js.map +1 -1
  750. package/dist/runtime/query/plugin-hooks.d.ts.map +1 -1
  751. package/dist/runtime/query/plugin-hooks.js +6 -2
  752. package/dist/runtime/query/plugin-hooks.js.map +1 -1
  753. package/dist/runtime/query/prompt.d.ts.map +1 -1
  754. package/dist/runtime/query/prompt.js +25 -1
  755. package/dist/runtime/query/prompt.js.map +1 -1
  756. package/dist/runtime/query/question-park.d.ts +85 -0
  757. package/dist/runtime/query/question-park.d.ts.map +1 -0
  758. package/dist/runtime/query/question-park.js +71 -0
  759. package/dist/runtime/query/question-park.js.map +1 -0
  760. package/dist/runtime/query/result.d.ts +9 -0
  761. package/dist/runtime/query/result.d.ts.map +1 -1
  762. package/dist/runtime/query/result.js +49 -0
  763. package/dist/runtime/query/result.js.map +1 -1
  764. package/dist/runtime/query/resume-pending.d.ts +106 -0
  765. package/dist/runtime/query/resume-pending.d.ts.map +1 -0
  766. package/dist/runtime/query/resume-pending.js +312 -0
  767. package/dist/runtime/query/resume-pending.js.map +1 -0
  768. package/dist/runtime/query/run-state.d.ts +52 -0
  769. package/dist/runtime/query/run-state.d.ts.map +1 -0
  770. package/dist/runtime/query/run-state.js +87 -0
  771. package/dist/runtime/query/run-state.js.map +1 -0
  772. package/dist/runtime/query/tool-grants.d.ts +44 -0
  773. package/dist/runtime/query/tool-grants.d.ts.map +1 -0
  774. package/dist/runtime/query/tool-grants.js +63 -0
  775. package/dist/runtime/query/tool-grants.js.map +1 -0
  776. package/dist/runtime/query/tool-output-budget.d.ts +83 -0
  777. package/dist/runtime/query/tool-output-budget.d.ts.map +1 -0
  778. package/dist/runtime/query/tool-output-budget.js +147 -0
  779. package/dist/runtime/query/tool-output-budget.js.map +1 -0
  780. package/dist/runtime/query/tool-pause.d.ts +41 -0
  781. package/dist/runtime/query/tool-pause.d.ts.map +1 -0
  782. package/dist/runtime/query/tool-pause.js +94 -0
  783. package/dist/runtime/query/tool-pause.js.map +1 -0
  784. package/dist/runtime/query/tooling.d.ts +10 -1
  785. package/dist/runtime/query/tooling.d.ts.map +1 -1
  786. package/dist/runtime/query/tooling.js +13 -0
  787. package/dist/runtime/query/tooling.js.map +1 -1
  788. package/dist/sandbox/__tests__/isolation.test.d.ts +2 -0
  789. package/dist/sandbox/__tests__/isolation.test.d.ts.map +1 -0
  790. package/dist/sandbox/__tests__/isolation.test.js +139 -0
  791. package/dist/sandbox/__tests__/isolation.test.js.map +1 -0
  792. package/dist/sandbox/__tests__/output-truncation.test.d.ts +2 -0
  793. package/dist/sandbox/__tests__/output-truncation.test.d.ts.map +1 -0
  794. package/dist/sandbox/__tests__/output-truncation.test.js +88 -0
  795. package/dist/sandbox/__tests__/output-truncation.test.js.map +1 -0
  796. package/dist/sandbox/__tests__/resource-limits.test.d.ts +2 -0
  797. package/dist/sandbox/__tests__/resource-limits.test.d.ts.map +1 -0
  798. package/dist/sandbox/__tests__/resource-limits.test.js +86 -0
  799. package/dist/sandbox/__tests__/resource-limits.test.js.map +1 -0
  800. package/dist/sandbox/factory.js +1 -1
  801. package/dist/sandbox/factory.js.map +1 -1
  802. package/dist/sandbox/index.d.ts +2 -0
  803. package/dist/sandbox/index.d.ts.map +1 -1
  804. package/dist/sandbox/index.js +1 -0
  805. package/dist/sandbox/index.js.map +1 -1
  806. package/dist/sandbox/isolation.d.ts +15 -0
  807. package/dist/sandbox/isolation.d.ts.map +1 -0
  808. package/dist/sandbox/isolation.js +53 -0
  809. package/dist/sandbox/isolation.js.map +1 -0
  810. package/dist/sandbox/provider/local.d.ts +51 -2
  811. package/dist/sandbox/provider/local.d.ts.map +1 -1
  812. package/dist/sandbox/provider/local.js +139 -69
  813. package/dist/sandbox/provider/local.js.map +1 -1
  814. package/dist/session/__tests__/integration/spawn-rollback.test.d.ts +1 -1
  815. package/dist/session/__tests__/integration/spawn-rollback.test.js +1 -1
  816. package/dist/session/handoff/broadcast.js +3 -3
  817. package/dist/session/handoff/broadcast.js.map +1 -1
  818. package/dist/session/migration/__tests__/filesystem.test.js +5 -1
  819. package/dist/session/migration/__tests__/filesystem.test.js.map +1 -1
  820. package/dist/session/migration/filesystem.d.ts.map +1 -1
  821. package/dist/session/migration/filesystem.js +7 -13
  822. package/dist/session/migration/filesystem.js.map +1 -1
  823. package/dist/session/migration/marker.d.ts.map +1 -1
  824. package/dist/session/migration/marker.js +5 -10
  825. package/dist/session/migration/marker.js.map +1 -1
  826. package/dist/session/retention/disk-backend.d.ts.map +1 -1
  827. package/dist/session/retention/disk-backend.js +5 -12
  828. package/dist/session/retention/disk-backend.js.map +1 -1
  829. package/dist/session/workspace/__tests__/git-worktree.test.js +6 -2
  830. package/dist/session/workspace/__tests__/git-worktree.test.js.map +1 -1
  831. package/dist/session/workspace/__tests__/shared-run.test.js +15 -0
  832. package/dist/session/workspace/__tests__/shared-run.test.js.map +1 -1
  833. package/dist/session/workspace/shared-run.js +8 -1
  834. package/dist/session/workspace/shared-run.js.map +1 -1
  835. package/dist/skills/__tests__/frontmatter.test.d.ts +2 -0
  836. package/dist/skills/__tests__/frontmatter.test.d.ts.map +1 -0
  837. package/dist/skills/__tests__/frontmatter.test.js +106 -0
  838. package/dist/skills/__tests__/frontmatter.test.js.map +1 -0
  839. package/dist/skills/loader.d.ts.map +1 -1
  840. package/dist/skills/loader.js +42 -3
  841. package/dist/skills/loader.js.map +1 -1
  842. package/dist/store/__tests__/checkpoint-integrity.test.d.ts +2 -0
  843. package/dist/store/__tests__/checkpoint-integrity.test.d.ts.map +1 -0
  844. package/dist/store/__tests__/checkpoint-integrity.test.js +112 -0
  845. package/dist/store/__tests__/checkpoint-integrity.test.js.map +1 -0
  846. package/dist/store/__tests__/checkpoint-versioning.test.d.ts +2 -0
  847. package/dist/store/__tests__/checkpoint-versioning.test.d.ts.map +1 -0
  848. package/dist/store/__tests__/checkpoint-versioning.test.js +151 -0
  849. package/dist/store/__tests__/checkpoint-versioning.test.js.map +1 -0
  850. package/dist/store/__tests__/schema-adoption.test.d.ts +2 -0
  851. package/dist/store/__tests__/schema-adoption.test.d.ts.map +1 -0
  852. package/dist/store/__tests__/schema-adoption.test.js +90 -0
  853. package/dist/store/__tests__/schema-adoption.test.js.map +1 -0
  854. package/dist/store/__tests__/schema.test.d.ts +2 -0
  855. package/dist/store/__tests__/schema.test.d.ts.map +1 -0
  856. package/dist/store/__tests__/schema.test.js +93 -0
  857. package/dist/store/__tests__/schema.test.js.map +1 -0
  858. package/dist/store/memory/disk.d.ts.map +1 -1
  859. package/dist/store/memory/disk.js +16 -16
  860. package/dist/store/memory/disk.js.map +1 -1
  861. package/dist/store/run/disk.d.ts +23 -0
  862. package/dist/store/run/disk.d.ts.map +1 -1
  863. package/dist/store/run/disk.js +148 -23
  864. package/dist/store/run/disk.js.map +1 -1
  865. package/dist/store/schema.d.ts +73 -0
  866. package/dist/store/schema.d.ts.map +1 -0
  867. package/dist/store/schema.js +103 -0
  868. package/dist/store/schema.js.map +1 -0
  869. package/dist/store/session/disk.d.ts.map +1 -1
  870. package/dist/store/session/disk.js +25 -13
  871. package/dist/store/session/disk.js.map +1 -1
  872. package/dist/store/task/__tests__/lookup-across-runs.test.d.ts +2 -0
  873. package/dist/store/task/__tests__/lookup-across-runs.test.d.ts.map +1 -0
  874. package/dist/store/task/__tests__/lookup-across-runs.test.js +60 -0
  875. package/dist/store/task/__tests__/lookup-across-runs.test.js.map +1 -0
  876. package/dist/store/task/disk.d.ts +18 -0
  877. package/dist/store/task/disk.d.ts.map +1 -1
  878. package/dist/store/task/disk.js +61 -16
  879. package/dist/store/task/disk.js.map +1 -1
  880. package/dist/store/thread/disk.d.ts.map +1 -1
  881. package/dist/store/thread/disk.js +20 -11
  882. package/dist/store/thread/disk.js.map +1 -1
  883. package/dist/streaming/coalesce.d.ts +1 -1
  884. package/dist/streaming/coalesce.js +1 -1
  885. package/dist/telemetry/__tests__/latency-instruments.test.d.ts +2 -0
  886. package/dist/telemetry/__tests__/latency-instruments.test.d.ts.map +1 -0
  887. package/dist/telemetry/__tests__/latency-instruments.test.js +82 -0
  888. package/dist/telemetry/__tests__/latency-instruments.test.js.map +1 -0
  889. package/dist/telemetry/__tests__/metrics.test.d.ts +2 -0
  890. package/dist/telemetry/__tests__/metrics.test.d.ts.map +1 -0
  891. package/dist/telemetry/__tests__/metrics.test.js +112 -0
  892. package/dist/telemetry/__tests__/metrics.test.js.map +1 -0
  893. package/dist/telemetry/__tests__/span-hierarchy.test.d.ts +2 -0
  894. package/dist/telemetry/__tests__/span-hierarchy.test.d.ts.map +1 -0
  895. package/dist/telemetry/__tests__/span-hierarchy.test.js +84 -0
  896. package/dist/telemetry/__tests__/span-hierarchy.test.js.map +1 -0
  897. package/dist/telemetry/attributes.d.ts +44 -0
  898. package/dist/telemetry/attributes.d.ts.map +1 -1
  899. package/dist/telemetry/attributes.js +80 -0
  900. package/dist/telemetry/attributes.js.map +1 -1
  901. package/dist/telemetry/metrics.d.ts +41 -0
  902. package/dist/telemetry/metrics.d.ts.map +1 -0
  903. package/dist/telemetry/metrics.js +160 -0
  904. package/dist/telemetry/metrics.js.map +1 -0
  905. package/dist/test-setup.d.ts +3 -3
  906. package/dist/test-setup.js +3 -3
  907. package/dist/test-support/paths.d.ts +14 -0
  908. package/dist/test-support/paths.d.ts.map +1 -0
  909. package/dist/test-support/paths.js +16 -0
  910. package/dist/test-support/paths.js.map +1 -0
  911. package/dist/tools/__tests__/path-containment.test.d.ts +2 -0
  912. package/dist/tools/__tests__/path-containment.test.d.ts.map +1 -0
  913. package/dist/tools/__tests__/path-containment.test.js +133 -0
  914. package/dist/tools/__tests__/path-containment.test.js.map +1 -0
  915. package/dist/tools/__tests__/posix-path.test.d.ts +2 -0
  916. package/dist/tools/__tests__/posix-path.test.d.ts.map +1 -0
  917. package/dist/tools/__tests__/posix-path.test.js +67 -0
  918. package/dist/tools/__tests__/posix-path.test.js.map +1 -0
  919. package/dist/tools/__tests__/sandboxed-search.test.d.ts +2 -0
  920. package/dist/tools/__tests__/sandboxed-search.test.d.ts.map +1 -0
  921. package/dist/tools/__tests__/sandboxed-search.test.js +114 -0
  922. package/dist/tools/__tests__/sandboxed-search.test.js.map +1 -0
  923. package/dist/tools/builtins/__tests__/computer-use.test.js +12 -2
  924. package/dist/tools/builtins/__tests__/computer-use.test.js.map +1 -1
  925. package/dist/tools/builtins/__tests__/edit.test.js +10 -216
  926. package/dist/tools/builtins/__tests__/edit.test.js.map +1 -1
  927. package/dist/tools/builtins/__tests__/payload-budget.test.js +16 -30
  928. package/dist/tools/builtins/__tests__/payload-budget.test.js.map +1 -1
  929. package/dist/tools/builtins/__tests__/read-file.test.js +23 -1
  930. package/dist/tools/builtins/__tests__/read-file.test.js.map +1 -1
  931. package/dist/tools/builtins/__tests__/write-file.test.js +13 -37
  932. package/dist/tools/builtins/__tests__/write-file.test.js.map +1 -1
  933. package/dist/tools/builtins/bash.d.ts.map +1 -1
  934. package/dist/tools/builtins/bash.js +33 -4
  935. package/dist/tools/builtins/bash.js.map +1 -1
  936. package/dist/tools/builtins/computer-use.d.ts +22 -22
  937. package/dist/tools/builtins/computer-use.d.ts.map +1 -1
  938. package/dist/tools/builtins/computer-use.js +7 -1
  939. package/dist/tools/builtins/computer-use.js.map +1 -1
  940. package/dist/tools/builtins/edit.d.ts +5 -2
  941. package/dist/tools/builtins/edit.d.ts.map +1 -1
  942. package/dist/tools/builtins/edit.js +125 -105
  943. package/dist/tools/builtins/edit.js.map +1 -1
  944. package/dist/tools/builtins/glob.d.ts +1 -1
  945. package/dist/tools/builtins/glob.d.ts.map +1 -1
  946. package/dist/tools/builtins/glob.js +63 -6
  947. package/dist/tools/builtins/glob.js.map +1 -1
  948. package/dist/tools/builtins/grep.d.ts.map +1 -1
  949. package/dist/tools/builtins/grep.js +61 -8
  950. package/dist/tools/builtins/grep.js.map +1 -1
  951. package/dist/tools/builtins/index.js +6 -6
  952. package/dist/tools/builtins/index.js.map +1 -1
  953. package/dist/tools/builtins/ls.d.ts.map +1 -1
  954. package/dist/tools/builtins/ls.js +4 -2
  955. package/dist/tools/builtins/ls.js.map +1 -1
  956. package/dist/tools/builtins/read-file.d.ts +1 -1
  957. package/dist/tools/builtins/read-file.d.ts.map +1 -1
  958. package/dist/tools/builtins/read-file.js +38 -4
  959. package/dist/tools/builtins/read-file.js.map +1 -1
  960. package/dist/tools/builtins/write-file.d.ts +2 -1
  961. package/dist/tools/builtins/write-file.d.ts.map +1 -1
  962. package/dist/tools/builtins/write-file.js +40 -66
  963. package/dist/tools/builtins/write-file.js.map +1 -1
  964. package/dist/tools/coordinator/__tests__/agent.test.d.ts +1 -1
  965. package/dist/tools/coordinator/__tests__/agent.test.js +2 -2
  966. package/dist/tools/coordinator/__tests__/agent.test.js.map +1 -1
  967. package/dist/tools/coordinator/agent.d.ts +20 -9
  968. package/dist/tools/coordinator/agent.d.ts.map +1 -1
  969. package/dist/tools/coordinator/agent.js +2 -1
  970. package/dist/tools/coordinator/agent.js.map +1 -1
  971. package/dist/tools/coordinator/index.d.ts +20 -0
  972. package/dist/tools/coordinator/index.d.ts.map +1 -1
  973. package/dist/tools/coordinator/index.js +80 -32
  974. package/dist/tools/coordinator/index.js.map +1 -1
  975. package/dist/tools/defineTool.d.ts +16 -0
  976. package/dist/tools/defineTool.d.ts.map +1 -1
  977. package/dist/tools/defineTool.js +4 -0
  978. package/dist/tools/defineTool.js.map +1 -1
  979. package/dist/tools/glob-match.d.ts +18 -0
  980. package/dist/tools/glob-match.d.ts.map +1 -0
  981. package/dist/tools/glob-match.js +50 -0
  982. package/dist/tools/glob-match.js.map +1 -0
  983. package/dist/tools/paths.d.ts +20 -0
  984. package/dist/tools/paths.d.ts.map +1 -0
  985. package/dist/tools/paths.js +41 -0
  986. package/dist/tools/paths.js.map +1 -0
  987. package/dist/tools/posix-path.d.ts +24 -0
  988. package/dist/tools/posix-path.d.ts.map +1 -0
  989. package/dist/tools/posix-path.js +67 -0
  990. package/dist/tools/posix-path.js.map +1 -0
  991. package/dist/types/advisory/config.d.ts +21 -3
  992. package/dist/types/advisory/config.d.ts.map +1 -1
  993. package/dist/types/advisory/result.d.ts +12 -10
  994. package/dist/types/advisory/result.d.ts.map +1 -1
  995. package/dist/types/agent/base.d.ts +20 -2
  996. package/dist/types/agent/base.d.ts.map +1 -1
  997. package/dist/types/agent/factory.d.ts +15 -9
  998. package/dist/types/agent/factory.d.ts.map +1 -1
  999. package/dist/types/agent/gateway.d.ts +18 -0
  1000. package/dist/types/agent/gateway.d.ts.map +1 -1
  1001. package/dist/types/agent/manager.d.ts +19 -0
  1002. package/dist/types/agent/manager.d.ts.map +1 -1
  1003. package/dist/types/agent/reactive.d.ts +48 -0
  1004. package/dist/types/agent/reactive.d.ts.map +1 -1
  1005. package/dist/types/agent/router.d.ts +11 -0
  1006. package/dist/types/agent/router.d.ts.map +1 -1
  1007. package/dist/types/agent/supervisor.d.ts +2 -2
  1008. package/dist/types/agent/task.d.ts +24 -0
  1009. package/dist/types/agent/task.d.ts.map +1 -1
  1010. package/dist/types/bidi/index.d.ts +140 -0
  1011. package/dist/types/bidi/index.d.ts.map +1 -0
  1012. package/dist/types/bidi/index.js +2 -0
  1013. package/dist/types/bidi/index.js.map +1 -0
  1014. package/dist/types/common/__tests__/numeric-correctness.test.d.ts +2 -0
  1015. package/dist/types/common/__tests__/numeric-correctness.test.d.ts.map +1 -0
  1016. package/dist/types/common/__tests__/numeric-correctness.test.js +108 -0
  1017. package/dist/types/common/__tests__/numeric-correctness.test.js.map +1 -0
  1018. package/dist/types/common/index.d.ts +1 -1
  1019. package/dist/types/common/index.d.ts.map +1 -1
  1020. package/dist/types/common/index.js +15 -4
  1021. package/dist/types/common/index.js.map +1 -1
  1022. package/dist/types/connector/mcp.d.ts +29 -0
  1023. package/dist/types/connector/mcp.d.ts.map +1 -1
  1024. package/dist/types/errors/__tests__/catalog.test.d.ts +2 -0
  1025. package/dist/types/errors/__tests__/catalog.test.d.ts.map +1 -0
  1026. package/dist/types/errors/__tests__/catalog.test.js +106 -0
  1027. package/dist/types/errors/__tests__/catalog.test.js.map +1 -0
  1028. package/dist/types/errors/__tests__/errors.test.d.ts +2 -0
  1029. package/dist/types/errors/__tests__/errors.test.d.ts.map +1 -0
  1030. package/dist/types/errors/__tests__/errors.test.js +104 -0
  1031. package/dist/types/errors/__tests__/errors.test.js.map +1 -0
  1032. package/dist/types/errors/catalog.d.ts +83 -0
  1033. package/dist/types/errors/catalog.d.ts.map +1 -0
  1034. package/dist/types/errors/catalog.js +148 -0
  1035. package/dist/types/errors/catalog.js.map +1 -0
  1036. package/dist/types/errors/index.d.ts +73 -0
  1037. package/dist/types/errors/index.d.ts.map +1 -0
  1038. package/dist/types/errors/index.js +70 -0
  1039. package/dist/types/errors/index.js.map +1 -0
  1040. package/dist/types/guardrail/index.d.ts +57 -0
  1041. package/dist/types/guardrail/index.d.ts.map +1 -0
  1042. package/dist/types/guardrail/index.js +2 -0
  1043. package/dist/types/guardrail/index.js.map +1 -0
  1044. package/dist/types/hitl/index.d.ts +97 -14
  1045. package/dist/types/hitl/index.d.ts.map +1 -1
  1046. package/dist/types/hitl/index.js.map +1 -1
  1047. package/dist/types/ids/index.d.ts +2 -2
  1048. package/dist/types/message/content.d.ts +22 -0
  1049. package/dist/types/message/content.d.ts.map +1 -0
  1050. package/dist/types/message/content.js +49 -0
  1051. package/dist/types/message/content.js.map +1 -0
  1052. package/dist/types/message/index.d.ts +193 -8
  1053. package/dist/types/message/index.d.ts.map +1 -1
  1054. package/dist/types/message/index.js +19 -4
  1055. package/dist/types/message/index.js.map +1 -1
  1056. package/dist/types/plugin/index.d.ts +78 -3
  1057. package/dist/types/plugin/index.d.ts.map +1 -1
  1058. package/dist/types/plugin/index.js +0 -1
  1059. package/dist/types/plugin/index.js.map +1 -1
  1060. package/dist/types/provider/__tests__/cause-chain.test.d.ts +2 -0
  1061. package/dist/types/provider/__tests__/cause-chain.test.d.ts.map +1 -0
  1062. package/dist/types/provider/__tests__/cause-chain.test.js +108 -0
  1063. package/dist/types/provider/__tests__/cause-chain.test.js.map +1 -0
  1064. package/dist/types/provider/__tests__/declared-retryable.test.d.ts +2 -0
  1065. package/dist/types/provider/__tests__/declared-retryable.test.d.ts.map +1 -0
  1066. package/dist/types/provider/__tests__/declared-retryable.test.js +70 -0
  1067. package/dist/types/provider/__tests__/declared-retryable.test.js.map +1 -0
  1068. package/dist/types/provider/__tests__/overflow-classification.test.d.ts +2 -0
  1069. package/dist/types/provider/__tests__/overflow-classification.test.d.ts.map +1 -0
  1070. package/dist/types/provider/__tests__/overflow-classification.test.js +80 -0
  1071. package/dist/types/provider/__tests__/overflow-classification.test.js.map +1 -0
  1072. package/dist/types/provider/chat.d.ts +27 -1
  1073. package/dist/types/provider/chat.d.ts.map +1 -1
  1074. package/dist/types/provider/config.d.ts +83 -2
  1075. package/dist/types/provider/config.d.ts.map +1 -1
  1076. package/dist/types/provider/errors.d.ts +102 -0
  1077. package/dist/types/provider/errors.d.ts.map +1 -0
  1078. package/dist/types/provider/errors.js +393 -0
  1079. package/dist/types/provider/errors.js.map +1 -0
  1080. package/dist/types/provider/index.d.ts +3 -2
  1081. package/dist/types/provider/index.d.ts.map +1 -1
  1082. package/dist/types/provider/stream.d.ts +59 -3
  1083. package/dist/types/provider/stream.d.ts.map +1 -1
  1084. package/dist/types/rag/embedding.d.ts +11 -2
  1085. package/dist/types/rag/embedding.d.ts.map +1 -1
  1086. package/dist/types/run/__tests__/step.test.d.ts +2 -0
  1087. package/dist/types/run/__tests__/step.test.d.ts.map +1 -0
  1088. package/dist/types/run/__tests__/step.test.js +107 -0
  1089. package/dist/types/run/__tests__/step.test.js.map +1 -0
  1090. package/dist/types/run/answer-review.d.ts +46 -0
  1091. package/dist/types/run/answer-review.d.ts.map +1 -0
  1092. package/dist/types/run/answer-review.js +2 -0
  1093. package/dist/types/run/answer-review.js.map +1 -0
  1094. package/dist/types/run/config.d.ts +33 -0
  1095. package/dist/types/run/config.d.ts.map +1 -1
  1096. package/dist/types/run/derive-status.d.ts +29 -0
  1097. package/dist/types/run/derive-status.d.ts.map +1 -0
  1098. package/dist/types/run/derive-status.js +49 -0
  1099. package/dist/types/run/derive-status.js.map +1 -0
  1100. package/dist/types/run/entity.d.ts +16 -0
  1101. package/dist/types/run/entity.d.ts.map +1 -1
  1102. package/dist/types/run/events.d.ts +228 -4
  1103. package/dist/types/run/events.d.ts.map +1 -1
  1104. package/dist/types/run/events.js +7 -0
  1105. package/dist/types/run/events.js.map +1 -1
  1106. package/dist/types/run/index.d.ts +5 -0
  1107. package/dist/types/run/index.d.ts.map +1 -1
  1108. package/dist/types/run/index.js +3 -0
  1109. package/dist/types/run/index.js.map +1 -1
  1110. package/dist/types/run/memory-promotion.d.ts +90 -0
  1111. package/dist/types/run/memory-promotion.d.ts.map +1 -0
  1112. package/dist/types/run/memory-promotion.js +38 -0
  1113. package/dist/types/run/memory-promotion.js.map +1 -0
  1114. package/dist/types/run/prepare-step.d.ts +100 -0
  1115. package/dist/types/run/prepare-step.d.ts.map +1 -0
  1116. package/dist/types/run/prepare-step.js +2 -0
  1117. package/dist/types/run/prepare-step.js.map +1 -0
  1118. package/dist/types/run/state.d.ts +85 -0
  1119. package/dist/types/run/state.d.ts.map +1 -0
  1120. package/dist/types/run/state.js +33 -0
  1121. package/dist/types/run/state.js.map +1 -0
  1122. package/dist/types/run/step.d.ts +77 -0
  1123. package/dist/types/run/step.d.ts.map +1 -0
  1124. package/dist/types/run/step.js +26 -0
  1125. package/dist/types/run/step.js.map +1 -0
  1126. package/dist/types/run/stop-reason.d.ts +31 -3
  1127. package/dist/types/run/stop-reason.d.ts.map +1 -1
  1128. package/dist/types/sandbox/index.d.ts +82 -10
  1129. package/dist/types/sandbox/index.d.ts.map +1 -1
  1130. package/dist/types/sandbox/index.js +13 -1
  1131. package/dist/types/sandbox/index.js.map +1 -1
  1132. package/dist/types/structured-output/index.d.ts +6 -5
  1133. package/dist/types/structured-output/index.d.ts.map +1 -1
  1134. package/dist/types/thread/entity.d.ts +2 -3
  1135. package/dist/types/thread/entity.d.ts.map +1 -1
  1136. package/dist/types/tool/index.d.ts +193 -1
  1137. package/dist/types/tool/index.d.ts.map +1 -1
  1138. package/dist/types/tool/index.js +1 -1
  1139. package/dist/types/tool/index.js.map +1 -1
  1140. package/dist/types/tool/repair.d.ts +49 -0
  1141. package/dist/types/tool/repair.d.ts.map +1 -0
  1142. package/dist/types/tool/repair.js +2 -0
  1143. package/dist/types/tool/repair.js.map +1 -0
  1144. package/dist/types/toolset/index.d.ts +2 -5
  1145. package/dist/types/toolset/index.d.ts.map +1 -1
  1146. package/dist/utils/__tests__/atomic-write.test.d.ts +2 -0
  1147. package/dist/utils/__tests__/atomic-write.test.d.ts.map +1 -0
  1148. package/dist/utils/__tests__/atomic-write.test.js +91 -0
  1149. package/dist/utils/__tests__/atomic-write.test.js.map +1 -0
  1150. package/dist/utils/__tests__/shell-compress.test.d.ts +2 -0
  1151. package/dist/utils/__tests__/shell-compress.test.d.ts.map +1 -0
  1152. package/dist/utils/__tests__/shell-compress.test.js +41 -0
  1153. package/dist/utils/__tests__/shell-compress.test.js.map +1 -0
  1154. package/dist/utils/atomic-write.d.ts +10 -0
  1155. package/dist/utils/atomic-write.d.ts.map +1 -0
  1156. package/dist/utils/atomic-write.js +83 -0
  1157. package/dist/utils/atomic-write.js.map +1 -0
  1158. package/dist/utils/shell-compress.d.ts.map +1 -1
  1159. package/dist/utils/shell-compress.js +12 -1
  1160. package/dist/utils/shell-compress.js.map +1 -1
  1161. package/dist/verification/presets.d.ts +6 -3
  1162. package/dist/verification/presets.d.ts.map +1 -1
  1163. package/dist/verification/presets.js +6 -3
  1164. package/dist/verification/presets.js.map +1 -1
  1165. package/package.json +3 -3
  1166. package/src/__tests__/default-path-conformance.test.ts +306 -0
  1167. package/src/__tests__/shipped-defaults.test.ts +111 -0
  1168. package/src/advisory/__tests__/budget-enforcement.test.ts +208 -0
  1169. package/src/advisory/__tests__/structured-response.test.ts +128 -0
  1170. package/src/advisory/budget.ts +27 -0
  1171. package/src/advisory/context.ts +19 -0
  1172. package/src/advisory/executor.test.ts +12 -8
  1173. package/src/advisory/executor.ts +37 -25
  1174. package/src/advisory/index.ts +2 -0
  1175. package/src/advisory/parse.ts +63 -0
  1176. package/src/agents/AbstractAgent.ts +82 -0
  1177. package/src/agents/PipelineAgent.ts +19 -0
  1178. package/src/agents/ReactiveAgent.ts +47 -0
  1179. package/src/agents/RouterAgent.ts +32 -2
  1180. package/src/agents/SupervisorAgent.ts +31 -1
  1181. package/src/agents/__tests__/idempotency-key.test.ts +177 -0
  1182. package/src/agents/__tests__/invocation-lock.test.ts +154 -0
  1183. package/src/agents/__tests__/reachability.test.ts +224 -0
  1184. package/src/agents/__tests__/supervisor-ledger.test.ts +1 -1
  1185. package/src/agents/index.ts +1 -1
  1186. package/src/agents/lock.ts +0 -2
  1187. package/src/bridge/a2a/late-events.test.ts +84 -0
  1188. package/src/bridge/a2a/mapper.test.ts +1 -1
  1189. package/src/bridge/a2a/mapper.ts +70 -5
  1190. package/src/bridge/a2a/message.ts +5 -1
  1191. package/src/bridge/sse/late-events.test.ts +182 -0
  1192. package/src/bridge/sse/mapper.ts +109 -0
  1193. package/src/compaction/__tests__/cleared-result-recovery.test.ts +118 -0
  1194. package/src/compaction/__tests__/context-window.test.ts +77 -0
  1195. package/src/compaction/__tests__/dangling.test.ts +16 -6
  1196. package/src/compaction/__tests__/lossiness.test.ts +187 -0
  1197. package/src/compaction/__tests__/state-across-resume.test.ts +99 -0
  1198. package/src/compaction/__tests__/tool-result-editing.test.ts +213 -0
  1199. package/src/compaction/__tests__/trim-opens-on-user.test.ts +118 -0
  1200. package/src/compaction/__tests__/verifier-model.test.ts +83 -0
  1201. package/src/compaction/context-window.ts +107 -0
  1202. package/src/compaction/dangling.ts +48 -1
  1203. package/src/compaction/extractor.ts +56 -1
  1204. package/src/compaction/index.ts +16 -0
  1205. package/src/compaction/manager.ts +53 -8
  1206. package/src/compaction/managers/structured.ts +7 -6
  1207. package/src/compaction/retention.ts +89 -0
  1208. package/src/compaction/serializer.ts +42 -11
  1209. package/src/compaction/tool-result-editing.ts +229 -0
  1210. package/src/compaction/types.ts +10 -0
  1211. package/src/compaction/verifier.ts +24 -1
  1212. package/src/compaction/wire.ts +91 -0
  1213. package/src/config/runtime.ts +63 -19
  1214. package/src/connector/builtins/http.test.ts +13 -0
  1215. package/src/connector/builtins/http.ts +15 -1
  1216. package/src/connector/mcp/__tests__/discovery-policy.test.ts +133 -0
  1217. package/src/connector/mcp/__tests__/list-pagination.test.ts +98 -0
  1218. package/src/connector/mcp/__tests__/policy.test.ts +100 -0
  1219. package/src/connector/mcp/__tests__/protocol-negotiation.test.ts +105 -0
  1220. package/src/connector/mcp/__tests__/schema-fidelity.test.ts +250 -0
  1221. package/src/connector/mcp/__tests__/structured-result.test.ts +143 -0
  1222. package/src/connector/mcp/adapter.ts +306 -19
  1223. package/src/connector/mcp/client.test.ts +159 -0
  1224. package/src/connector/mcp/client.ts +160 -13
  1225. package/src/connector/mcp/discovery.ts +74 -5
  1226. package/src/connector/mcp/http-sse.ts +16 -1
  1227. package/src/connector/mcp/index.ts +4 -0
  1228. package/src/connector/mcp/policy.ts +134 -0
  1229. package/src/connector/mcp/schema-refs.ts +126 -0
  1230. package/src/constants/agent/index.ts +3 -0
  1231. package/src/constants/compaction/index.ts +1 -1
  1232. package/src/constants/continuation.ts +4 -4
  1233. package/src/constants/mcp/index.ts +34 -0
  1234. package/src/constants/plugin/index.ts +24 -2
  1235. package/src/constants/sandbox/index.ts +6 -5
  1236. package/src/constants/telemetry/index.ts +8 -0
  1237. package/src/constants/tools/index.ts +20 -0
  1238. package/src/contracts/api.ts +40 -0
  1239. package/src/contracts/run-status.ts +43 -0
  1240. package/src/eval/__tests__/case-deadline.test.ts +147 -0
  1241. package/src/eval/__tests__/end-to-end.test.ts +259 -0
  1242. package/src/eval/__tests__/harness-honesty.test.ts +99 -0
  1243. package/src/eval/__tests__/harness.test.ts +249 -0
  1244. package/src/eval/__tests__/judge.test.ts +306 -0
  1245. package/src/eval/__tests__/scorer-gates.test.ts +156 -0
  1246. package/src/eval/experiment.ts +335 -0
  1247. package/src/eval/from-run.ts +48 -0
  1248. package/src/eval/index.ts +21 -0
  1249. package/src/eval/judge.ts +240 -0
  1250. package/src/eval/scorers.ts +169 -0
  1251. package/src/eval/types.ts +157 -0
  1252. package/src/execution/local.test.ts +46 -0
  1253. package/src/execution/local.ts +6 -1
  1254. package/src/gateway/__tests__/list-survives-eviction.test.ts +124 -0
  1255. package/src/gateway/local.ts +117 -3
  1256. package/src/manager/agent/__tests__/dispose-cancels.test.ts +104 -0
  1257. package/src/manager/agent/__tests__/lifecycle.test.ts +297 -0
  1258. package/src/manager/agent/lifecycle.ts +127 -7
  1259. package/src/manager/run/persistence.ts +132 -1
  1260. package/src/plugin/__tests__/enable-contributions.test.ts +3 -3
  1261. package/src/plugin/__tests__/hook-ordering.test.ts +252 -0
  1262. package/src/plugin/__tests__/install-time-refusal.test.ts +68 -0
  1263. package/src/plugin/__tests__/lifecycle.test.ts +95 -73
  1264. package/src/plugin/lifecycle.ts +76 -22
  1265. package/src/plugin/loader.ts +37 -1
  1266. package/src/probe/registry.test.ts +49 -6
  1267. package/src/probe/registry.ts +21 -1
  1268. package/src/provider/__tests__/mock-script.test.ts +179 -0
  1269. package/src/provider/__tests__/registry.test.ts +1 -1
  1270. package/src/provider/__tests__/retry-visibility.test.ts +160 -0
  1271. package/src/provider/__tests__/retry.test.ts +296 -0
  1272. package/src/provider/capabilities.ts +4 -0
  1273. package/src/provider/collect.test.ts +1 -1
  1274. package/src/provider/collect.ts +1 -1
  1275. package/src/provider/index.ts +8 -0
  1276. package/src/provider/mock-register.ts +5 -2
  1277. package/src/provider/mock.ts +117 -34
  1278. package/src/provider/retry.ts +199 -0
  1279. package/src/public-runtime.ts +111 -7
  1280. package/src/public-types.ts +5 -1
  1281. package/src/rag/__tests__/retrieval-math.test.ts +54 -0
  1282. package/src/rag/chunking.ts +32 -1
  1283. package/src/rag/embedding.test.ts +65 -10
  1284. package/src/rag/embedding.ts +26 -5
  1285. package/src/rag/index.ts +1 -1
  1286. package/src/rag/retriever.ts +92 -21
  1287. package/src/rag/vector-store.test.ts +2 -2
  1288. package/src/registry/ManagedRegistry.test.ts +1 -1
  1289. package/src/registry/Registry.test.ts +1 -1
  1290. package/src/registry/connector/definitions.test.ts +1 -1
  1291. package/src/registry/tool/__tests__/schema.test.ts +208 -0
  1292. package/src/registry/tool/__tests__/tool-name-validation.test.ts +88 -0
  1293. package/src/registry/tool/execute.ts +104 -22
  1294. package/src/registry/tool/schema.ts +80 -0
  1295. package/src/registry/toolset/catalog.test.ts +0 -1
  1296. package/src/registry/toolset/catalog.ts +0 -1
  1297. package/src/run/reporter.ts +84 -0
  1298. package/src/runtime/bidi/__tests__/session.test.ts +232 -0
  1299. package/src/runtime/bidi/index.ts +4 -0
  1300. package/src/runtime/bidi/mock.ts +90 -0
  1301. package/src/runtime/bidi/session.ts +205 -0
  1302. package/src/runtime/decision/parser.test.ts +46 -0
  1303. package/src/runtime/decision/parser.ts +8 -1
  1304. package/src/runtime/query/__tests__/answer-review.test.ts +172 -0
  1305. package/src/runtime/query/__tests__/capability-negotiation.test.ts +25 -129
  1306. package/src/runtime/query/__tests__/checkpoint-store.test.ts +13 -50
  1307. package/src/runtime/query/__tests__/checkpoint.test.ts +2 -10
  1308. package/src/runtime/query/__tests__/citations.test.ts +111 -0
  1309. package/src/runtime/query/__tests__/content-budget.test.ts +83 -0
  1310. package/src/runtime/query/__tests__/context.test.ts +3 -2
  1311. package/src/runtime/query/__tests__/deferred-tools.test.ts +23 -148
  1312. package/src/runtime/query/__tests__/delegation-resume.test.ts +159 -0
  1313. package/src/runtime/query/__tests__/document-attachments.test.ts +125 -0
  1314. package/src/runtime/query/__tests__/durable-park-and-trace.test.ts +245 -0
  1315. package/src/runtime/query/__tests__/durable-question-park.test.ts +305 -0
  1316. package/src/runtime/query/__tests__/durable-run-state.test.ts +194 -0
  1317. package/src/runtime/query/__tests__/executor-plugin-hooks.test.ts +24 -20
  1318. package/src/runtime/query/__tests__/finish-reason-vs-tool-calls.test.ts +206 -0
  1319. package/src/runtime/query/__tests__/grant-cannot-overrule-the-gate.test.ts +115 -0
  1320. package/src/runtime/query/__tests__/guardrails-e2e.test.ts +164 -0
  1321. package/src/runtime/query/__tests__/guardrails.test.ts +236 -0
  1322. package/src/runtime/query/__tests__/hook-model-visibility.test.ts +143 -0
  1323. package/src/runtime/query/__tests__/long-document-flow.test.ts +9 -2
  1324. package/src/runtime/query/__tests__/memory-promotion.test.ts +148 -0
  1325. package/src/runtime/query/__tests__/orphan-task-turn-end.test.ts +5 -31
  1326. package/src/runtime/query/__tests__/overflow-relief.test.ts +139 -0
  1327. package/src/runtime/query/__tests__/partial-batch-resume.test.ts +264 -0
  1328. package/src/runtime/query/__tests__/plugin-hooks.test.ts +0 -1
  1329. package/src/runtime/query/__tests__/prepare-step-chain.test.ts +127 -0
  1330. package/src/runtime/query/__tests__/prepare-step.test.ts +233 -0
  1331. package/src/runtime/query/__tests__/probe-veto-is-error.test.ts +145 -0
  1332. package/src/runtime/query/__tests__/prompt.test.ts +1 -1
  1333. package/src/runtime/query/__tests__/provider-fault-pauses.test.ts +104 -0
  1334. package/src/runtime/query/__tests__/reasoning-stream.test.ts +177 -0
  1335. package/src/runtime/query/__tests__/recoverable-settle.test.ts +145 -0
  1336. package/src/runtime/query/__tests__/resume-budget.test.ts +158 -0
  1337. package/src/runtime/query/__tests__/resume-pending.test.ts +273 -0
  1338. package/src/runtime/query/__tests__/rich-content-budget-wiring.test.ts +109 -0
  1339. package/src/runtime/query/__tests__/stop-condition.test.ts +265 -0
  1340. package/src/runtime/query/__tests__/structured-output.test.ts +268 -0
  1341. package/src/runtime/query/__tests__/structured-run-failure.test.ts +144 -0
  1342. package/src/runtime/query/__tests__/terminal-tool.test.ts +148 -0
  1343. package/src/runtime/query/__tests__/tool-grants.test.ts +102 -0
  1344. package/src/runtime/query/__tests__/tool-output-budget.test.ts +0 -0
  1345. package/src/runtime/query/__tests__/tool-pause.test.ts +209 -0
  1346. package/src/runtime/query/__tests__/tool-progress.test.ts +154 -0
  1347. package/src/runtime/query/__tests__/tool-repair-retry.test.ts +255 -0
  1348. package/src/runtime/query/__tests__/tool-result-wire.test.ts +135 -0
  1349. package/src/runtime/query/__tests__/tool-timeout.test.ts +230 -0
  1350. package/src/runtime/query/__tests__/trace-parenting.test.ts +128 -0
  1351. package/src/runtime/query/__tests__/wiring-seams.test.ts +288 -0
  1352. package/src/runtime/query/checkpoint.ts +242 -10
  1353. package/src/runtime/query/events.ts +2 -2
  1354. package/src/runtime/query/executor.ts +847 -39
  1355. package/src/runtime/query/guard.ts +26 -1
  1356. package/src/runtime/query/guardrail-presets.ts +124 -0
  1357. package/src/runtime/query/guardrails.ts +144 -0
  1358. package/src/runtime/query/index.ts +759 -12
  1359. package/src/runtime/query/iteration/__tests__/cancel-settles-turn.test.ts +177 -0
  1360. package/src/runtime/query/iteration/index.ts +656 -40
  1361. package/src/runtime/query/iteration/phases/advisory.test.ts +4 -1
  1362. package/src/runtime/query/iteration/phases/advisory.ts +14 -2
  1363. package/src/runtime/query/iteration/phases/checkpoint.ts +2 -2
  1364. package/src/runtime/query/iteration/phases/compaction-safe-cut.test.ts +3 -1
  1365. package/src/runtime/query/iteration/phases/compaction.ts +329 -64
  1366. package/src/runtime/query/iteration/phases/context-measurement.test.ts +196 -0
  1367. package/src/runtime/query/iteration/phases/context.ts +155 -2
  1368. package/src/runtime/query/iteration/phases/hitl-answer-question.test.ts +3 -1
  1369. package/src/runtime/query/iteration/phases/overflow-escalation.test.ts +167 -0
  1370. package/src/runtime/query/iteration/phases/plan.ts +10 -2
  1371. package/src/runtime/query/iteration/phases/retained-messages.test.ts +226 -0
  1372. package/src/runtime/query/iteration/phases/tool-review-denials.test.ts +270 -0
  1373. package/src/runtime/query/iteration/phases/tool-review.ts +140 -51
  1374. package/src/runtime/query/iteration/phases/working-memory-compaction.test.ts +21 -9
  1375. package/src/runtime/query/iteration/stream-turn.ts +285 -15
  1376. package/src/runtime/query/plugin-hooks.ts +6 -2
  1377. package/src/runtime/query/prompt.ts +26 -1
  1378. package/src/runtime/query/question-park.ts +120 -0
  1379. package/src/runtime/query/result.ts +64 -0
  1380. package/src/runtime/query/resume-pending.ts +394 -0
  1381. package/src/runtime/query/run-state.ts +124 -0
  1382. package/src/runtime/query/tool-grants.ts +83 -0
  1383. package/src/runtime/query/tool-output-budget.ts +183 -0
  1384. package/src/runtime/query/tool-pause.ts +119 -0
  1385. package/src/runtime/query/tooling.ts +23 -1
  1386. package/src/sandbox/__tests__/isolation.test.ts +164 -0
  1387. package/src/sandbox/__tests__/output-truncation.test.ts +106 -0
  1388. package/src/sandbox/__tests__/resource-limits.test.ts +106 -0
  1389. package/src/sandbox/factory.ts +1 -1
  1390. package/src/sandbox/index.ts +2 -0
  1391. package/src/sandbox/isolation.ts +72 -0
  1392. package/src/sandbox/provider/local.ts +173 -74
  1393. package/src/session/__tests__/integration/spawn-rollback.test.ts +1 -1
  1394. package/src/session/handoff/broadcast.ts +3 -3
  1395. package/src/session/migration/__tests__/filesystem.test.ts +26 -19
  1396. package/src/session/migration/filesystem.ts +7 -12
  1397. package/src/session/migration/marker.ts +5 -9
  1398. package/src/session/retention/disk-backend.ts +5 -11
  1399. package/src/session/workspace/__tests__/git-worktree.test.ts +6 -2
  1400. package/src/session/workspace/__tests__/shared-run.test.ts +18 -0
  1401. package/src/session/workspace/shared-run.ts +7 -1
  1402. package/src/skills/__tests__/frontmatter.test.ts +135 -0
  1403. package/src/skills/loader.ts +47 -3
  1404. package/src/store/__tests__/checkpoint-integrity.test.ts +134 -0
  1405. package/src/store/__tests__/checkpoint-versioning.test.ts +194 -0
  1406. package/src/store/__tests__/schema-adoption.test.ts +118 -0
  1407. package/src/store/__tests__/schema.test.ts +119 -0
  1408. package/src/store/memory/disk.ts +17 -16
  1409. package/src/store/run/disk.ts +169 -22
  1410. package/src/store/schema.ts +135 -0
  1411. package/src/store/session/disk.ts +30 -21
  1412. package/src/store/task/__tests__/lookup-across-runs.test.ts +77 -0
  1413. package/src/store/task/disk.ts +59 -15
  1414. package/src/store/thread/disk.ts +21 -10
  1415. package/src/streaming/coalesce.ts +1 -1
  1416. package/src/telemetry/__tests__/latency-instruments.test.ts +120 -0
  1417. package/src/telemetry/__tests__/metrics.test.ts +160 -0
  1418. package/src/telemetry/__tests__/span-hierarchy.test.ts +94 -0
  1419. package/src/telemetry/attributes.ts +109 -0
  1420. package/src/telemetry/metrics.ts +190 -0
  1421. package/src/test-setup.ts +3 -3
  1422. package/src/test-support/paths.ts +17 -0
  1423. package/src/tools/__tests__/path-containment.test.ts +167 -0
  1424. package/src/tools/__tests__/posix-path.test.ts +83 -0
  1425. package/src/tools/__tests__/sandboxed-search.test.ts +140 -0
  1426. package/src/tools/builtins/__tests__/computer-use.test.ts +15 -2
  1427. package/src/tools/builtins/__tests__/edit.test.ts +10 -270
  1428. package/src/tools/builtins/__tests__/payload-budget.test.ts +20 -40
  1429. package/src/tools/builtins/__tests__/read-file.test.ts +29 -1
  1430. package/src/tools/builtins/__tests__/write-file.test.ts +13 -52
  1431. package/src/tools/builtins/bash.ts +34 -4
  1432. package/src/tools/builtins/computer-use.ts +7 -1
  1433. package/src/tools/builtins/edit.ts +162 -116
  1434. package/src/tools/builtins/glob.ts +71 -6
  1435. package/src/tools/builtins/grep.ts +82 -8
  1436. package/src/tools/builtins/index.ts +6 -6
  1437. package/src/tools/builtins/ls.ts +4 -2
  1438. package/src/tools/builtins/read-file.ts +40 -4
  1439. package/src/tools/builtins/write-file.ts +44 -68
  1440. package/src/tools/coordinator/__tests__/agent.test.ts +2 -2
  1441. package/src/tools/coordinator/agent.ts +23 -10
  1442. package/src/tools/coordinator/index.ts +111 -31
  1443. package/src/tools/defineTool.ts +20 -0
  1444. package/src/tools/glob-match.ts +50 -0
  1445. package/src/tools/paths.ts +44 -0
  1446. package/src/tools/posix-path.ts +73 -0
  1447. package/src/types/advisory/config.ts +21 -3
  1448. package/src/types/advisory/result.ts +12 -7
  1449. package/src/types/agent/base.ts +22 -2
  1450. package/src/types/agent/factory.ts +15 -9
  1451. package/src/types/agent/gateway.ts +20 -0
  1452. package/src/types/agent/manager.ts +19 -0
  1453. package/src/types/agent/reactive.ts +50 -0
  1454. package/src/types/agent/router.ts +11 -0
  1455. package/src/types/agent/supervisor.ts +2 -2
  1456. package/src/types/agent/task.ts +26 -0
  1457. package/src/types/bidi/index.ts +119 -0
  1458. package/src/types/common/__tests__/numeric-correctness.test.ts +123 -0
  1459. package/src/types/common/index.ts +16 -4
  1460. package/src/types/computer-use/index.ts +1 -1
  1461. package/src/types/connector/mcp.ts +29 -0
  1462. package/src/types/errors/__tests__/catalog.test.ts +130 -0
  1463. package/src/types/errors/__tests__/errors.test.ts +118 -0
  1464. package/src/types/errors/catalog.ts +211 -0
  1465. package/src/types/errors/index.ts +121 -0
  1466. package/src/types/guardrail/index.ts +62 -0
  1467. package/src/types/hitl/index.ts +102 -20
  1468. package/src/types/ids/index.ts +2 -2
  1469. package/src/types/message/content.ts +50 -0
  1470. package/src/types/message/index.ts +210 -9
  1471. package/src/types/plugin/index.ts +82 -2
  1472. package/src/types/provider/__tests__/cause-chain.test.ts +132 -0
  1473. package/src/types/provider/__tests__/declared-retryable.test.ts +85 -0
  1474. package/src/types/provider/__tests__/overflow-classification.test.ts +113 -0
  1475. package/src/types/provider/chat.ts +29 -1
  1476. package/src/types/provider/config.ts +83 -2
  1477. package/src/types/provider/errors.ts +472 -0
  1478. package/src/types/provider/index.ts +5 -0
  1479. package/src/types/provider/stream.ts +63 -3
  1480. package/src/types/rag/embedding.ts +11 -2
  1481. package/src/types/run/__tests__/step.test.ts +123 -0
  1482. package/src/types/run/answer-review.ts +51 -0
  1483. package/src/types/run/config.ts +35 -0
  1484. package/src/types/run/derive-status.ts +62 -0
  1485. package/src/types/run/entity.ts +18 -0
  1486. package/src/types/run/events.ts +230 -3
  1487. package/src/types/run/index.ts +5 -0
  1488. package/src/types/run/memory-promotion.ts +112 -0
  1489. package/src/types/run/prepare-step.ts +109 -0
  1490. package/src/types/run/state.ts +111 -0
  1491. package/src/types/run/step.ts +95 -0
  1492. package/src/types/run/stop-reason.ts +29 -1
  1493. package/src/types/sandbox/index.ts +93 -12
  1494. package/src/types/structured-output/index.ts +6 -5
  1495. package/src/types/thread/entity.ts +2 -3
  1496. package/src/types/tool/index.ts +204 -1
  1497. package/src/types/tool/repair.ts +54 -0
  1498. package/src/types/toolset/index.ts +2 -6
  1499. package/src/utils/__tests__/atomic-write.test.ts +108 -0
  1500. package/src/utils/__tests__/shell-compress.test.ts +49 -0
  1501. package/src/utils/atomic-write.ts +85 -0
  1502. package/src/utils/shell-compress.ts +12 -1
  1503. package/src/verification/presets.ts +6 -3
@@ -0,0 +1,63 @@
1
+ import type { LLMProvider } from '../types/provider/interface.js';
2
+ import type { Scorer } from './types.js';
3
+ /**
4
+ * Grade an open-ended answer with a model.
5
+ *
6
+ * Every other scorer here is a pure function over the run, which is what
7
+ * makes them cheap and reproducible — and also what makes them unable to
8
+ * say anything about whether an answer is *good*. `containsScorer` can
9
+ * check that a required phrase appears; it cannot tell a correct
10
+ * explanation from a fluent wrong one. That was the hole: the dimension
11
+ * most worth guarding was the one with no scorer behind it.
12
+ *
13
+ * A judge is a network call, so this scorer can fail in ways a pure
14
+ * function cannot. It throws rather than returning a low score, and the
15
+ * harness files a throw as `unavailable` — a rate limit must not read as
16
+ * a regression. See `Score.unavailable`.
17
+ */
18
+ export interface JudgeScorerConfig {
19
+ /** Defaults to `'judge'`. Give it a distinct name to run several. */
20
+ name?: string;
21
+ provider: LLMProvider;
22
+ model: string;
23
+ /**
24
+ * What "good" means, in the caller's words. REQUIRED.
25
+ *
26
+ * A judge asked to rate quality with no rubric rates fluency, which
27
+ * correlates with almost nothing worth measuring and drifts whenever
28
+ * the judge model changes. Making this optional would make the scorer
29
+ * easy to use and its output meaningless, so it is not optional.
30
+ */
31
+ rubric: string;
32
+ /**
33
+ * Highest grade on the scale. Default 4.
34
+ *
35
+ * An integer scale, not a 0..1 float: models place a continuous score
36
+ * poorly and cluster on round numbers, while a short ordinal scale
37
+ * against a written rubric is a judgement they can actually make. The
38
+ * result is divided down to 0..1 for the report.
39
+ *
40
+ * The default is EVEN on purpose — an odd scale has a midpoint, and a
41
+ * midpoint is where an uncertain judge parks. Forcing a side produces a
42
+ * signal; a pile of 3-out-of-5s does not.
43
+ */
44
+ scale?: number;
45
+ /**
46
+ * Show the judge which tools the run called. Default false.
47
+ *
48
+ * Useful when the rubric is about method rather than answer, and a
49
+ * needless cost otherwise — the trajectory is usually longer than the
50
+ * answer it produced.
51
+ */
52
+ includeTrajectory?: boolean;
53
+ /**
54
+ * Cap on the answer text handed to the judge. Default 20000.
55
+ *
56
+ * Truncation is disclosed IN the prompt. A judge shown a silently cut
57
+ * answer marks it down for stopping mid-sentence, which scores our
58
+ * truncation rather than the run.
59
+ */
60
+ maxOutputChars?: number;
61
+ }
62
+ export declare function judgeScorer(config: JudgeScorerConfig): Scorer;
63
+ //# sourceMappingURL=judge.d.ts.map
@@ -0,0 +1 @@
1
+ {"version":3,"file":"judge.d.ts","sourceRoot":"","sources":["../../src/eval/judge.ts"],"names":[],"mappings":"AACA,OAAO,KAAK,EAAE,WAAW,EAAE,MAAM,gCAAgC,CAAA;AACjE,OAAO,KAAK,EAA4B,MAAM,EAAE,MAAM,YAAY,CAAA;AAElE;;;;;;;;;;;;;;GAcG;AACH,MAAM,WAAW,iBAAiB;IACjC,qEAAqE;IACrE,IAAI,CAAC,EAAE,MAAM,CAAA;IACb,QAAQ,EAAE,WAAW,CAAA;IACrB,KAAK,EAAE,MAAM,CAAA;IACb;;;;;;;OAOG;IACH,MAAM,EAAE,MAAM,CAAA;IACd;;;;;;;;;;;OAWG;IACH,KAAK,CAAC,EAAE,MAAM,CAAA;IACd;;;;;;OAMG;IACH,iBAAiB,CAAC,EAAE,OAAO,CAAA;IAC3B;;;;;;OAMG;IACH,cAAc,CAAC,EAAE,MAAM,CAAA;CACvB;AAwID,wBAAgB,WAAW,CAAC,MAAM,EAAE,iBAAiB,GAAG,MAAM,CAyC7D"}
@@ -0,0 +1,136 @@
1
+ import { collect } from '../provider/collect.js';
2
+ const DEFAULT_SCALE = 4;
3
+ const DEFAULT_MAX_OUTPUT_CHARS = 20_000;
4
+ /**
5
+ * Pull the verdict out of the reply.
6
+ *
7
+ * Models wrap JSON in prose or fences however firmly asked not to, so the
8
+ * first balanced object is extracted rather than the whole reply parsed.
9
+ * What is NOT tolerated is a missing or out-of-range grade: that is an
10
+ * unusable measurement, and the caller has to hear about it rather than
11
+ * receive a number invented from a fallback.
12
+ */
13
+ function parseVerdict(reply, scale) {
14
+ const start = reply.indexOf('{');
15
+ if (start === -1)
16
+ throw new Error(`judge returned no JSON object: ${reply.slice(0, 200)}`);
17
+ let depth = 0;
18
+ let inString = false;
19
+ let escaped = false;
20
+ let end = -1;
21
+ for (let i = start; i < reply.length; i++) {
22
+ const ch = reply[i];
23
+ if (escaped) {
24
+ escaped = false;
25
+ continue;
26
+ }
27
+ if (ch === '\\') {
28
+ escaped = true;
29
+ continue;
30
+ }
31
+ if (ch === '"') {
32
+ inString = !inString;
33
+ continue;
34
+ }
35
+ if (inString)
36
+ continue;
37
+ if (ch === '{')
38
+ depth++;
39
+ else if (ch === '}') {
40
+ depth--;
41
+ if (depth === 0) {
42
+ end = i;
43
+ break;
44
+ }
45
+ }
46
+ }
47
+ if (end === -1)
48
+ throw new Error(`judge returned an unterminated JSON object: ${reply.slice(0, 200)}`);
49
+ let parsed;
50
+ try {
51
+ parsed = JSON.parse(reply.slice(start, end + 1));
52
+ }
53
+ catch (err) {
54
+ throw new Error(`judge returned unparsable JSON: ${err instanceof Error ? err.message : String(err)}`);
55
+ }
56
+ const grade = parsed.grade;
57
+ if (typeof grade !== 'number' || !Number.isFinite(grade)) {
58
+ throw new Error(`judge returned no numeric grade: ${JSON.stringify(parsed).slice(0, 200)}`);
59
+ }
60
+ if (grade < 0 || grade > scale) {
61
+ // Clamping instead would quietly turn a judge that misread the scale
62
+ // into a confident score, and a judge that misread the scale did not
63
+ // apply the rubric either.
64
+ throw new Error(`judge returned ${grade}, outside the 0..${scale} scale it was given`);
65
+ }
66
+ const reason = parsed.reason;
67
+ return {
68
+ grade,
69
+ reason: typeof reason === 'string' && reason.length > 0 ? reason : '(judge gave no reason)',
70
+ };
71
+ }
72
+ function buildPrompt(config, run, evalCase, scale) {
73
+ const limit = config.maxOutputChars ?? DEFAULT_MAX_OUTPUT_CHARS;
74
+ const answer = run.output ?? '';
75
+ const truncated = answer.length > limit;
76
+ const sections = [
77
+ 'Grade the ANSWER against the RUBRIC.',
78
+ '',
79
+ `RUBRIC:\n${config.rubric}`,
80
+ '',
81
+ `TASK:\n${typeof evalCase.input === 'string' ? evalCase.input : JSON.stringify(evalCase.input)}`,
82
+ ];
83
+ if (evalCase.expected !== undefined) {
84
+ sections.push('', `REFERENCE (one good answer, not the only one):\n${typeof evalCase.expected === 'string'
85
+ ? evalCase.expected
86
+ : JSON.stringify(evalCase.expected)}`);
87
+ }
88
+ if (config.includeTrajectory === true) {
89
+ sections.push('', `TOOLS CALLED, in order:\n${run.toolCalls.length > 0 ? run.toolCalls.join(' -> ') : '(none)'}`);
90
+ }
91
+ sections.push('', `ANSWER:\n${truncated ? answer.slice(0, limit) : answer}`);
92
+ if (truncated) {
93
+ // Disclosed, so the judge does not mark the answer down for an
94
+ // ending we removed.
95
+ sections.push('', `(The answer above was cut at ${limit} characters by the harness, not by the agent. Do not penalise it for ending abruptly.)`);
96
+ }
97
+ sections.push('', `Reply with JSON and nothing else: {"grade": <integer 0..${scale}>, "reason": "<one or two sentences>"}.`, `0 means the rubric is not met at all; ${scale} means it is fully met. Judge only against the rubric.`);
98
+ return sections.join('\n');
99
+ }
100
+ export function judgeScorer(config) {
101
+ const scale = config.scale ?? DEFAULT_SCALE;
102
+ if (!Number.isInteger(scale) || scale < 1) {
103
+ throw new Error(`judgeScorer: scale must be a positive integer, got ${scale}`);
104
+ }
105
+ if (config.rubric.trim().length === 0) {
106
+ throw new Error('judgeScorer: a rubric is required. A judge with no rubric rates fluency, which is not what anyone means to measure.');
107
+ }
108
+ return {
109
+ name: config.name ?? 'judge',
110
+ async score(run, evalCase) {
111
+ const response = await collect(config.provider.chatStream({
112
+ model: config.model,
113
+ messages: [{ role: 'user', content: buildPrompt(config, run, evalCase, scale) }],
114
+ // The same run must grade the same way twice, or a
115
+ // regression cannot be told from sampling noise.
116
+ temperature: 0,
117
+ maxTokens: 512,
118
+ }));
119
+ const verdict = parseVerdict(response.message.content ?? '', scale);
120
+ return {
121
+ score: verdict.grade / scale,
122
+ reason: verdict.reason,
123
+ details: {
124
+ grade: verdict.grade,
125
+ scale,
126
+ // Carried so a suite can account for what the judging
127
+ // itself cost. A judge is the most expensive scorer there
128
+ // is, and a bill nobody can attribute is a bill nobody
129
+ // controls.
130
+ judgeTokens: response.usage.totalTokens,
131
+ },
132
+ };
133
+ },
134
+ };
135
+ }
136
+ //# sourceMappingURL=judge.js.map
@@ -0,0 +1 @@
1
+ {"version":3,"file":"judge.js","sourceRoot":"","sources":["../../src/eval/judge.ts"],"names":[],"mappings":"AAAA,OAAO,EAAE,OAAO,EAAE,MAAM,wBAAwB,CAAA;AAgEhD,MAAM,aAAa,GAAG,CAAC,CAAA;AACvB,MAAM,wBAAwB,GAAG,MAAM,CAAA;AAOvC;;;;;;;;GAQG;AACH,SAAS,YAAY,CAAC,KAAa,EAAE,KAAa;IACjD,MAAM,KAAK,GAAG,KAAK,CAAC,OAAO,CAAC,GAAG,CAAC,CAAA;IAChC,IAAI,KAAK,KAAK,CAAC,CAAC;QAAE,MAAM,IAAI,KAAK,CAAC,kCAAkC,KAAK,CAAC,KAAK,CAAC,CAAC,EAAE,GAAG,CAAC,EAAE,CAAC,CAAA;IAE1F,IAAI,KAAK,GAAG,CAAC,CAAA;IACb,IAAI,QAAQ,GAAG,KAAK,CAAA;IACpB,IAAI,OAAO,GAAG,KAAK,CAAA;IACnB,IAAI,GAAG,GAAG,CAAC,CAAC,CAAA;IACZ,KAAK,IAAI,CAAC,GAAG,KAAK,EAAE,CAAC,GAAG,KAAK,CAAC,MAAM,EAAE,CAAC,EAAE,EAAE,CAAC;QAC3C,MAAM,EAAE,GAAG,KAAK,CAAC,CAAC,CAAC,CAAA;QACnB,IAAI,OAAO,EAAE,CAAC;YACb,OAAO,GAAG,KAAK,CAAA;YACf,SAAQ;QACT,CAAC;QACD,IAAI,EAAE,KAAK,IAAI,EAAE,CAAC;YACjB,OAAO,GAAG,IAAI,CAAA;YACd,SAAQ;QACT,CAAC;QACD,IAAI,EAAE,KAAK,GAAG,EAAE,CAAC;YAChB,QAAQ,GAAG,CAAC,QAAQ,CAAA;YACpB,SAAQ;QACT,CAAC;QACD,IAAI,QAAQ;YAAE,SAAQ;QACtB,IAAI,EAAE,KAAK,GAAG;YAAE,KAAK,EAAE,CAAA;aAClB,IAAI,EAAE,KAAK,GAAG,EAAE,CAAC;YACrB,KAAK,EAAE,CAAA;YACP,IAAI,KAAK,KAAK,CAAC,EAAE,CAAC;gBACjB,GAAG,GAAG,CAAC,CAAA;gBACP,MAAK;YACN,CAAC;QACF,CAAC;IACF,CAAC;IACD,IAAI,GAAG,KAAK,CAAC,CAAC;QACb,MAAM,IAAI,KAAK,CAAC,+CAA+C,KAAK,CAAC,KAAK,CAAC,CAAC,EAAE,GAAG,CAAC,EAAE,CAAC,CAAA;IAEtF,IAAI,MAAe,CAAA;IACnB,IAAI,CAAC;QACJ,MAAM,GAAG,IAAI,CAAC,KAAK,CAAC,KAAK,CAAC,KAAK,CAAC,KAAK,EAAE,GAAG,GAAG,CAAC,CAAC,CAAC,CAAA;IACjD,CAAC;IAAC,OAAO,GAAG,EAAE,CAAC;QACd,MAAM,IAAI,KAAK,CACd,mCAAmC,GAAG,YAAY,KAAK,CAAC,CAAC,CAAC,GAAG,CAAC,OAAO,CAAC,CAAC,CAAC,MAAM,CAAC,GAAG,CAAC,EAAE,CACrF,CAAA;IACF,CAAC;IAED,MAAM,KAAK,GAAI,MAA8B,CAAC,KAAK,CAAA;IACnD,IAAI,OAAO,KAAK,KAAK,QAAQ,IAAI,CAAC,MAAM,CAAC,QAAQ,CAAC,KAAK,CAAC,EAAE,CAAC;QAC1D,MAAM,IAAI,KAAK,CAAC,oCAAoC,IAAI,CAAC,SAAS,CAAC,MAAM,CAAC,CAAC,KAAK,CAAC,CAAC,EAAE,GAAG,CAAC,EAAE,CAAC,CAAA;IAC5F,CAAC;IACD,IAAI,KAAK,GAAG,CAAC,IAAI,KAAK,GAAG,KAAK,EAAE,CAAC;QAChC,qEAAqE;QACrE,qEAAqE;QACrE,2BAA2B;QAC3B,MAAM,IAAI,KAAK,CAAC,kBAAkB,KAAK,oBAAoB,KAAK,qBAAqB,CAAC,CAAA;IACvF,CAAC;IAED,MAAM,MAAM,GAAI,MAA+B,CAAC,MAAM,CAAA;IACtD,OAAO;QACN,KAAK;QACL,MAAM,EAAE,OAAO,MAAM,KAAK,QAAQ,IAAI,MAAM,CAAC,MAAM,GAAG,CAAC,CAAC,CAAC,CAAC,MAAM,CAAC,CAAC,CAAC,wBAAwB;KAC3F,CAAA;AACF,CAAC;AAED,SAAS,WAAW,CACnB,MAAyB,EACzB,GAAY,EACZ,QAAkB,EAClB,KAAa;IAEb,MAAM,KAAK,GAAG,MAAM,CAAC,cAAc,IAAI,wBAAwB,CAAA;IAC/D,MAAM,MAAM,GAAG,GAAG,CAAC,MAAM,IAAI,EAAE,CAAA;IAC/B,MAAM,SAAS,GAAG,MAAM,CAAC,MAAM,GAAG,KAAK,CAAA;IAEvC,MAAM,QAAQ,GAAa;QAC1B,sCAAsC;QACtC,EAAE;QACF,YAAY,MAAM,CAAC,MAAM,EAAE;QAC3B,EAAE;QACF,UAAU,OAAO,QAAQ,CAAC,KAAK,KAAK,QAAQ,CAAC,CAAC,CAAC,QAAQ,CAAC,KAAK,CAAC,CAAC,CAAC,IAAI,CAAC,SAAS,CAAC,QAAQ,CAAC,KAAK,CAAC,EAAE;KAChG,CAAA;IAED,IAAI,QAAQ,CAAC,QAAQ,KAAK,SAAS,EAAE,CAAC;QACrC,QAAQ,CAAC,IAAI,CACZ,EAAE,EACF,mDACC,OAAO,QAAQ,CAAC,QAAQ,KAAK,QAAQ;YACpC,CAAC,CAAC,QAAQ,CAAC,QAAQ;YACnB,CAAC,CAAC,IAAI,CAAC,SAAS,CAAC,QAAQ,CAAC,QAAQ,CACpC,EAAE,CACF,CAAA;IACF,CAAC;IAED,IAAI,MAAM,CAAC,iBAAiB,KAAK,IAAI,EAAE,CAAC;QACvC,QAAQ,CAAC,IAAI,CACZ,EAAE,EACF,4BAA4B,GAAG,CAAC,SAAS,CAAC,MAAM,GAAG,CAAC,CAAC,CAAC,CAAC,GAAG,CAAC,SAAS,CAAC,IAAI,CAAC,MAAM,CAAC,CAAC,CAAC,CAAC,QAAQ,EAAE,CAC9F,CAAA;IACF,CAAC;IAED,QAAQ,CAAC,IAAI,CAAC,EAAE,EAAE,YAAY,SAAS,CAAC,CAAC,CAAC,MAAM,CAAC,KAAK,CAAC,CAAC,EAAE,KAAK,CAAC,CAAC,CAAC,CAAC,MAAM,EAAE,CAAC,CAAA;IAC5E,IAAI,SAAS,EAAE,CAAC;QACf,+DAA+D;QAC/D,qBAAqB;QACrB,QAAQ,CAAC,IAAI,CACZ,EAAE,EACF,gCAAgC,KAAK,wFAAwF,CAC7H,CAAA;IACF,CAAC;IAED,QAAQ,CAAC,IAAI,CACZ,EAAE,EACF,2DAA2D,KAAK,yCAAyC,EACzG,yCAAyC,KAAK,wDAAwD,CACtG,CAAA;IAED,OAAO,QAAQ,CAAC,IAAI,CAAC,IAAI,CAAC,CAAA;AAC3B,CAAC;AAED,MAAM,UAAU,WAAW,CAAC,MAAyB;IACpD,MAAM,KAAK,GAAG,MAAM,CAAC,KAAK,IAAI,aAAa,CAAA;IAC3C,IAAI,CAAC,MAAM,CAAC,SAAS,CAAC,KAAK,CAAC,IAAI,KAAK,GAAG,CAAC,EAAE,CAAC;QAC3C,MAAM,IAAI,KAAK,CAAC,sDAAsD,KAAK,EAAE,CAAC,CAAA;IAC/E,CAAC;IACD,IAAI,MAAM,CAAC,MAAM,CAAC,IAAI,EAAE,CAAC,MAAM,KAAK,CAAC,EAAE,CAAC;QACvC,MAAM,IAAI,KAAK,CACd,qHAAqH,CACrH,CAAA;IACF,CAAC;IAED,OAAO;QACN,IAAI,EAAE,MAAM,CAAC,IAAI,IAAI,OAAO;QAC5B,KAAK,CAAC,KAAK,CAAC,GAAY,EAAE,QAAkB;YAC3C,MAAM,QAAQ,GAAG,MAAM,OAAO,CAC7B,MAAM,CAAC,QAAQ,CAAC,UAAU,CAAC;gBAC1B,KAAK,EAAE,MAAM,CAAC,KAAK;gBACnB,QAAQ,EAAE,CAAC,EAAE,IAAI,EAAE,MAAM,EAAE,OAAO,EAAE,WAAW,CAAC,MAAM,EAAE,GAAG,EAAE,QAAQ,EAAE,KAAK,CAAC,EAAE,CAAC;gBAChF,mDAAmD;gBACnD,iDAAiD;gBACjD,WAAW,EAAE,CAAC;gBACd,SAAS,EAAE,GAAG;aACd,CAAC,CACF,CAAA;YAED,MAAM,OAAO,GAAG,YAAY,CAAC,QAAQ,CAAC,OAAO,CAAC,OAAO,IAAI,EAAE,EAAE,KAAK,CAAC,CAAA;YACnE,OAAO;gBACN,KAAK,EAAE,OAAO,CAAC,KAAK,GAAG,KAAK;gBAC5B,MAAM,EAAE,OAAO,CAAC,MAAM;gBACtB,OAAO,EAAE;oBACR,KAAK,EAAE,OAAO,CAAC,KAAK;oBACpB,KAAK;oBACL,sDAAsD;oBACtD,0DAA0D;oBAC1D,uDAAuD;oBACvD,YAAY;oBACZ,WAAW,EAAE,QAAQ,CAAC,KAAK,CAAC,WAAW;iBACvC;aACD,CAAA;QACF,CAAC;KACD,CAAA;AACF,CAAC"}
@@ -0,0 +1,40 @@
1
+ import type { EvalCase, EvalRun, Score, Scorer } from './types.js';
2
+ /**
3
+ * How closely the run's tool sequence matched the expected one, as F1 over
4
+ * the longest common subsequence.
5
+ *
6
+ * Trajectory, not final answer. Namzu's most load-bearing behavior is
7
+ * tuned by constants nobody could measure — `search_tools` activates the
8
+ * top 5 deferred tools, compaction fires at 0.7, six state lists cap at
9
+ * 25 — and changing any of them, or a tool description, or the
10
+ * deferred-tools prompt block, could silently make the agent take four
11
+ * tool calls where it took one. Final-answer scoring cannot see that;
12
+ * this can.
13
+ *
14
+ * Extra calls cut precision, missing calls cut recall, so a run that does
15
+ * the right thing wastefully and a run that skips a step score
16
+ * differently — which is the distinction a final-answer score collapses.
17
+ */
18
+ export declare function trajectoryScorer(): Scorer;
19
+ /** The run settled cleanly rather than erroring or being cut off. */
20
+ export declare function completionScorer(acceptable?: readonly string[]): Scorer;
21
+ /**
22
+ * The run stayed within a step budget.
23
+ *
24
+ * A regression that makes the agent take four turns where it took one is
25
+ * invisible to correctness scoring and very visible on the bill.
26
+ */
27
+ export declare function stepBudgetScorer(maxSteps: number): Scorer;
28
+ /** The final text contains every required substring. */
29
+ export declare function containsScorer(...required: string[]): Scorer;
30
+ /**
31
+ * Judge the run with a caller-supplied predicate.
32
+ *
33
+ * The escape hatch for anything the built-in scorers do not cover. For
34
+ * a model-graded judge, reach for `judgeScorer` first: it handles the
35
+ * parts that are easy to get wrong — a rubric it refuses to run without,
36
+ * an ordinal scale rather than a float, disclosed truncation, and a
37
+ * failed call that reports as unjudged rather than as a zero.
38
+ */
39
+ export declare function customScorer(name: string, fn: (run: EvalRun, evalCase: EvalCase) => Score | Promise<Score>): Scorer;
40
+ //# sourceMappingURL=scorers.d.ts.map
@@ -0,0 +1 @@
1
+ {"version":3,"file":"scorers.d.ts","sourceRoot":"","sources":["../../src/eval/scorers.ts"],"names":[],"mappings":"AAAA,OAAO,KAAK,EAAE,QAAQ,EAAE,OAAO,EAAE,KAAK,EAAE,MAAM,EAAE,MAAM,YAAY,CAAA;AAyBlE;;;;;;;;;;;;;;;GAeG;AACH,wBAAgB,gBAAgB,IAAI,MAAM,CAwCzC;AAED,qEAAqE;AACrE,wBAAgB,gBAAgB,CAC/B,UAAU,GAAE,SAAS,MAAM,EAAmC,GAC5D,MAAM,CAqBR;AAED;;;;;GAKG;AACH,wBAAgB,gBAAgB,CAAC,QAAQ,EAAE,MAAM,GAAG,MAAM,CAezD;AAED,wDAAwD;AACxD,wBAAgB,cAAc,CAAC,GAAG,QAAQ,EAAE,MAAM,EAAE,GAAG,MAAM,CAmB5D;AAED;;;;;;;;GAQG;AACH,wBAAgB,YAAY,CAC3B,IAAI,EAAE,MAAM,EACZ,EAAE,EAAE,CAAC,GAAG,EAAE,OAAO,EAAE,QAAQ,EAAE,QAAQ,KAAK,KAAK,GAAG,OAAO,CAAC,KAAK,CAAC,GAC9D,MAAM,CAER"}
@@ -0,0 +1,151 @@
1
+ /**
2
+ * Longest common subsequence length between two tool sequences.
3
+ *
4
+ * Subsequence, not set intersection: order carries meaning in a
5
+ * trajectory. Reading a file before editing it is not the same run as
6
+ * editing then reading, and a set-based score cannot tell them apart.
7
+ */
8
+ function lcsLength(a, b) {
9
+ const table = Array.from({ length: a.length + 1 }, () => new Array(b.length + 1).fill(0));
10
+ for (let i = 1; i <= a.length; i++) {
11
+ for (let j = 1; j <= b.length; j++) {
12
+ const row = table[i];
13
+ const prev = table[i - 1];
14
+ if (!row || !prev)
15
+ continue;
16
+ row[j] =
17
+ a[i - 1] === b[j - 1] ? (prev[j - 1] ?? 0) + 1 : Math.max(prev[j] ?? 0, row[j - 1] ?? 0);
18
+ }
19
+ }
20
+ return table[a.length]?.[b.length] ?? 0;
21
+ }
22
+ /**
23
+ * How closely the run's tool sequence matched the expected one, as F1 over
24
+ * the longest common subsequence.
25
+ *
26
+ * Trajectory, not final answer. Namzu's most load-bearing behavior is
27
+ * tuned by constants nobody could measure — `search_tools` activates the
28
+ * top 5 deferred tools, compaction fires at 0.7, six state lists cap at
29
+ * 25 — and changing any of them, or a tool description, or the
30
+ * deferred-tools prompt block, could silently make the agent take four
31
+ * tool calls where it took one. Final-answer scoring cannot see that;
32
+ * this can.
33
+ *
34
+ * Extra calls cut precision, missing calls cut recall, so a run that does
35
+ * the right thing wastefully and a run that skips a step score
36
+ * differently — which is the distinction a final-answer score collapses.
37
+ */
38
+ export function trajectoryScorer() {
39
+ return {
40
+ name: 'trajectory',
41
+ score(run, evalCase) {
42
+ const expected = evalCase.expectedTools ?? [];
43
+ const actual = run.toolCalls;
44
+ if (expected.length === 0 && actual.length === 0) {
45
+ return { score: 1, reason: 'no tools expected and none called' };
46
+ }
47
+ if (expected.length === 0) {
48
+ return {
49
+ score: 0,
50
+ reason: `expected no tool calls, got ${actual.length}: ${actual.join(' → ')}`,
51
+ details: { actual },
52
+ };
53
+ }
54
+ if (actual.length === 0) {
55
+ return {
56
+ score: 0,
57
+ reason: `expected ${expected.join(' → ')}, but no tools were called`,
58
+ details: { expected },
59
+ };
60
+ }
61
+ const matched = lcsLength(expected, actual);
62
+ const precision = matched / actual.length;
63
+ const recall = matched / expected.length;
64
+ const f1 = precision + recall === 0 ? 0 : (2 * precision * recall) / (precision + recall);
65
+ return {
66
+ score: f1,
67
+ reason: f1 === 1
68
+ ? `exact trajectory: ${actual.join(' → ')}`
69
+ : `matched ${matched}/${expected.length} in order; expected ${expected.join(' → ')}, got ${actual.join(' → ')}`,
70
+ details: { precision, recall, matched, expected, actual },
71
+ };
72
+ },
73
+ };
74
+ }
75
+ /** The run settled cleanly rather than erroring or being cut off. */
76
+ export function completionScorer(acceptable = ['end_turn', 'stop_condition']) {
77
+ return {
78
+ name: 'completion',
79
+ // Binary by construction: the run either settled cleanly or it did
80
+ // not, so there is no score between 0 and 1 for a mean to soften.
81
+ // Averaged in, a hard 0 here is carried by three good fuzzy scores.
82
+ severity: 'gate',
83
+ score(run) {
84
+ if (run.error) {
85
+ return { score: 0, reason: `run failed: ${run.error}` };
86
+ }
87
+ const ok = run.stopReason !== undefined && acceptable.includes(run.stopReason);
88
+ return {
89
+ score: ok ? 1 : 0,
90
+ reason: ok
91
+ ? `settled as ${run.stopReason}`
92
+ : `settled as ${run.stopReason ?? 'unknown'}, expected one of ${acceptable.join(', ')}`,
93
+ details: { stopReason: run.stopReason },
94
+ };
95
+ },
96
+ };
97
+ }
98
+ /**
99
+ * The run stayed within a step budget.
100
+ *
101
+ * A regression that makes the agent take four turns where it took one is
102
+ * invisible to correctness scoring and very visible on the bill.
103
+ */
104
+ export function stepBudgetScorer(maxSteps) {
105
+ return {
106
+ name: 'step-budget',
107
+ score(run) {
108
+ const used = run.steps.length;
109
+ return {
110
+ score: used <= maxSteps ? 1 : Math.max(0, maxSteps / used),
111
+ reason: used <= maxSteps
112
+ ? `${used}/${maxSteps} steps`
113
+ : `over budget: ${used} steps against a ${maxSteps} allowance`,
114
+ details: { used, maxSteps },
115
+ };
116
+ },
117
+ };
118
+ }
119
+ /** The final text contains every required substring. */
120
+ export function containsScorer(...required) {
121
+ return {
122
+ name: 'contains',
123
+ // A required fragment is required. Partial credit here reads as
124
+ // "most of the answer was there", which is not what was asked.
125
+ severity: 'gate',
126
+ score(run) {
127
+ const text = run.output ?? '';
128
+ const missing = required.filter((r) => !text.includes(r));
129
+ return {
130
+ score: required.length === 0 ? 1 : (required.length - missing.length) / required.length,
131
+ reason: missing.length === 0
132
+ ? `found all ${required.length} required fragments`
133
+ : `missing: ${missing.join(', ')}`,
134
+ details: { missing },
135
+ };
136
+ },
137
+ };
138
+ }
139
+ /**
140
+ * Judge the run with a caller-supplied predicate.
141
+ *
142
+ * The escape hatch for anything the built-in scorers do not cover. For
143
+ * a model-graded judge, reach for `judgeScorer` first: it handles the
144
+ * parts that are easy to get wrong — a rubric it refuses to run without,
145
+ * an ordinal scale rather than a float, disclosed truncation, and a
146
+ * failed call that reports as unjudged rather than as a zero.
147
+ */
148
+ export function customScorer(name, fn) {
149
+ return { name, score: fn };
150
+ }
151
+ //# sourceMappingURL=scorers.js.map
@@ -0,0 +1 @@
1
+ {"version":3,"file":"scorers.js","sourceRoot":"","sources":["../../src/eval/scorers.ts"],"names":[],"mappings":"AAEA;;;;;;GAMG;AACH,SAAS,SAAS,CAAC,CAAoB,EAAE,CAAoB;IAC5D,MAAM,KAAK,GAAe,KAAK,CAAC,IAAI,CAAC,EAAE,MAAM,EAAE,CAAC,CAAC,MAAM,GAAG,CAAC,EAAE,EAAE,GAAG,EAAE,CACnE,IAAI,KAAK,CAAC,CAAC,CAAC,MAAM,GAAG,CAAC,CAAC,CAAC,IAAI,CAAC,CAAC,CAAC,CAC/B,CAAA;IACD,KAAK,IAAI,CAAC,GAAG,CAAC,EAAE,CAAC,IAAI,CAAC,CAAC,MAAM,EAAE,CAAC,EAAE,EAAE,CAAC;QACpC,KAAK,IAAI,CAAC,GAAG,CAAC,EAAE,CAAC,IAAI,CAAC,CAAC,MAAM,EAAE,CAAC,EAAE,EAAE,CAAC;YACpC,MAAM,GAAG,GAAG,KAAK,CAAC,CAAC,CAAC,CAAA;YACpB,MAAM,IAAI,GAAG,KAAK,CAAC,CAAC,GAAG,CAAC,CAAC,CAAA;YACzB,IAAI,CAAC,GAAG,IAAI,CAAC,IAAI;gBAAE,SAAQ;YAC3B,GAAG,CAAC,CAAC,CAAC;gBACL,CAAC,CAAC,CAAC,GAAG,CAAC,CAAC,KAAK,CAAC,CAAC,CAAC,GAAG,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC,IAAI,CAAC,CAAC,GAAG,CAAC,CAAC,IAAI,CAAC,CAAC,GAAG,CAAC,CAAC,CAAC,CAAC,IAAI,CAAC,GAAG,CAAC,IAAI,CAAC,CAAC,CAAC,IAAI,CAAC,EAAE,GAAG,CAAC,CAAC,GAAG,CAAC,CAAC,IAAI,CAAC,CAAC,CAAA;QAC1F,CAAC;IACF,CAAC;IACD,OAAO,KAAK,CAAC,CAAC,CAAC,MAAM,CAAC,EAAE,CAAC,CAAC,CAAC,MAAM,CAAC,IAAI,CAAC,CAAA;AACxC,CAAC;AAED;;;;;;;;;;;;;;;GAeG;AACH,MAAM,UAAU,gBAAgB;IAC/B,OAAO;QACN,IAAI,EAAE,YAAY;QAClB,KAAK,CAAC,GAAY,EAAE,QAAkB;YACrC,MAAM,QAAQ,GAAG,QAAQ,CAAC,aAAa,IAAI,EAAE,CAAA;YAC7C,MAAM,MAAM,GAAG,GAAG,CAAC,SAAS,CAAA;YAE5B,IAAI,QAAQ,CAAC,MAAM,KAAK,CAAC,IAAI,MAAM,CAAC,MAAM,KAAK,CAAC,EAAE,CAAC;gBAClD,OAAO,EAAE,KAAK,EAAE,CAAC,EAAE,MAAM,EAAE,mCAAmC,EAAE,CAAA;YACjE,CAAC;YACD,IAAI,QAAQ,CAAC,MAAM,KAAK,CAAC,EAAE,CAAC;gBAC3B,OAAO;oBACN,KAAK,EAAE,CAAC;oBACR,MAAM,EAAE,+BAA+B,MAAM,CAAC,MAAM,KAAK,MAAM,CAAC,IAAI,CAAC,KAAK,CAAC,EAAE;oBAC7E,OAAO,EAAE,EAAE,MAAM,EAAE;iBACnB,CAAA;YACF,CAAC;YACD,IAAI,MAAM,CAAC,MAAM,KAAK,CAAC,EAAE,CAAC;gBACzB,OAAO;oBACN,KAAK,EAAE,CAAC;oBACR,MAAM,EAAE,YAAY,QAAQ,CAAC,IAAI,CAAC,KAAK,CAAC,4BAA4B;oBACpE,OAAO,EAAE,EAAE,QAAQ,EAAE;iBACrB,CAAA;YACF,CAAC;YAED,MAAM,OAAO,GAAG,SAAS,CAAC,QAAQ,EAAE,MAAM,CAAC,CAAA;YAC3C,MAAM,SAAS,GAAG,OAAO,GAAG,MAAM,CAAC,MAAM,CAAA;YACzC,MAAM,MAAM,GAAG,OAAO,GAAG,QAAQ,CAAC,MAAM,CAAA;YACxC,MAAM,EAAE,GAAG,SAAS,GAAG,MAAM,KAAK,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC,GAAG,SAAS,GAAG,MAAM,CAAC,GAAG,CAAC,SAAS,GAAG,MAAM,CAAC,CAAA;YAEzF,OAAO;gBACN,KAAK,EAAE,EAAE;gBACT,MAAM,EACL,EAAE,KAAK,CAAC;oBACP,CAAC,CAAC,qBAAqB,MAAM,CAAC,IAAI,CAAC,KAAK,CAAC,EAAE;oBAC3C,CAAC,CAAC,WAAW,OAAO,IAAI,QAAQ,CAAC,MAAM,uBAAuB,QAAQ,CAAC,IAAI,CAAC,KAAK,CAAC,SAAS,MAAM,CAAC,IAAI,CAAC,KAAK,CAAC,EAAE;gBACjH,OAAO,EAAE,EAAE,SAAS,EAAE,MAAM,EAAE,OAAO,EAAE,QAAQ,EAAE,MAAM,EAAE;aACzD,CAAA;QACF,CAAC;KACD,CAAA;AACF,CAAC;AAED,qEAAqE;AACrE,MAAM,UAAU,gBAAgB,CAC/B,aAAgC,CAAC,UAAU,EAAE,gBAAgB,CAAC;IAE9D,OAAO;QACN,IAAI,EAAE,YAAY;QAClB,mEAAmE;QACnE,kEAAkE;QAClE,oEAAoE;QACpE,QAAQ,EAAE,MAAM;QAChB,KAAK,CAAC,GAAY;YACjB,IAAI,GAAG,CAAC,KAAK,EAAE,CAAC;gBACf,OAAO,EAAE,KAAK,EAAE,CAAC,EAAE,MAAM,EAAE,eAAe,GAAG,CAAC,KAAK,EAAE,EAAE,CAAA;YACxD,CAAC;YACD,MAAM,EAAE,GAAG,GAAG,CAAC,UAAU,KAAK,SAAS,IAAI,UAAU,CAAC,QAAQ,CAAC,GAAG,CAAC,UAAU,CAAC,CAAA;YAC9E,OAAO;gBACN,KAAK,EAAE,EAAE,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC;gBACjB,MAAM,EAAE,EAAE;oBACT,CAAC,CAAC,cAAc,GAAG,CAAC,UAAU,EAAE;oBAChC,CAAC,CAAC,cAAc,GAAG,CAAC,UAAU,IAAI,SAAS,qBAAqB,UAAU,CAAC,IAAI,CAAC,IAAI,CAAC,EAAE;gBACxF,OAAO,EAAE,EAAE,UAAU,EAAE,GAAG,CAAC,UAAU,EAAE;aACvC,CAAA;QACF,CAAC;KACD,CAAA;AACF,CAAC;AAED;;;;;GAKG;AACH,MAAM,UAAU,gBAAgB,CAAC,QAAgB;IAChD,OAAO;QACN,IAAI,EAAE,aAAa;QACnB,KAAK,CAAC,GAAY;YACjB,MAAM,IAAI,GAAG,GAAG,CAAC,KAAK,CAAC,MAAM,CAAA;YAC7B,OAAO;gBACN,KAAK,EAAE,IAAI,IAAI,QAAQ,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC,IAAI,CAAC,GAAG,CAAC,CAAC,EAAE,QAAQ,GAAG,IAAI,CAAC;gBAC1D,MAAM,EACL,IAAI,IAAI,QAAQ;oBACf,CAAC,CAAC,GAAG,IAAI,IAAI,QAAQ,QAAQ;oBAC7B,CAAC,CAAC,gBAAgB,IAAI,oBAAoB,QAAQ,YAAY;gBAChE,OAAO,EAAE,EAAE,IAAI,EAAE,QAAQ,EAAE;aAC3B,CAAA;QACF,CAAC;KACD,CAAA;AACF,CAAC;AAED,wDAAwD;AACxD,MAAM,UAAU,cAAc,CAAC,GAAG,QAAkB;IACnD,OAAO;QACN,IAAI,EAAE,UAAU;QAChB,gEAAgE;QAChE,+DAA+D;QAC/D,QAAQ,EAAE,MAAM;QAChB,KAAK,CAAC,GAAY;YACjB,MAAM,IAAI,GAAG,GAAG,CAAC,MAAM,IAAI,EAAE,CAAA;YAC7B,MAAM,OAAO,GAAG,QAAQ,CAAC,MAAM,CAAC,CAAC,CAAC,EAAE,EAAE,CAAC,CAAC,IAAI,CAAC,QAAQ,CAAC,CAAC,CAAC,CAAC,CAAA;YACzD,OAAO;gBACN,KAAK,EAAE,QAAQ,CAAC,MAAM,KAAK,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC,QAAQ,CAAC,MAAM,GAAG,OAAO,CAAC,MAAM,CAAC,GAAG,QAAQ,CAAC,MAAM;gBACvF,MAAM,EACL,OAAO,CAAC,MAAM,KAAK,CAAC;oBACnB,CAAC,CAAC,aAAa,QAAQ,CAAC,MAAM,qBAAqB;oBACnD,CAAC,CAAC,YAAY,OAAO,CAAC,IAAI,CAAC,IAAI,CAAC,EAAE;gBACpC,OAAO,EAAE,EAAE,OAAO,EAAE;aACpB,CAAA;QACF,CAAC;KACD,CAAA;AACF,CAAC;AAED;;;;;;;;GAQG;AACH,MAAM,UAAU,YAAY,CAC3B,IAAY,EACZ,EAAgE;IAEhE,OAAO,EAAE,IAAI,EAAE,KAAK,EAAE,EAAE,EAAE,CAAA;AAC3B,CAAC"}
@@ -0,0 +1,150 @@
1
+ import type { StepResult } from '../types/run/step.js';
2
+ /**
3
+ * One case in a dataset: an input, and what a good run looks like.
4
+ *
5
+ * `TInput` is the agent's own input type, so a dataset is checked against
6
+ * the thing it tests rather than against `unknown`.
7
+ */
8
+ export interface EvalCase<TInput = unknown> {
9
+ name: string;
10
+ input: TInput;
11
+ /** Free-form expectation, consumed by whichever scorers you attach. */
12
+ expected?: unknown;
13
+ /**
14
+ * Tool names the run is expected to call, in order. Consumed by
15
+ * `trajectoryScorer`.
16
+ */
17
+ expectedTools?: readonly string[];
18
+ /** Per-case scorer overrides; falls back to the dataset's scorers. */
19
+ scorers?: readonly Scorer[];
20
+ }
21
+ /** What actually happened, handed to every scorer. */
22
+ export interface EvalRun {
23
+ /** Final assistant text, when the run produced one. */
24
+ output: string | null;
25
+ /** Schema-validated final output, when `structuredOutput` was requested. */
26
+ structuredOutput?: unknown;
27
+ steps: readonly StepResult[];
28
+ /** Tool names in call order, flattened across steps. */
29
+ toolCalls: readonly string[];
30
+ stopReason?: string;
31
+ totalTokens: number;
32
+ totalCostUsd: number;
33
+ durationMs: number;
34
+ /** Set when the run threw rather than settling. */
35
+ error?: string;
36
+ }
37
+ /**
38
+ * A judgement about one run.
39
+ *
40
+ * `reason` is REQUIRED, not optional. A bare number tells you a run got
41
+ * worse without telling you how, which is exactly when you need to know —
42
+ * so every scorer has to be able to explain itself, and one that cannot
43
+ * is a scorer that should not exist.
44
+ */
45
+ export interface Score {
46
+ /** 0..1. Higher is better. Ignored when `unavailable` is set. */
47
+ score: number;
48
+ reason: string;
49
+ /** Anything the scorer wants to surface (counts, diffs, matched items). */
50
+ details?: Record<string, unknown>;
51
+ /**
52
+ * The scorer could not produce a judgement — it threw, its provider was
53
+ * unreachable, its verdict did not parse.
54
+ *
55
+ * A failed measurement is not a measurement of zero. Scoring it zero
56
+ * says "the run was bad" when the truth is "we do not know", and the
57
+ * two demand opposite responses: one is a regression to chase, the
58
+ * other is a broken harness to fix. This mattered little while every
59
+ * scorer was a pure function — a throw there is a bug — and matters a
60
+ * great deal now that a scorer can be a network call, where one 429
61
+ * would otherwise turn a green suite red and send somebody hunting a
62
+ * regression that never happened.
63
+ *
64
+ * An unavailable score is excluded from the case mean's numerator AND
65
+ * its denominator, and the case it belongs to is reported as
66
+ * inconclusive rather than passed or failed.
67
+ */
68
+ unavailable?: boolean;
69
+ }
70
+ /**
71
+ * How a scorer's verdict affects the case's.
72
+ *
73
+ * A single unweighted mean over every scorer, compared to one suite-wide
74
+ * threshold, made the two halves of the design fight each other. At the
75
+ * default threshold of 1 the harness never reports a false pass — but a
76
+ * trajectory F1 and a graded judge can essentially never reach 1, so every
77
+ * real suite lowers it. And every step down buys tolerance for the fuzzy
78
+ * scorers by buying the same tolerance for the deterministic ones: at 0.75,
79
+ * a trajectory score of 0 alongside three perfect scores averages to 0.75
80
+ * and reports **passed**. The exact regression the harness exists to catch
81
+ * comes back green.
82
+ *
83
+ * - `gate` — a miss fails the case outright, whatever the mean says.
84
+ * For the deterministic checks: it called the wrong tools, it did not
85
+ * finish, the required phrase is absent.
86
+ * - `soft` — contributes to the mean only. For the fuzzy ones, where a
87
+ * number below 1 is normal and a threshold is a judgement call.
88
+ *
89
+ * Defaults to `soft`, so a suite that sets nothing behaves exactly as
90
+ * before.
91
+ */
92
+ export type ScorerSeverity = 'gate' | 'soft';
93
+ export interface Scorer {
94
+ name: string;
95
+ score(run: EvalRun, evalCase: EvalCase): Score | Promise<Score>;
96
+ /** See {@link ScorerSeverity}. Default `soft`. */
97
+ severity?: ScorerSeverity;
98
+ /**
99
+ * The score this scorer must reach. Defaults to the experiment's
100
+ * `passThreshold`.
101
+ *
102
+ * Per-scorer because "good enough" is not one number across dimensions:
103
+ * a trajectory match at 0.8 may be fine while a completion check at 0.8
104
+ * is meaningless — it either finished or it did not.
105
+ */
106
+ threshold?: number;
107
+ }
108
+ /**
109
+ * `inconclusive` is a first-class outcome: every scorer that ran was
110
+ * unavailable, so there is no evidence either way. Folding it into
111
+ * `failed` invents a regression and folding it into `passed` invents
112
+ * evidence; both are worse than saying nothing was measured.
113
+ */
114
+ export type CaseStatus = 'passed' | 'failed' | 'inconclusive';
115
+ export interface CaseResult {
116
+ case: string;
117
+ run: EvalRun;
118
+ scores: Record<string, Score>;
119
+ /** Mean of this case's AVAILABLE scores. Zero when none were. */
120
+ mean: number;
121
+ status: CaseStatus;
122
+ /** True only for `status: 'passed'`. */
123
+ passed: boolean;
124
+ /**
125
+ * Gate scorers that missed their threshold, by name.
126
+ *
127
+ * Named rather than counted: a case reported failed with a mean of 0.75
128
+ * otherwise sends somebody to read four scores and guess which one
129
+ * mattered. Absent when no gate missed.
130
+ */
131
+ failedGates?: readonly string[];
132
+ }
133
+ export interface ExperimentReport {
134
+ name: string;
135
+ cases: readonly CaseResult[];
136
+ /** Mean score across the cases that produced one. */
137
+ mean: number;
138
+ passed: number;
139
+ failed: number;
140
+ /** Cases where no scorer could produce a judgement. */
141
+ inconclusive: number;
142
+ /**
143
+ * Per-scorer means over the cases each one actually judged, for
144
+ * spotting which dimension regressed. A scorer that was unavailable on
145
+ * some cases is averaged over the rest rather than dragged toward zero.
146
+ */
147
+ byScorer: Record<string, number>;
148
+ durationMs: number;
149
+ }
150
+ //# sourceMappingURL=types.d.ts.map
@@ -0,0 +1 @@
1
+ {"version":3,"file":"types.d.ts","sourceRoot":"","sources":["../../src/eval/types.ts"],"names":[],"mappings":"AAAA,OAAO,KAAK,EAAE,UAAU,EAAE,MAAM,sBAAsB,CAAA;AAEtD;;;;;GAKG;AACH,MAAM,WAAW,QAAQ,CAAC,MAAM,GAAG,OAAO;IACzC,IAAI,EAAE,MAAM,CAAA;IACZ,KAAK,EAAE,MAAM,CAAA;IACb,uEAAuE;IACvE,QAAQ,CAAC,EAAE,OAAO,CAAA;IAClB;;;OAGG;IACH,aAAa,CAAC,EAAE,SAAS,MAAM,EAAE,CAAA;IACjC,sEAAsE;IACtE,OAAO,CAAC,EAAE,SAAS,MAAM,EAAE,CAAA;CAC3B;AAED,sDAAsD;AACtD,MAAM,WAAW,OAAO;IACvB,uDAAuD;IACvD,MAAM,EAAE,MAAM,GAAG,IAAI,CAAA;IACrB,4EAA4E;IAC5E,gBAAgB,CAAC,EAAE,OAAO,CAAA;IAC1B,KAAK,EAAE,SAAS,UAAU,EAAE,CAAA;IAC5B,wDAAwD;IACxD,SAAS,EAAE,SAAS,MAAM,EAAE,CAAA;IAC5B,UAAU,CAAC,EAAE,MAAM,CAAA;IACnB,WAAW,EAAE,MAAM,CAAA;IACnB,YAAY,EAAE,MAAM,CAAA;IACpB,UAAU,EAAE,MAAM,CAAA;IAClB,mDAAmD;IACnD,KAAK,CAAC,EAAE,MAAM,CAAA;CACd;AAED;;;;;;;GAOG;AACH,MAAM,WAAW,KAAK;IACrB,iEAAiE;IACjE,KAAK,EAAE,MAAM,CAAA;IACb,MAAM,EAAE,MAAM,CAAA;IACd,2EAA2E;IAC3E,OAAO,CAAC,EAAE,MAAM,CAAC,MAAM,EAAE,OAAO,CAAC,CAAA;IACjC;;;;;;;;;;;;;;;;OAgBG;IACH,WAAW,CAAC,EAAE,OAAO,CAAA;CACrB;AAED;;;;;;;;;;;;;;;;;;;;;GAqBG;AACH,MAAM,MAAM,cAAc,GAAG,MAAM,GAAG,MAAM,CAAA;AAE5C,MAAM,WAAW,MAAM;IACtB,IAAI,EAAE,MAAM,CAAA;IACZ,KAAK,CAAC,GAAG,EAAE,OAAO,EAAE,QAAQ,EAAE,QAAQ,GAAG,KAAK,GAAG,OAAO,CAAC,KAAK,CAAC,CAAA;IAC/D,kDAAkD;IAClD,QAAQ,CAAC,EAAE,cAAc,CAAA;IACzB;;;;;;;OAOG;IACH,SAAS,CAAC,EAAE,MAAM,CAAA;CAClB;AAED;;;;;GAKG;AACH,MAAM,MAAM,UAAU,GAAG,QAAQ,GAAG,QAAQ,GAAG,cAAc,CAAA;AAE7D,MAAM,WAAW,UAAU;IAC1B,IAAI,EAAE,MAAM,CAAA;IACZ,GAAG,EAAE,OAAO,CAAA;IACZ,MAAM,EAAE,MAAM,CAAC,MAAM,EAAE,KAAK,CAAC,CAAA;IAC7B,iEAAiE;IACjE,IAAI,EAAE,MAAM,CAAA;IACZ,MAAM,EAAE,UAAU,CAAA;IAClB,wCAAwC;IACxC,MAAM,EAAE,OAAO,CAAA;IACf;;;;;;OAMG;IACH,WAAW,CAAC,EAAE,SAAS,MAAM,EAAE,CAAA;CAC/B;AAED,MAAM,WAAW,gBAAgB;IAChC,IAAI,EAAE,MAAM,CAAA;IACZ,KAAK,EAAE,SAAS,UAAU,EAAE,CAAA;IAC5B,qDAAqD;IACrD,IAAI,EAAE,MAAM,CAAA;IACZ,MAAM,EAAE,MAAM,CAAA;IACd,MAAM,EAAE,MAAM,CAAA;IACd,uDAAuD;IACvD,YAAY,EAAE,MAAM,CAAA;IACpB;;;;OAIG;IACH,QAAQ,EAAE,MAAM,CAAC,MAAM,EAAE,MAAM,CAAC,CAAA;IAChC,UAAU,EAAE,MAAM,CAAA;CAClB"}
@@ -0,0 +1,2 @@
1
+ export {};
2
+ //# sourceMappingURL=types.js.map
@@ -0,0 +1 @@
1
+ {"version":3,"file":"types.js","sourceRoot":"","sources":["../../src/eval/types.ts"],"names":[],"mappings":""}
@@ -1 +1 @@
1
- {"version":3,"file":"local.d.ts","sourceRoot":"","sources":["../../src/execution/local.ts"],"names":[],"mappings":"AAGA,OAAO,KAAK,EAAE,2BAA2B,EAAE,MAAM,6BAA6B,CAAA;AAC9E,OAAO,KAAK,EACX,eAAe,EACf,cAAc,EACd,aAAa,EACb,mBAAmB,EACnB,oBAAoB,EACpB,MAAM,6BAA6B,CAAA;AACpC,OAAO,EAAE,oBAAoB,EAAE,MAAM,WAAW,CAAA;AAEhD,MAAM,WAAW,4BAA4B;IAC5C,EAAE,EAAE,MAAM,CAAA;IACV,GAAG,EAAE,MAAM,CAAA;IACX,QAAQ,CAAC,EAAE,OAAO,CAAA;IAClB,OAAO,CAAC,EAAE,MAAM,CAAC,MAAM,EAAE,MAAM,CAAC,CAAA;IAChC,YAAY,CAAC,EAAE,mBAAmB,EAAE,CAAA;IACpC,KAAK,CAAC,EAAE,MAAM,CAAA;CACd;AAED,qBAAa,qBAAsB,SAAQ,oBAAqB,YAAW,eAAe;IACzF,QAAQ,CAAC,EAAE,EAAE,MAAM,CAAA;IACnB,QAAQ,CAAC,WAAW,EAAE,oBAAoB,CAAU;IAEpD,OAAO,CAAC,GAAG,CAAQ;IACnB,OAAO,CAAC,QAAQ,CAAS;IACzB,OAAO,CAAC,OAAO,CAAwB;IACvC,OAAO,CAAC,YAAY,CAAuB;IAC3C,OAAO,CAAC,KAAK,CAAoB;gBAErB,OAAO,EAAE,4BAA4B;cAUjC,YAAY,IAAI,OAAO,CAAC,IAAI,CAAC;cAO7B,UAAU,IAAI,OAAO,CAAC,IAAI,CAAC;IAE3C,MAAM,IAAI,MAAM;IAIhB,MAAM,CAAC,MAAM,EAAE,MAAM,GAAG,IAAI;IAQ5B,WAAW,IAAI,OAAO;IAItB,WAAW,CAAC,YAAY,EAAE,MAAM,GAAG,MAAM;IAIzC,SAAS,CAAC,GAAG,EAAE,MAAM,GAAG,MAAM,GAAG,SAAS;IAI1C,UAAU,IAAI,MAAM,CAAC,MAAM,EAAE,MAAM,CAAC;IAIpC,eAAe,IAAI,mBAAmB,EAAE;IAIxC,aAAa,CAAC,GAAG,EAAE,mBAAmB,GAAG,OAAO;IAI1C,cAAc,CACnB,OAAO,EAAE,MAAM,EACf,IAAI,GAAE,MAAM,EAAO,EACnB,OAAO,CAAC,EAAE,cAAc,GACtB,OAAO,CAAC,aAAa,CAAC;IAuDzB,QAAQ,IAAI,2BAA2B;CAWvC"}
1
+ {"version":3,"file":"local.d.ts","sourceRoot":"","sources":["../../src/execution/local.ts"],"names":[],"mappings":"AAGA,OAAO,KAAK,EAAE,2BAA2B,EAAE,MAAM,6BAA6B,CAAA;AAC9E,OAAO,KAAK,EACX,eAAe,EACf,cAAc,EACd,aAAa,EACb,mBAAmB,EACnB,oBAAoB,EACpB,MAAM,6BAA6B,CAAA;AACpC,OAAO,EAAE,oBAAoB,EAAE,MAAM,WAAW,CAAA;AAEhD,MAAM,WAAW,4BAA4B;IAC5C,EAAE,EAAE,MAAM,CAAA;IACV,GAAG,EAAE,MAAM,CAAA;IACX,QAAQ,CAAC,EAAE,OAAO,CAAA;IAClB,OAAO,CAAC,EAAE,MAAM,CAAC,MAAM,EAAE,MAAM,CAAC,CAAA;IAChC,YAAY,CAAC,EAAE,mBAAmB,EAAE,CAAA;IACpC,KAAK,CAAC,EAAE,MAAM,CAAA;CACd;AAED,qBAAa,qBAAsB,SAAQ,oBAAqB,YAAW,eAAe;IACzF,QAAQ,CAAC,EAAE,EAAE,MAAM,CAAA;IACnB,QAAQ,CAAC,WAAW,EAAE,oBAAoB,CAAU;IAEpD,OAAO,CAAC,GAAG,CAAQ;IACnB,OAAO,CAAC,QAAQ,CAAS;IACzB,OAAO,CAAC,OAAO,CAAwB;IACvC,OAAO,CAAC,YAAY,CAAuB;IAC3C,OAAO,CAAC,KAAK,CAAoB;gBAErB,OAAO,EAAE,4BAA4B;cAUjC,YAAY,IAAI,OAAO,CAAC,IAAI,CAAC;cAO7B,UAAU,IAAI,OAAO,CAAC,IAAI,CAAC;IAE3C,MAAM,IAAI,MAAM;IAIhB,MAAM,CAAC,MAAM,EAAE,MAAM,GAAG,IAAI;IAQ5B,WAAW,IAAI,OAAO;IAItB,WAAW,CAAC,YAAY,EAAE,MAAM,GAAG,MAAM;IAIzC,SAAS,CAAC,GAAG,EAAE,MAAM,GAAG,MAAM,GAAG,SAAS;IAI1C,UAAU,IAAI,MAAM,CAAC,MAAM,EAAE,MAAM,CAAC;IAIpC,eAAe,IAAI,mBAAmB,EAAE;IAIxC,aAAa,CAAC,GAAG,EAAE,mBAAmB,GAAG,OAAO;IAI1C,cAAc,CACnB,OAAO,EAAE,MAAM,EACf,IAAI,GAAE,MAAM,EAAO,EACnB,OAAO,CAAC,EAAE,cAAc,GACtB,OAAO,CAAC,aAAa,CAAC;IA4DzB,QAAQ,IAAI,2BAA2B;CAWvC"}
@@ -66,7 +66,12 @@ export class LocalExecutionContext extends BaseExecutionContext {
66
66
  const cwd = options?.cwd ? resolve(this.cwd, options.cwd) : this.cwd;
67
67
  const env = { ...process.env, ...this.envVars, ...options?.env };
68
68
  const timeoutMs = options?.timeoutMs ?? 30_000;
69
- const shell = options?.shell ?? this.shell ?? true;
69
+ // node's spawn() only keeps argv separation when it execs the binary directly;
70
+ // with shell enabled it re-joins command+args into one unescaped string and hands
71
+ // it to sh/cmd.exe, so any caller-supplied arg becomes shell syntax rather than a
72
+ // literal value. Since `command`/`args` here are whatever a tool call passed in,
73
+ // shell interpretation must be an explicit opt-in, never the silent default.
74
+ const shell = options?.shell ?? this.shell ?? false;
70
75
  return new Promise((resolvePromise) => {
71
76
  const start = performance.now();
72
77
  const proc = spawn(command, args, {
@@ -1 +1 @@
1
- {"version":3,"file":"local.js","sourceRoot":"","sources":["../../src/execution/local.ts"],"names":[],"mappings":"AAAA,OAAO,EAAE,KAAK,EAAE,MAAM,oBAAoB,CAAA;AAC1C,OAAO,EAAE,UAAU,EAAE,MAAM,SAAS,CAAA;AACpC,OAAO,EAAE,OAAO,EAAE,MAAM,WAAW,CAAA;AASnC,OAAO,EAAE,oBAAoB,EAAE,MAAM,WAAW,CAAA;AAWhD,MAAM,OAAO,qBAAsB,SAAQ,oBAAoB;IACrD,EAAE,CAAQ;IACV,WAAW,GAAyB,OAAO,CAAA;IAE5C,GAAG,CAAQ;IACX,QAAQ,CAAS;IACjB,OAAO,CAAwB;IAC/B,YAAY,CAAuB;IACnC,KAAK,CAAoB;IAEjC,YAAY,OAAqC;QAChD,KAAK,EAAE,CAAA;QACP,IAAI,CAAC,EAAE,GAAG,OAAO,CAAC,EAAE,CAAA;QACpB,IAAI,CAAC,GAAG,GAAG,OAAO,CAAC,OAAO,CAAC,GAAG,CAAC,CAAA;QAC/B,IAAI,CAAC,QAAQ,GAAG,OAAO,CAAC,QAAQ,IAAI,IAAI,CAAA;QACxC,IAAI,CAAC,OAAO,GAAG,OAAO,CAAC,OAAO,IAAI,EAAE,CAAA;QACpC,IAAI,CAAC,YAAY,GAAG,OAAO,CAAC,YAAY,IAAI,CAAC,YAAY,EAAE,SAAS,EAAE,OAAO,CAAC,CAAA;QAC9E,IAAI,CAAC,KAAK,GAAG,OAAO,CAAC,KAAK,CAAA;IAC3B,CAAC;IAES,KAAK,CAAC,YAAY;QAC3B,IAAI,CAAC,UAAU,CAAC,IAAI,CAAC,GAAG,CAAC,EAAE,CAAC;YAC3B,MAAM,IAAI,KAAK,CAAC,qCAAqC,IAAI,CAAC,GAAG,EAAE,CAAC,CAAA;QACjE,CAAC;QACD,IAAI,CAAC,GAAG,CAAC,IAAI,CAAC,gCAAgC,IAAI,CAAC,GAAG,EAAE,CAAC,CAAA;IAC1D,CAAC;IAES,KAAK,CAAC,UAAU,KAAmB,CAAC;IAE9C,MAAM;QACL,OAAO,IAAI,CAAC,GAAG,CAAA;IAChB,CAAC;IAED,MAAM,CAAC,MAAc;QACpB,MAAM,QAAQ,GAAG,OAAO,CAAC,MAAM,CAAC,CAAA;QAChC,IAAI,CAAC,UAAU,CAAC,QAAQ,CAAC,EAAE,CAAC;YAC3B,MAAM,IAAI,KAAK,CAAC,qCAAqC,QAAQ,EAAE,CAAC,CAAA;QACjE,CAAC;QACD,IAAI,CAAC,GAAG,GAAG,QAAQ,CAAA;IACpB,CAAC;IAED,WAAW;QACV,OAAO,IAAI,CAAC,QAAQ,CAAA;IACrB,CAAC;IAED,WAAW,CAAC,YAAoB;QAC/B,OAAO,OAAO,CAAC,IAAI,CAAC,GAAG,EAAE,YAAY,CAAC,CAAA;IACvC,CAAC;IAED,SAAS,CAAC,GAAW;QACpB,OAAO,IAAI,CAAC,OAAO,CAAC,GAAG,CAAC,IAAI,OAAO,CAAC,GAAG,CAAC,GAAG,CAAC,CAAA;IAC7C,CAAC;IAED,UAAU;QACT,OAAO,EAAE,GAAG,IAAI,CAAC,OAAO,EAAE,CAAA;IAC3B,CAAC;IAED,eAAe;QACd,OAAO,CAAC,GAAG,IAAI,CAAC,YAAY,CAAC,CAAA;IAC9B,CAAC;IAED,aAAa,CAAC,GAAwB;QACrC,OAAO,IAAI,CAAC,YAAY,CAAC,QAAQ,CAAC,GAAG,CAAC,CAAA;IACvC,CAAC;IAED,KAAK,CAAC,cAAc,CACnB,OAAe,EACf,OAAiB,EAAE,EACnB,OAAwB;QAExB,IAAI,CAAC,IAAI,CAAC,aAAa,CAAC,SAAS,CAAC,IAAI,CAAC,IAAI,CAAC,aAAa,CAAC,OAAO,CAAC,EAAE,CAAC;YACpE,OAAO;gBACN,QAAQ,EAAE,CAAC;gBACX,MAAM,EAAE,EAAE;gBACV,MAAM,EAAE,yEAAyE;gBACjF,UAAU,EAAE,CAAC;aACb,CAAA;QACF,CAAC;QAED,MAAM,GAAG,GAAG,OAAO,EAAE,GAAG,CAAC,CAAC,CAAC,OAAO,CAAC,IAAI,CAAC,GAAG,EAAE,OAAO,CAAC,GAAG,CAAC,CAAC,CAAC,CAAC,IAAI,CAAC,GAAG,CAAA;QACpE,MAAM,GAAG,GAAG,EAAE,GAAG,OAAO,CAAC,GAAG,EAAE,GAAG,IAAI,CAAC,OAAO,EAAE,GAAG,OAAO,EAAE,GAAG,EAAE,CAAA;QAChE,MAAM,SAAS,GAAG,OAAO,EAAE,SAAS,IAAI,MAAM,CAAA;QAC9C,MAAM,KAAK,GAAG,OAAO,EAAE,KAAK,IAAI,IAAI,CAAC,KAAK,IAAI,IAAI,CAAA;QAElD,OAAO,IAAI,OAAO,CAAgB,CAAC,cAAc,EAAE,EAAE;YACpD,MAAM,KAAK,GAAG,WAAW,CAAC,GAAG,EAAE,CAAA;YAC/B,MAAM,IAAI,GAAG,KAAK,CAAC,OAAO,EAAE,IAAI,EAAE;gBACjC,GAAG;gBACH,GAAG;gBACH,KAAK;gBACL,OAAO,EAAE,SAAS;aAClB,CAAC,CAAA;YAEF,IAAI,MAAM,GAAG,EAAE,CAAA;YACf,IAAI,MAAM,GAAG,EAAE,CAAA;YAEf,IAAI,CAAC,MAAM,EAAE,EAAE,CAAC,MAAM,EAAE,CAAC,IAAY,EAAE,EAAE;gBACxC,MAAM,IAAI,IAAI,CAAC,QAAQ,EAAE,CAAA;YAC1B,CAAC,CAAC,CAAA;YAEF,IAAI,CAAC,MAAM,EAAE,EAAE,CAAC,MAAM,EAAE,CAAC,IAAY,EAAE,EAAE;gBACxC,MAAM,IAAI,IAAI,CAAC,QAAQ,EAAE,CAAA;YAC1B,CAAC,CAAC,CAAA;YAEF,IAAI,CAAC,EAAE,CAAC,OAAO,EAAE,CAAC,IAAI,EAAE,EAAE;gBACzB,cAAc,CAAC;oBACd,QAAQ,EAAE,IAAI,IAAI,CAAC;oBACnB,MAAM;oBACN,MAAM;oBACN,UAAU,EAAE,IAAI,CAAC,KAAK,CAAC,WAAW,CAAC,GAAG,EAAE,GAAG,KAAK,CAAC;iBACjD,CAAC,CAAA;YACH,CAAC,CAAC,CAAA;YAEF,IAAI,CAAC,EAAE,CAAC,OAAO,EAAE,CAAC,GAAG,EAAE,EAAE;gBACxB,cAAc,CAAC;oBACd,QAAQ,EAAE,CAAC;oBACX,MAAM;oBACN,MAAM,EAAE,GAAG,CAAC,OAAO;oBACnB,UAAU,EAAE,IAAI,CAAC,KAAK,CAAC,WAAW,CAAC,GAAG,EAAE,GAAG,KAAK,CAAC;iBACjD,CAAC,CAAA;YACH,CAAC,CAAC,CAAA;QACH,CAAC,CAAC,CAAA;IACH,CAAC;IAED,QAAQ;QACP,OAAO;YACN,EAAE,EAAE,IAAI,CAAC,EAAE;YACX,WAAW,EAAE,OAAO;YACpB,GAAG,EAAE,IAAI,CAAC,GAAG;YACb,QAAQ,EAAE,IAAI,CAAC,QAAQ;YACvB,OAAO,EAAE,IAAI,CAAC,OAAO;YACrB,YAAY,EAAE,IAAI,CAAC,YAAY;YAC/B,KAAK,EAAE,IAAI,CAAC,KAAK;SACjB,CAAA;IACF,CAAC;CACD"}
1
+ {"version":3,"file":"local.js","sourceRoot":"","sources":["../../src/execution/local.ts"],"names":[],"mappings":"AAAA,OAAO,EAAE,KAAK,EAAE,MAAM,oBAAoB,CAAA;AAC1C,OAAO,EAAE,UAAU,EAAE,MAAM,SAAS,CAAA;AACpC,OAAO,EAAE,OAAO,EAAE,MAAM,WAAW,CAAA;AASnC,OAAO,EAAE,oBAAoB,EAAE,MAAM,WAAW,CAAA;AAWhD,MAAM,OAAO,qBAAsB,SAAQ,oBAAoB;IACrD,EAAE,CAAQ;IACV,WAAW,GAAyB,OAAO,CAAA;IAE5C,GAAG,CAAQ;IACX,QAAQ,CAAS;IACjB,OAAO,CAAwB;IAC/B,YAAY,CAAuB;IACnC,KAAK,CAAoB;IAEjC,YAAY,OAAqC;QAChD,KAAK,EAAE,CAAA;QACP,IAAI,CAAC,EAAE,GAAG,OAAO,CAAC,EAAE,CAAA;QACpB,IAAI,CAAC,GAAG,GAAG,OAAO,CAAC,OAAO,CAAC,GAAG,CAAC,CAAA;QAC/B,IAAI,CAAC,QAAQ,GAAG,OAAO,CAAC,QAAQ,IAAI,IAAI,CAAA;QACxC,IAAI,CAAC,OAAO,GAAG,OAAO,CAAC,OAAO,IAAI,EAAE,CAAA;QACpC,IAAI,CAAC,YAAY,GAAG,OAAO,CAAC,YAAY,IAAI,CAAC,YAAY,EAAE,SAAS,EAAE,OAAO,CAAC,CAAA;QAC9E,IAAI,CAAC,KAAK,GAAG,OAAO,CAAC,KAAK,CAAA;IAC3B,CAAC;IAES,KAAK,CAAC,YAAY;QAC3B,IAAI,CAAC,UAAU,CAAC,IAAI,CAAC,GAAG,CAAC,EAAE,CAAC;YAC3B,MAAM,IAAI,KAAK,CAAC,qCAAqC,IAAI,CAAC,GAAG,EAAE,CAAC,CAAA;QACjE,CAAC;QACD,IAAI,CAAC,GAAG,CAAC,IAAI,CAAC,gCAAgC,IAAI,CAAC,GAAG,EAAE,CAAC,CAAA;IAC1D,CAAC;IAES,KAAK,CAAC,UAAU,KAAmB,CAAC;IAE9C,MAAM;QACL,OAAO,IAAI,CAAC,GAAG,CAAA;IAChB,CAAC;IAED,MAAM,CAAC,MAAc;QACpB,MAAM,QAAQ,GAAG,OAAO,CAAC,MAAM,CAAC,CAAA;QAChC,IAAI,CAAC,UAAU,CAAC,QAAQ,CAAC,EAAE,CAAC;YAC3B,MAAM,IAAI,KAAK,CAAC,qCAAqC,QAAQ,EAAE,CAAC,CAAA;QACjE,CAAC;QACD,IAAI,CAAC,GAAG,GAAG,QAAQ,CAAA;IACpB,CAAC;IAED,WAAW;QACV,OAAO,IAAI,CAAC,QAAQ,CAAA;IACrB,CAAC;IAED,WAAW,CAAC,YAAoB;QAC/B,OAAO,OAAO,CAAC,IAAI,CAAC,GAAG,EAAE,YAAY,CAAC,CAAA;IACvC,CAAC;IAED,SAAS,CAAC,GAAW;QACpB,OAAO,IAAI,CAAC,OAAO,CAAC,GAAG,CAAC,IAAI,OAAO,CAAC,GAAG,CAAC,GAAG,CAAC,CAAA;IAC7C,CAAC;IAED,UAAU;QACT,OAAO,EAAE,GAAG,IAAI,CAAC,OAAO,EAAE,CAAA;IAC3B,CAAC;IAED,eAAe;QACd,OAAO,CAAC,GAAG,IAAI,CAAC,YAAY,CAAC,CAAA;IAC9B,CAAC;IAED,aAAa,CAAC,GAAwB;QACrC,OAAO,IAAI,CAAC,YAAY,CAAC,QAAQ,CAAC,GAAG,CAAC,CAAA;IACvC,CAAC;IAED,KAAK,CAAC,cAAc,CACnB,OAAe,EACf,OAAiB,EAAE,EACnB,OAAwB;QAExB,IAAI,CAAC,IAAI,CAAC,aAAa,CAAC,SAAS,CAAC,IAAI,CAAC,IAAI,CAAC,aAAa,CAAC,OAAO,CAAC,EAAE,CAAC;YACpE,OAAO;gBACN,QAAQ,EAAE,CAAC;gBACX,MAAM,EAAE,EAAE;gBACV,MAAM,EAAE,yEAAyE;gBACjF,UAAU,EAAE,CAAC;aACb,CAAA;QACF,CAAC;QAED,MAAM,GAAG,GAAG,OAAO,EAAE,GAAG,CAAC,CAAC,CAAC,OAAO,CAAC,IAAI,CAAC,GAAG,EAAE,OAAO,CAAC,GAAG,CAAC,CAAC,CAAC,CAAC,IAAI,CAAC,GAAG,CAAA;QACpE,MAAM,GAAG,GAAG,EAAE,GAAG,OAAO,CAAC,GAAG,EAAE,GAAG,IAAI,CAAC,OAAO,EAAE,GAAG,OAAO,EAAE,GAAG,EAAE,CAAA;QAChE,MAAM,SAAS,GAAG,OAAO,EAAE,SAAS,IAAI,MAAM,CAAA;QAC9C,+EAA+E;QAC/E,kFAAkF;QAClF,kFAAkF;QAClF,iFAAiF;QACjF,6EAA6E;QAC7E,MAAM,KAAK,GAAG,OAAO,EAAE,KAAK,IAAI,IAAI,CAAC,KAAK,IAAI,KAAK,CAAA;QAEnD,OAAO,IAAI,OAAO,CAAgB,CAAC,cAAc,EAAE,EAAE;YACpD,MAAM,KAAK,GAAG,WAAW,CAAC,GAAG,EAAE,CAAA;YAC/B,MAAM,IAAI,GAAG,KAAK,CAAC,OAAO,EAAE,IAAI,EAAE;gBACjC,GAAG;gBACH,GAAG;gBACH,KAAK;gBACL,OAAO,EAAE,SAAS;aAClB,CAAC,CAAA;YAEF,IAAI,MAAM,GAAG,EAAE,CAAA;YACf,IAAI,MAAM,GAAG,EAAE,CAAA;YAEf,IAAI,CAAC,MAAM,EAAE,EAAE,CAAC,MAAM,EAAE,CAAC,IAAY,EAAE,EAAE;gBACxC,MAAM,IAAI,IAAI,CAAC,QAAQ,EAAE,CAAA;YAC1B,CAAC,CAAC,CAAA;YAEF,IAAI,CAAC,MAAM,EAAE,EAAE,CAAC,MAAM,EAAE,CAAC,IAAY,EAAE,EAAE;gBACxC,MAAM,IAAI,IAAI,CAAC,QAAQ,EAAE,CAAA;YAC1B,CAAC,CAAC,CAAA;YAEF,IAAI,CAAC,EAAE,CAAC,OAAO,EAAE,CAAC,IAAI,EAAE,EAAE;gBACzB,cAAc,CAAC;oBACd,QAAQ,EAAE,IAAI,IAAI,CAAC;oBACnB,MAAM;oBACN,MAAM;oBACN,UAAU,EAAE,IAAI,CAAC,KAAK,CAAC,WAAW,CAAC,GAAG,EAAE,GAAG,KAAK,CAAC;iBACjD,CAAC,CAAA;YACH,CAAC,CAAC,CAAA;YAEF,IAAI,CAAC,EAAE,CAAC,OAAO,EAAE,CAAC,GAAG,EAAE,EAAE;gBACxB,cAAc,CAAC;oBACd,QAAQ,EAAE,CAAC;oBACX,MAAM;oBACN,MAAM,EAAE,GAAG,CAAC,OAAO;oBACnB,UAAU,EAAE,IAAI,CAAC,KAAK,CAAC,WAAW,CAAC,GAAG,EAAE,GAAG,KAAK,CAAC;iBACjD,CAAC,CAAA;YACH,CAAC,CAAC,CAAA;QACH,CAAC,CAAC,CAAA;IACH,CAAC;IAED,QAAQ;QACP,OAAO;YACN,EAAE,EAAE,IAAI,CAAC,EAAE;YACX,WAAW,EAAE,OAAO;YACpB,GAAG,EAAE,IAAI,CAAC,GAAG;YACb,QAAQ,EAAE,IAAI,CAAC,QAAQ;YACvB,OAAO,EAAE,IAAI,CAAC,OAAO;YACrB,YAAY,EAAE,IAAI,CAAC,YAAY;YAC/B,KAAK,EAAE,IAAI,CAAC,KAAK;SACjB,CAAA;IACF,CAAC;CACD"}
@@ -0,0 +1,2 @@
1
+ export {};
2
+ //# sourceMappingURL=local.test.d.ts.map
@@ -0,0 +1 @@
1
+ {"version":3,"file":"local.test.d.ts","sourceRoot":"","sources":["../../src/execution/local.test.ts"],"names":[],"mappings":""}