niceeval 0.6.1 → 0.7.0

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 (378) hide show
  1. package/INDEX.md +23 -23
  2. package/README.zh.md +6 -6
  3. package/dist/agents/types.d.ts +69 -7
  4. package/dist/context/types.d.ts +32 -12
  5. package/dist/i18n/en.d.ts +54 -0
  6. package/dist/i18n/zh-CN.d.ts +57 -3
  7. package/dist/o11y/types.d.ts +16 -2
  8. package/dist/report/aggregate.d.ts +32 -24
  9. package/dist/report/aggregate.js +158 -50
  10. package/dist/report/built-in/index.d.ts +2 -0
  11. package/dist/report/built-in/index.js +8 -0
  12. package/dist/report/components.d.ts +93 -160
  13. package/dist/report/components.js +377 -114
  14. package/dist/report/compute.d.ts +87 -81
  15. package/dist/report/compute.js +597 -417
  16. package/dist/report/flag.d.ts +32 -6
  17. package/dist/report/flag.js +92 -4
  18. package/dist/report/format.d.ts +19 -11
  19. package/dist/report/format.js +30 -13
  20. package/dist/report/index.d.ts +16 -16
  21. package/dist/report/index.js +20 -21
  22. package/dist/report/load.js +3 -2
  23. package/dist/report/locale.d.ts +57 -33
  24. package/dist/report/locale.js +122 -56
  25. package/dist/report/metrics.d.ts +23 -4
  26. package/dist/report/metrics.js +110 -25
  27. package/dist/report/primitives.d.ts +48 -15
  28. package/dist/report/primitives.js +135 -26
  29. package/dist/report/react/AttemptList.d.ts +9 -7
  30. package/dist/report/react/AttemptList.js +17 -10
  31. package/dist/report/react/DeltaTable.js +19 -18
  32. package/dist/report/react/EvalList.d.ts +4 -4
  33. package/dist/report/react/EvalList.js +0 -0
  34. package/dist/report/react/ExperimentComparison.d.ts +10 -0
  35. package/dist/report/react/ExperimentComparison.js +12 -0
  36. package/dist/report/react/ExperimentList.d.ts +4 -3
  37. package/dist/report/react/ExperimentList.js +17 -18
  38. package/dist/report/react/MetricBars.js +5 -4
  39. package/dist/report/react/MetricLine.js +12 -5
  40. package/dist/report/react/MetricMatrix.js +1 -1
  41. package/dist/report/react/MetricScatter.js +59 -28
  42. package/dist/report/react/MetricTable.js +2 -12
  43. package/dist/report/react/ScopeSummary.d.ts +10 -0
  44. package/dist/report/react/ScopeSummary.js +17 -0
  45. package/dist/report/react/Scoreboard.js +6 -6
  46. package/dist/report/react/cell.js +2 -2
  47. package/dist/report/react/chart-math.d.ts +23 -6
  48. package/dist/report/react/chart-math.js +71 -19
  49. package/dist/report/react/fixtures.d.ts +5 -9
  50. package/dist/report/react/fixtures.js +110 -147
  51. package/dist/report/react/index.d.ts +15 -5
  52. package/dist/report/react/index.js +18 -7
  53. package/dist/report/report.d.ts +137 -16
  54. package/dist/report/report.js +259 -28
  55. package/dist/report/text/faces.d.ts +17 -19
  56. package/dist/report/text/faces.js +253 -184
  57. package/dist/report/text/plot.js +1 -1
  58. package/dist/report/text/table.js +38 -7
  59. package/dist/report/tree.d.ts +90 -40
  60. package/dist/report/tree.js +252 -94
  61. package/dist/report/types.d.ts +247 -284
  62. package/dist/report/types.js +4 -3
  63. package/dist/report/web.d.ts +21 -5
  64. package/dist/report/web.js +42 -16
  65. package/dist/results/select.d.ts +38 -16
  66. package/dist/results/select.js +73 -25
  67. package/dist/results/types.d.ts +49 -14
  68. package/dist/runner/feedback/sink.d.ts +110 -0
  69. package/dist/runner/types.d.ts +513 -22
  70. package/dist/sandbox/docker.d.ts +23 -2
  71. package/dist/sandbox/e2b.d.ts +15 -1
  72. package/dist/sandbox/errors.d.ts +30 -3
  73. package/dist/sandbox/io-retry.d.ts +17 -0
  74. package/dist/sandbox/registry.d.ts +2 -0
  75. package/dist/sandbox/resolve.d.ts +18 -5
  76. package/dist/sandbox/retry.d.ts +11 -1
  77. package/dist/sandbox/types.d.ts +39 -5
  78. package/dist/sandbox/vercel.d.ts +7 -1
  79. package/dist/scoring/coverage.d.ts +30 -0
  80. package/dist/scoring/display.d.ts +21 -0
  81. package/dist/scoring/display.js +120 -0
  82. package/dist/scoring/types.d.ts +103 -20
  83. package/dist/shared/aggregate.d.ts +4 -2
  84. package/dist/shared/aggregate.js +8 -7
  85. package/dist/shared/types.d.ts +28 -0
  86. package/dist/tty-line.d.ts +0 -4
  87. package/dist/util.d.ts +23 -0
  88. package/docs-site/zh/README.md +44 -0
  89. package/docs-site/zh/examples/ai-agent-application.mdx +63 -0
  90. package/docs-site/zh/examples/coding-agent-extensions.mdx +57 -0
  91. package/docs-site/zh/examples/index.mdx +50 -0
  92. package/docs-site/zh/{concepts → explanation}/adapter.mdx +31 -13
  93. package/docs-site/zh/{concepts → explanation}/assert.mdx +7 -7
  94. package/docs-site/zh/{concepts → explanation}/drive.mdx +8 -8
  95. package/docs-site/zh/{concepts → explanation}/evals.mdx +4 -4
  96. package/docs-site/zh/{concepts → explanation}/experiment.mdx +8 -8
  97. package/docs-site/zh/{concepts → explanation}/hitl.mdx +8 -8
  98. package/docs-site/zh/{concepts → explanation}/judge.mdx +5 -5
  99. package/docs-site/zh/{concepts → explanation}/overview.mdx +11 -11
  100. package/docs-site/zh/{guides → explanation}/runner.mdx +18 -8
  101. package/docs-site/zh/{concepts → explanation}/tier.mdx +6 -6
  102. package/docs-site/zh/{guides → how-to}/agent-feedback-loop.mdx +35 -33
  103. package/docs-site/zh/{guides → how-to}/authoring.mdx +35 -2
  104. package/docs-site/zh/{guides → how-to}/ci-integration.mdx +23 -12
  105. package/docs-site/zh/{guides → how-to}/connect-otel.mdx +6 -6
  106. package/docs-site/zh/{guides → how-to}/connect-your-agent.mdx +47 -21
  107. package/docs-site/zh/{guides → how-to}/custom-reports.mdx +34 -39
  108. package/docs-site/zh/{guides → how-to}/dataset-fanout.mdx +25 -3
  109. package/docs-site/zh/{guides → how-to}/experiments.mdx +12 -5
  110. package/docs-site/zh/how-to/publish-report.mdx +105 -0
  111. package/docs-site/zh/{guides → how-to}/reporters.mdx +2 -2
  112. package/docs-site/zh/{guides → how-to}/sandbox-agent.mdx +56 -7
  113. package/docs-site/zh/how-to/sandbox-providers.mdx +350 -0
  114. package/docs-site/zh/{guides → how-to}/scoring-guide.mdx +4 -4
  115. package/docs-site/zh/{guides → how-to}/viewing-results.mdx +82 -39
  116. package/docs-site/zh/{guides → how-to}/write-experiment.mdx +6 -4
  117. package/docs-site/zh/{guides → how-to}/write-send.mdx +30 -14
  118. package/docs-site/zh/index.mdx +24 -26
  119. package/docs-site/zh/introduction.mdx +8 -8
  120. package/docs-site/zh/reference/builtin-agents.mdx +32 -5
  121. package/docs-site/zh/reference/capabilities.mdx +8 -8
  122. package/docs-site/zh/reference/cli.mdx +40 -12
  123. package/docs-site/zh/reference/define-agent.mdx +58 -5
  124. package/docs-site/zh/reference/define-config.mdx +1 -1
  125. package/docs-site/zh/reference/define-eval.mdx +42 -9
  126. package/docs-site/zh/reference/events.mdx +3 -3
  127. package/docs-site/zh/reference/expect.mdx +26 -1
  128. package/docs-site/zh/{guides → reference}/official-adapters.mdx +32 -8
  129. package/docs-site/zh/{guides → reference}/report-components.mdx +45 -33
  130. package/docs-site/zh/{guides → reference}/results-data.mdx +21 -13
  131. package/docs-site/zh/troubleshooting/debug-sandbox.mdx +57 -0
  132. package/docs-site/zh/troubleshooting/debugging.mdx +212 -0
  133. package/docs-site/zh/{quickstart.mdx → tutorials/quickstart.mdx} +5 -17
  134. package/package.json +10 -2
  135. package/src/agents/ai-sdk-otel.test.ts +1 -0
  136. package/src/agents/ai-sdk.test.ts +3 -0
  137. package/src/agents/ai-sdk.ts +3 -0
  138. package/src/agents/bub-install-spec.test.ts +34 -0
  139. package/src/agents/bub-install-spec.ts +32 -0
  140. package/src/agents/bub.ts +31 -32
  141. package/src/agents/claude-code.test.ts +130 -9
  142. package/src/agents/claude-code.ts +76 -4
  143. package/src/agents/codex.test.ts +189 -40
  144. package/src/agents/codex.ts +155 -14
  145. package/src/agents/coding-cli-versions.test.ts +15 -0
  146. package/src/agents/coding-cli-versions.ts +3 -0
  147. package/src/agents/index.ts +13 -2
  148. package/src/agents/langgraph.test.ts +204 -0
  149. package/src/agents/langgraph.ts +495 -0
  150. package/src/agents/marketplace.ts +85 -0
  151. package/src/agents/native-config.test.ts +179 -0
  152. package/src/agents/native-config.ts +267 -0
  153. package/src/agents/openai-compat.test.ts +1 -0
  154. package/src/agents/openai-compat.ts +1 -1
  155. package/src/agents/openclaw.test.ts +31 -0
  156. package/src/agents/openclaw.ts +171 -0
  157. package/src/agents/plugin-config.test.ts +1 -0
  158. package/src/agents/sdk-streams.test.ts +79 -0
  159. package/src/agents/sdk-streams.ts +55 -10
  160. package/src/agents/skills.test.ts +1 -0
  161. package/src/agents/streaming.test.ts +3 -9
  162. package/src/agents/streaming.ts +2 -2
  163. package/src/agents/types.ts +71 -8
  164. package/src/agents/ui-message-stream.test.ts +3 -0
  165. package/src/cli.ts +446 -124
  166. package/src/context/context.test.ts +51 -12
  167. package/src/context/context.ts +162 -30
  168. package/src/context/session.test.ts +2 -1
  169. package/src/context/session.ts +115 -7
  170. package/src/context/types.ts +30 -12
  171. package/src/define.test.ts +13 -8
  172. package/src/define.ts +25 -4
  173. package/src/expect/index.ts +53 -23
  174. package/src/i18n/en.ts +81 -17
  175. package/src/i18n/zh-CN.ts +80 -17
  176. package/src/o11y/cost.test.ts +1 -0
  177. package/src/o11y/execution-tree.test.ts +1 -20
  178. package/src/o11y/otlp/mappers/claude-code.test.ts +1 -0
  179. package/src/o11y/otlp/parse.test.ts +1 -0
  180. package/src/o11y/otlp/turn-otel.test.ts +1 -0
  181. package/src/o11y/parsers/bub.test.ts +1 -0
  182. package/src/o11y/parsers/claude-code.test.ts +1 -34
  183. package/src/o11y/parsers/openclaw.test.ts +154 -0
  184. package/src/o11y/parsers/openclaw.ts +310 -0
  185. package/src/o11y/prices.json +746 -311
  186. package/src/o11y/tool-names.test.ts +1 -0
  187. package/src/o11y/types.ts +16 -2
  188. package/src/report/aggregate.ts +178 -61
  189. package/src/report/built-in/index.tsx +9 -0
  190. package/src/report/components.tsx +625 -279
  191. package/src/report/compute.ts +723 -491
  192. package/src/report/dual-render.test.tsx +741 -1024
  193. package/src/report/flag.ts +104 -12
  194. package/src/report/format.ts +32 -12
  195. package/src/report/index.ts +119 -46
  196. package/src/report/load.ts +3 -2
  197. package/src/report/locale.ts +136 -65
  198. package/src/report/metrics.ts +108 -25
  199. package/src/report/primitives.tsx +196 -45
  200. package/src/report/react/AttemptList.tsx +30 -43
  201. package/src/report/react/DeltaTable.tsx +63 -45
  202. package/src/report/react/EvalList.tsx +0 -0
  203. package/src/report/react/ExperimentComparison.tsx +73 -0
  204. package/src/report/react/ExperimentList.tsx +50 -32
  205. package/src/report/react/MetricBars.tsx +5 -4
  206. package/src/report/react/MetricLine.tsx +13 -8
  207. package/src/report/react/MetricMatrix.tsx +2 -2
  208. package/src/report/react/MetricScatter.tsx +86 -34
  209. package/src/report/react/MetricTable.tsx +4 -76
  210. package/src/report/react/ScopeSummary.tsx +86 -0
  211. package/src/report/react/Scoreboard.tsx +28 -10
  212. package/src/report/react/cell.tsx +2 -2
  213. package/src/report/react/chart-math.test.ts +85 -0
  214. package/src/report/react/chart-math.ts +101 -22
  215. package/src/report/react/enhance.js +89 -5
  216. package/src/report/react/fixtures.ts +114 -154
  217. package/src/report/react/index.tsx +24 -39
  218. package/src/report/react/render.test.tsx +138 -158
  219. package/src/report/react/styles.css +243 -82
  220. package/src/report/report.test.ts +779 -841
  221. package/src/report/report.ts +423 -41
  222. package/src/report/text/faces.ts +290 -193
  223. package/src/report/text/plot.ts +1 -1
  224. package/src/report/text/table.ts +44 -7
  225. package/src/report/tree.ts +362 -104
  226. package/src/report/types.ts +261 -271
  227. package/src/report/web.ts +63 -20
  228. package/src/results/annotated-source.test.ts +62 -9
  229. package/src/results/annotated-source.ts +64 -6
  230. package/src/results/attempt-evidence.test.ts +13 -11
  231. package/src/results/attempt-evidence.ts +20 -13
  232. package/src/results/attempt-source.ts +6 -3
  233. package/src/results/copy.ts +150 -60
  234. package/src/results/host-equivalence.test.ts +34 -20
  235. package/src/results/index.ts +12 -4
  236. package/src/results/locator.test.ts +1 -22
  237. package/src/results/open.ts +15 -5
  238. package/src/results/publish.ts +149 -0
  239. package/src/results/results.test.ts +89 -54
  240. package/src/results/select.ts +104 -34
  241. package/src/results/truncate.ts +90 -0
  242. package/src/results/types.ts +43 -14
  243. package/src/results/writer.ts +31 -13
  244. package/src/runner/attempt.test.ts +138 -7
  245. package/src/runner/attempt.ts +603 -104
  246. package/src/runner/discover.test.ts +47 -0
  247. package/src/runner/discover.ts +36 -2
  248. package/src/runner/eval-source.test.ts +1 -27
  249. package/src/runner/feedback/agent.test.ts +504 -0
  250. package/src/runner/feedback/agent.ts +409 -0
  251. package/src/runner/feedback/ci.test.ts +562 -0
  252. package/src/runner/feedback/ci.ts +401 -0
  253. package/src/runner/feedback/coordinator.test.ts +317 -0
  254. package/src/runner/feedback/coordinator.ts +397 -0
  255. package/src/runner/feedback/failure.ts +40 -0
  256. package/src/runner/feedback/human.test.ts +616 -0
  257. package/src/runner/feedback/human.ts +535 -0
  258. package/src/runner/feedback/index.ts +66 -0
  259. package/src/runner/feedback/io.ts +78 -0
  260. package/src/runner/feedback/profile.test.ts +50 -0
  261. package/src/runner/feedback/profile.ts +58 -0
  262. package/src/runner/feedback/reducer.test.ts +395 -0
  263. package/src/runner/feedback/reducer.ts +260 -0
  264. package/src/runner/feedback/renderer.ts +82 -0
  265. package/src/runner/feedback/sink.ts +203 -0
  266. package/src/runner/feedback/testing.ts +106 -0
  267. package/src/runner/ledger.test.ts +230 -0
  268. package/src/runner/ledger.ts +329 -0
  269. package/src/runner/report.test.ts +128 -3
  270. package/src/runner/report.ts +33 -9
  271. package/src/runner/reporters/artifacts.ts +8 -2
  272. package/src/runner/reporters/braintrust.test.ts +8 -7
  273. package/src/runner/reporters/braintrust.ts +9 -2
  274. package/src/runner/reporters/index.ts +2 -2
  275. package/src/runner/reporters/json.test.ts +162 -0
  276. package/src/runner/reporters/json.ts +35 -8
  277. package/src/runner/reporters/shared.ts +1 -5
  278. package/src/runner/run.test.ts +760 -3
  279. package/src/runner/run.ts +243 -37
  280. package/src/runner/sandbox-prep.ts +3 -42
  281. package/src/runner/timing.ts +158 -0
  282. package/src/runner/types.ts +518 -22
  283. package/src/sandbox/checkpoint.test.ts +55 -0
  284. package/src/sandbox/checkpoint.ts +29 -8
  285. package/src/sandbox/cli-commands.ts +407 -0
  286. package/src/sandbox/docker.ts +115 -16
  287. package/src/sandbox/e2b-agent-template.test.ts +56 -0
  288. package/src/sandbox/e2b-agent-template.ts +94 -0
  289. package/src/sandbox/e2b.ts +74 -9
  290. package/src/sandbox/errors.ts +111 -4
  291. package/src/sandbox/index.ts +2 -0
  292. package/src/sandbox/io-retry.test.ts +58 -0
  293. package/src/sandbox/io-retry.ts +45 -0
  294. package/src/sandbox/keep-registry.test.ts +86 -0
  295. package/src/sandbox/keep-registry.ts +142 -0
  296. package/src/sandbox/keep.ts +178 -0
  297. package/src/sandbox/paths.test.ts +1 -0
  298. package/src/sandbox/paths.ts +19 -8
  299. package/src/sandbox/registry.ts +20 -3
  300. package/src/sandbox/resolve.ts +76 -11
  301. package/src/sandbox/retry.test.ts +70 -0
  302. package/src/sandbox/retry.ts +46 -4
  303. package/src/sandbox/types.ts +44 -6
  304. package/src/sandbox/vercel.ts +43 -20
  305. package/src/scoring/collector.ts +60 -17
  306. package/src/scoring/coverage.ts +95 -0
  307. package/src/scoring/diff.ts +81 -0
  308. package/src/scoring/display.test.ts +121 -0
  309. package/src/scoring/display.ts +133 -0
  310. package/src/scoring/evidence.test.ts +189 -0
  311. package/src/scoring/judge.test.ts +142 -0
  312. package/src/scoring/judge.ts +15 -18
  313. package/src/scoring/scoped.ts +217 -50
  314. package/src/scoring/types.ts +117 -20
  315. package/src/scoring/verdict.ts +16 -4
  316. package/src/shared/aggregate.ts +8 -6
  317. package/src/shared/types.ts +31 -0
  318. package/src/show/compose.ts +50 -67
  319. package/src/show/index.ts +127 -56
  320. package/src/show/render.ts +662 -131
  321. package/src/show/report-host.test.ts +188 -0
  322. package/src/show/report-host.ts +375 -0
  323. package/src/show/show.test.ts +320 -54
  324. package/src/tty-line.ts +8 -26
  325. package/src/util.test.ts +1 -0
  326. package/src/util.ts +41 -0
  327. package/src/view/app/App.test.tsx +69 -0
  328. package/src/view/app/App.tsx +144 -48
  329. package/src/view/app/components/AttemptModal.tsx +423 -11
  330. package/src/view/app/components/CodeView.tsx +41 -14
  331. package/src/view/app/components/CopyControls.tsx +2 -2
  332. package/src/view/app/i18n.ts +37 -17
  333. package/src/view/app/lib/attempt-route.test.ts +1 -0
  334. package/src/view/app/lib/verdict.ts +7 -9
  335. package/src/view/app/main.tsx +13 -8
  336. package/src/view/app/pages/{RunsPage.tsx → AttemptsPage.tsx} +6 -6
  337. package/src/view/app/types.ts +4 -1
  338. package/src/view/artifact-serving.test.ts +2 -1
  339. package/src/view/client-dist/app.css +1 -1
  340. package/src/view/client-dist/app.js +17 -17
  341. package/src/view/data.test.ts +10 -3
  342. package/src/view/data.ts +155 -49
  343. package/src/view/index.ts +56 -41
  344. package/src/view/server.ts +37 -15
  345. package/src/view/shared/types.ts +34 -5
  346. package/src/view/styles.css +227 -0
  347. package/src/view/view-report.test.ts +167 -62
  348. package/dist/report/built-ins/experiment-comparison.d.ts +0 -1
  349. package/dist/report/built-ins/experiment-comparison.js +0 -13
  350. package/dist/report/built-ins/index.d.ts +0 -1
  351. package/dist/report/built-ins/index.js +0 -2
  352. package/dist/report/react/GroupSummary.d.ts +0 -8
  353. package/dist/report/react/GroupSummary.js +0 -8
  354. package/dist/report/react/RunOverview.d.ts +0 -8
  355. package/dist/report/react/RunOverview.js +0 -12
  356. package/docs-site/zh/example/ai-agent-application.mdx +0 -152
  357. package/docs-site/zh/example/claude-code-codex-plugin.mdx +0 -167
  358. package/docs-site/zh/example/claude-code-codex-skill.mdx +0 -152
  359. package/docs-site/zh/example/showcase.mdx +0 -39
  360. package/docs-site/zh/guides/publish-report.mdx +0 -91
  361. package/docs-site/zh/guides/sandbox-providers.mdx +0 -102
  362. package/src/report/built-in-user-parity.test.tsx +0 -640
  363. package/src/report/built-ins/experiment-comparison.tsx +0 -19
  364. package/src/report/built-ins/index.ts +0 -2
  365. package/src/report/react/GroupSummary.tsx +0 -66
  366. package/src/report/react/RunOverview.tsx +0 -109
  367. package/src/runner/reporters/console.ts +0 -70
  368. package/src/runner/reporters/live.test.ts +0 -56
  369. package/src/runner/reporters/live.ts +0 -247
  370. package/src/runner/reporters/quiet.test.ts +0 -66
  371. package/src/runner/reporters/quiet.ts +0 -49
  372. package/src/runner/reporters/table.ts +0 -277
  373. /package/docs-site/zh/{example/tier1-ai-sdk-v7.mdx → examples/integrations/ai-sdk-v7.mdx} +0 -0
  374. /package/docs-site/zh/{example/tier1-claude-sdk.mdx → examples/integrations/claude-sdk.mdx} +0 -0
  375. /package/docs-site/zh/{example/tier1-codex-sdk.mdx → examples/integrations/codex-sdk.mdx} +0 -0
  376. /package/docs-site/zh/{example/tier1-langgraph.mdx → examples/integrations/langgraph.mdx} +0 -0
  377. /package/docs-site/zh/{example/tier1-pi-sdk.mdx → examples/integrations/pi-sdk.mdx} +0 -0
  378. /package/docs-site/zh/{guides → how-to}/fixtures.mdx +0 -0
@@ -0,0 +1,158 @@
1
+ // attempt 级阶段计时:LifecyclePhase 主链/收尾段的耗时 + 阶段内时间树(hook / turn / command)。
2
+ // 契约见 docs/feature/results/architecture.md「result.json」的 PhaseTiming / TimingNode:
3
+ // 数组顺序即执行顺序;没执行的阶段不写 0 值条目;收尾段不计入 durationMs 口径但照常计时;
4
+ // 结果封口发生在 Effect Scope release 完成之后(sandbox.stop 也向这个 recorder 写入)。
5
+
6
+ import type { LifecyclePhase, PhaseTiming, TimingNode, TimingNodeKind } from "./types.ts";
7
+
8
+ /** 主链成员(enterPhase 推进;进入下一个即关闭上一个)。收尾段用 measureClosing 单独计时。 */
9
+ const CLOSING_PHASES: ReadonlySet<LifecyclePhase> = new Set([
10
+ "eval.teardown",
11
+ "agent.teardown",
12
+ "sandbox.teardown",
13
+ "sandbox.suspend",
14
+ "sandbox.stop",
15
+ ]);
16
+
17
+ export interface TimingRecorder {
18
+ /** 进入一个主链阶段:关闭上一个开着的主链条目,开一个新条目。 */
19
+ enter(phase: LifecyclePhase): void;
20
+ /** 把当前开着的阶段标记为 failed 并关闭(其后无主链条目)。 */
21
+ failCurrent(): void;
22
+ /** 关闭当前开着的主链条目(正常走完主链时在收尾前调用)。 */
23
+ closeCurrent(): void;
24
+ /** 计时一个收尾段(eval/agent/sandbox teardown、suspend、stop):无论成败都记条目,失败标 failed。 */
25
+ measureClosing<T>(phase: LifecyclePhase, fn: () => Promise<T> | T): Promise<T>;
26
+ /** 往「当前挂载点」挂一个时间树子节点:有 pushParent 的父节点则挂它下面(hook 内的 command),
27
+ * 否则挂当前开着的阶段;都没有时静默丢弃。 */
28
+ child(node: Omit<TimingNode, "id">): TimingNode | undefined;
29
+ /** 在指定的已有节点下挂子节点。 */
30
+ childOf(parent: TimingNode, node: Omit<TimingNode, "id">): TimingNode;
31
+ /** 把后续 child() 的挂载点压到 parent 下(hook 执行期间);与 popParent 成对。 */
32
+ pushParent(parent: TimingNode): void;
33
+ popParent(): void;
34
+ /** 直接补记一个已测好耗时的阶段条目(sandbox.stop 这类 Scope release 段用)。 */
35
+ record(phase: LifecyclePhase, durationMs: number, failed?: boolean): void;
36
+ /** 相对 attempt 单调时钟起点的当前偏移(ms)。 */
37
+ offsetNow(): number;
38
+ /** 封口:关闭残留的开条目,产出 PhaseTiming[](一个阶段都没记录时返回 undefined)。 */
39
+ finalize(): PhaseTiming[] | undefined;
40
+ }
41
+
42
+ interface OpenPhase {
43
+ name: LifecyclePhase;
44
+ startedAt: number;
45
+ children: TimingNode[];
46
+ }
47
+
48
+ export function createTimingRecorder(now: () => number = () => performance.now()): TimingRecorder {
49
+ const origin = now();
50
+ const phases: PhaseTiming[] = [];
51
+ let open: OpenPhase | undefined;
52
+ let nodeSeq = 0;
53
+ const parentStack: TimingNode[] = [];
54
+
55
+ const offset = () => Math.max(0, Math.round(now() - origin));
56
+ const nextId = () => `n${++nodeSeq}`;
57
+
58
+ function close(failed?: true): void {
59
+ if (!open) return;
60
+ const entry: PhaseTiming = {
61
+ name: open.name,
62
+ durationMs: Math.max(0, Math.round(now() - open.startedAt)),
63
+ ...(failed ? { failed: true as const } : {}),
64
+ ...(open.children.length > 0 ? { children: open.children } : {}),
65
+ };
66
+ phases.push(entry);
67
+ open = undefined;
68
+ }
69
+
70
+ return {
71
+ enter(phase) {
72
+ if (CLOSING_PHASES.has(phase)) {
73
+ // 收尾段经 measureClosing 计时;enter 只负责主链推进,这里防御性关掉主链残留。
74
+ close();
75
+ return;
76
+ }
77
+ // agent.run 是唯一的嵌套成员:只作归因值,不在 phases 里单列(不关闭 eval.run)。
78
+ if (phase === "agent.run") return;
79
+ close();
80
+ open = { name: phase, startedAt: now(), children: [] };
81
+ },
82
+ failCurrent() {
83
+ close(true);
84
+ },
85
+ closeCurrent() {
86
+ close();
87
+ },
88
+ async measureClosing(phase, fn) {
89
+ close();
90
+ const startedAt = now();
91
+ const children: TimingNode[] = [];
92
+ open = { name: phase, startedAt, children };
93
+ try {
94
+ const result = await fn();
95
+ close();
96
+ return result;
97
+ } catch (e) {
98
+ close(true);
99
+ throw e;
100
+ }
101
+ },
102
+ child(node) {
103
+ const full: TimingNode = { id: nextId(), ...node };
104
+ const top = parentStack[parentStack.length - 1];
105
+ if (top) {
106
+ (top.children ??= []).push(full);
107
+ return full;
108
+ }
109
+ if (!open) return undefined;
110
+ open.children.push(full);
111
+ return full;
112
+ },
113
+ childOf(parent, node) {
114
+ const full: TimingNode = { id: nextId(), ...node };
115
+ (parent.children ??= []).push(full);
116
+ return full;
117
+ },
118
+ pushParent(parent) {
119
+ parentStack.push(parent);
120
+ },
121
+ popParent() {
122
+ parentStack.pop();
123
+ },
124
+ record(phase, durationMs, failed) {
125
+ close();
126
+ phases.push({ name: phase, durationMs: Math.max(0, Math.round(durationMs)), ...(failed ? { failed: true as const } : {}) });
127
+ },
128
+ offsetNow: offset,
129
+ finalize() {
130
+ close();
131
+ return phases.length > 0 ? phases : undefined;
132
+ },
133
+ };
134
+ }
135
+
136
+ /** 命令的有界脱敏摘要:argv 拼接 + 160 字符截断;env 值与 stdout/stderr 不进入时间树。 */
137
+ export function commandDisplay(cmd: string, args?: readonly string[]): string {
138
+ const s = [cmd, ...(args ?? [])].join(" ");
139
+ return s.length > 160 ? `${s.slice(0, 159)}…` : s;
140
+ }
141
+
142
+ /** kind=command 节点的便捷构造。 */
143
+ export function commandNode(opts: {
144
+ display: string;
145
+ startOffsetMs: number;
146
+ durationMs: number;
147
+ exitCode?: number;
148
+ failed?: boolean;
149
+ }): Omit<TimingNode, "id"> {
150
+ return {
151
+ kind: "command" as TimingNodeKind,
152
+ label: opts.display.split(" ")[0] ?? opts.display,
153
+ startOffsetMs: opts.startOffsetMs,
154
+ durationMs: opts.durationMs,
155
+ ...(opts.failed ? { failed: true as const } : {}),
156
+ command: { display: opts.display, ...(opts.exitCode !== undefined ? { exitCode: opts.exitCode } : {}) },
157
+ };
158
+ }