@danceiny/gotry 0.0.1-rc.9 → 0.2.0-rc.27

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 (507) hide show
  1. package/README.md +172 -219
  2. package/README.zh-CN.md +222 -0
  3. package/bin/gotry-backend.js +53 -0
  4. package/bin/gotry-booking-copilot.js +53 -0
  5. package/bin/gotry-bootstrap.js +1744 -0
  6. package/bin/gotry-inner.js +733 -83
  7. package/bin/gotry-process-liveness.d.ts +27 -0
  8. package/bin/gotry-process-liveness.js +97 -0
  9. package/bin/gotry-runtime-resolution.d.ts +27 -0
  10. package/bin/gotry-runtime-resolution.js +49 -0
  11. package/bin/gotry.js +1 -1
  12. package/client/client.js +203 -0
  13. package/cordis.gotry-patch.yml +80 -56
  14. package/data/airline-airports.json +39 -0
  15. package/data/stations-12306-verify.json +524 -0
  16. package/dist/capabilities/agent-reach-deep.js +45 -122
  17. package/dist/capabilities/agent-reach.js +101 -146
  18. package/dist/capabilities/anything.js +135 -94
  19. package/dist/capabilities/artifacts.js +370 -0
  20. package/dist/capabilities/channel-health.js +131 -0
  21. package/dist/capabilities/channel-registry.js +256 -0
  22. package/dist/capabilities/doctor.js +546 -0
  23. package/dist/capabilities/effect.js +394 -0
  24. package/dist/capabilities/external-event.js +149 -0
  25. package/dist/capabilities/fact-log.js +55 -0
  26. package/dist/capabilities/flyai-config.js +337 -0
  27. package/dist/capabilities/flyai.js +947 -110
  28. package/dist/capabilities/fx-contract.js +293 -0
  29. package/dist/capabilities/geo-atlas.js +246 -0
  30. package/dist/capabilities/ground-transfer.js +509 -0
  31. package/dist/capabilities/hbcli.js +260 -123
  32. package/dist/capabilities/hotelbyte-transaction.js +194 -0
  33. package/dist/capabilities/incident-log.js +131 -61
  34. package/dist/capabilities/itinerary-artifact.js +190 -0
  35. package/dist/capabilities/itinerary-deck-artifact.js +194 -0
  36. package/dist/capabilities/itinerary-deck-export.js +339 -0
  37. package/dist/capabilities/lavish-local.js +924 -0
  38. package/dist/capabilities/model-override.js +38 -0
  39. package/dist/capabilities/opensky.js +54 -43
  40. package/dist/capabilities/resilience.js +191 -0
  41. package/dist/capabilities/session/action-cache.js +50 -44
  42. package/dist/capabilities/session/adapters/ctrip-flight.js +120 -62
  43. package/dist/capabilities/session/adapters/ctrip-hotel.js +222 -0
  44. package/dist/capabilities/session/adapters/dida-portal.js +274 -0
  45. package/dist/capabilities/session/adapters/meituan-local.js +35 -54
  46. package/dist/capabilities/session/adapters/rail-12306.js +342 -0
  47. package/dist/capabilities/session/benchmark.js +189 -0
  48. package/dist/capabilities/session/bridge-ledger.js +224 -0
  49. package/dist/capabilities/session/extension-bridge.js +636 -0
  50. package/dist/capabilities/session/extension-channel.js +109 -0
  51. package/dist/capabilities/session/extension-distribution.js +230 -0
  52. package/dist/capabilities/session/extract.js +31 -21
  53. package/dist/capabilities/session/golden-score.js +114 -0
  54. package/dist/capabilities/session/health-watch.js +215 -0
  55. package/dist/capabilities/session/read-guard.js +75 -24
  56. package/dist/capabilities/session/static-flight-golden.js +150 -0
  57. package/dist/capabilities/session/transport.js +97 -83
  58. package/dist/capabilities/session/wizard.js +117 -0
  59. package/dist/capabilities/session-consent.js +178 -0
  60. package/dist/capabilities/session-login.js +170 -0
  61. package/dist/capabilities/session-search.js +635 -44
  62. package/dist/capabilities/spawn-bounded.js +203 -0
  63. package/dist/capabilities/sponsor-plugin.js +341 -0
  64. package/dist/capabilities/visa-policy.js +140 -0
  65. package/dist/capabilities/weather.js +243 -138
  66. package/dist/data/session-golden-20.json +25 -0
  67. package/dist/data/sf-golden-manifest.json +103 -0
  68. package/dist/data/sf-static-routes.json +91 -0
  69. package/dist/scripts/action-cache-tests.js +24 -59
  70. package/dist/scripts/agent-planning-turn-deadline-e2e.js +242 -0
  71. package/dist/scripts/agent-planning-turn-deadline-tests.js +185 -0
  72. package/dist/scripts/agent-reach-deep-tests.js +30 -40
  73. package/dist/scripts/agent-reach-error-tests.js +354 -0
  74. package/dist/scripts/agent-reach-tests.js +26 -18
  75. package/dist/scripts/agent-reach-wrapper-tests.js +50 -50
  76. package/dist/scripts/anything-tests.js +116 -60
  77. package/dist/scripts/artifact-client-contract-tests.js +220 -0
  78. package/dist/scripts/artifacts-capability-tests.js +727 -0
  79. package/dist/scripts/async-collect.js +50 -11
  80. package/dist/scripts/benchmark-environment-bridge-e2e.js +741 -0
  81. package/dist/scripts/benchmark-environment-bridge-tests.js +1452 -0
  82. package/dist/scripts/booking-copilot-availability-ledger-binding-tests.js +126 -0
  83. package/dist/scripts/booking-copilot-availability-policy-tests.js +265 -0
  84. package/dist/scripts/booking-copilot-bin-proof-tests.js +66 -0
  85. package/dist/scripts/booking-copilot-cleanup-proof-helper.js +76 -0
  86. package/dist/scripts/booking-copilot-crossrepo-fixture-server.js +144 -0
  87. package/dist/scripts/booking-copilot-dsh-core-proof-tests.js +424 -0
  88. package/dist/scripts/booking-copilot-dsh-planner-proof-tests.js +2839 -0
  89. package/dist/scripts/booking-copilot-dsh-plugin-proof-tests.js +309 -0
  90. package/dist/scripts/booking-copilot-dsh-readiness-proof.js +238 -0
  91. package/dist/scripts/booking-copilot-dsh-warmer-proof.js +123 -0
  92. package/dist/scripts/booking-copilot-event-sequence-concurrency-proof-tests.js +156 -0
  93. package/dist/scripts/booking-copilot-gap-code-contract-proof-tests.js +70 -0
  94. package/dist/scripts/booking-copilot-intent-proof-tests.js +422 -0
  95. package/dist/scripts/booking-copilot-operation-ledger-concurrency-proof-tests.js +184 -0
  96. package/dist/scripts/booking-copilot-planner-live-probe.js +95 -0
  97. package/dist/scripts/booking-copilot-proof-runtime.js +42 -0
  98. package/dist/scripts/booking-copilot-receipt-ledger-concurrency-proof-tests.js +210 -0
  99. package/dist/scripts/booking-copilot-runtime-proof-tests.js +2349 -0
  100. package/dist/scripts/booking-copilot-server-proof-tests.js +784 -0
  101. package/dist/scripts/booking-copilot-startup-proof-tests.js +94 -0
  102. package/dist/scripts/booking-executor-tests.js +80 -0
  103. package/dist/scripts/booking-recovery-chain-tests.js +182 -0
  104. package/dist/scripts/booking-saga-tests.js +125 -0
  105. package/dist/scripts/booking-surface-contract-proof-tests.js +212 -0
  106. package/dist/scripts/bootstrap-tests.js +1479 -0
  107. package/dist/scripts/bridge-events-tests.js +239 -0
  108. package/dist/scripts/bridge-ledger-tests.js +428 -0
  109. package/dist/scripts/build-changelog.js +238 -0
  110. package/dist/scripts/build-metrics-report.js +331 -0
  111. package/dist/scripts/changelog-tests.js +118 -0
  112. package/dist/scripts/channel-probe-tests.js +103 -0
  113. package/dist/scripts/channel-probe.js +169 -0
  114. package/dist/scripts/channel-registry-tests.js +208 -0
  115. package/dist/scripts/companion-tests.js +25 -70
  116. package/dist/scripts/dida-cookie-probe.js +87 -0
  117. package/dist/scripts/dida-runner-stop-tests.js +311 -0
  118. package/dist/scripts/dida-sw-probe.js +114 -0
  119. package/dist/scripts/diff-test.js +17 -12
  120. package/dist/scripts/doctor-tests.js +564 -0
  121. package/dist/scripts/dsh-artifact-e2e.js +279 -0
  122. package/dist/scripts/dsh-artifact-web-e2e.js +675 -0
  123. package/dist/scripts/dsh-http-proxy-sse-proof.js +90 -0
  124. package/dist/scripts/dsh-runtime-closure-tests.js +182 -0
  125. package/dist/scripts/dsh-runtime-closure.js +162 -0
  126. package/dist/scripts/dsh-session-v3-migration-proof.js +109 -0
  127. package/dist/scripts/dsh-subprocess-local-proof.js +150 -0
  128. package/dist/scripts/dsh-target-closure-proof.js +178 -0
  129. package/dist/scripts/effect-tests.js +539 -0
  130. package/dist/scripts/engine-run.js +1 -3
  131. package/dist/scripts/engine-tests.js +16 -10
  132. package/dist/scripts/evaluation-cadence-tests.js +174 -0
  133. package/dist/scripts/evaluation-contract-tests.js +268 -0
  134. package/dist/scripts/extension-distribution-cli.js +39 -0
  135. package/dist/scripts/extension-distribution-tests.js +443 -0
  136. package/dist/scripts/extension-tests.js +988 -0
  137. package/dist/scripts/external-event-tests.js +350 -0
  138. package/dist/scripts/fact-gate-tests.js +869 -0
  139. package/dist/scripts/fixtures/extension-bridge-unref-child.mjs +21 -0
  140. package/dist/scripts/flight-hotel-anchor-363-registered-e2e.js +169 -0
  141. package/dist/scripts/flight-malformed-tests.js +159 -0
  142. package/dist/scripts/flyai-setup-tests.js +228 -0
  143. package/dist/scripts/flyai-setup-tool-tests.js +54 -0
  144. package/dist/scripts/flyai-tests.js +692 -0
  145. package/dist/scripts/flyai-tool-registration-contract-tests.js +91 -0
  146. package/dist/scripts/fx-contract-tests.js +201 -0
  147. package/dist/scripts/g5-guard-tests.js +131 -0
  148. package/dist/scripts/g5-guard.js +189 -0
  149. package/dist/scripts/gen-lunar-anchors.js +106 -0
  150. package/dist/scripts/geo-atlas-tests.js +283 -0
  151. package/dist/scripts/gotry-backend-kernel-tests.js +206 -0
  152. package/dist/scripts/gotry-try-demo-tests.js +79 -0
  153. package/dist/scripts/gotry-try-demo.js +49 -0
  154. package/dist/scripts/ground-transfer-tests.js +956 -0
  155. package/dist/scripts/hbcli-e2e-tests.js +159 -0
  156. package/dist/scripts/hbcli-live-optin-tests.js +96 -0
  157. package/dist/scripts/hbcli-release-contract-tests.js +122 -0
  158. package/dist/scripts/hbcli-tests.js +121 -69
  159. package/dist/scripts/health-watch-cli.js +73 -0
  160. package/dist/scripts/hotel-date-gate-tests.js +496 -0
  161. package/dist/scripts/hotelbyte-reconcile-tests.js +141 -0
  162. package/dist/scripts/hotelbyte-spawn-e2e-tests.js +373 -0
  163. package/dist/scripts/i18n-tests.js +41 -0
  164. package/dist/scripts/incident-tests.js +281 -87
  165. package/dist/scripts/issue-194-job-id-guard-tests.js +106 -0
  166. package/dist/scripts/issue-233-cancel-refund-commission-tests.js +392 -0
  167. package/dist/scripts/issue-271-liveness-tests.js +569 -0
  168. package/dist/scripts/issue-289-model-retry-real-tests.js +296 -0
  169. package/dist/scripts/issue-289-model-retry-tests.js +441 -0
  170. package/dist/scripts/issue-289-web-retry-e2e.js +630 -0
  171. package/dist/scripts/issue-338-home-city-e2e.js +577 -0
  172. package/dist/scripts/issue343-real-entry-e2e.js +160 -0
  173. package/dist/scripts/issue436-persisted-routing-e2e.js +247 -0
  174. package/dist/scripts/itinerary-artifact-tests.js +368 -0
  175. package/dist/scripts/itinerary-deck-artifact-tests.js +416 -0
  176. package/dist/scripts/itinerary-deck-export-tests.js +552 -0
  177. package/dist/scripts/itinerary-deck-tests.js +435 -0
  178. package/dist/scripts/itinerary-html-tests.js +426 -0
  179. package/dist/scripts/journey-tests.js +22 -27
  180. package/dist/scripts/kernel-manifest-evidence.js +109 -0
  181. package/dist/scripts/kernel-manifest-gate.js +126 -0
  182. package/dist/scripts/kernel-manifest-tests.js +212 -0
  183. package/dist/scripts/kernel-manifest-trace-entry.js +7 -0
  184. package/dist/scripts/kernel-manifest-trace-hook.mjs +28 -0
  185. package/dist/scripts/kernel-manifest.js +369 -0
  186. package/dist/scripts/lavish-local-tests.js +1226 -0
  187. package/dist/scripts/lavish-product-tools-tests.js +537 -0
  188. package/dist/scripts/ledger-repair-apply-tests.js +316 -0
  189. package/dist/scripts/ledger-repair-plan-tests.js +375 -0
  190. package/dist/scripts/ledger-tests.js +382 -246
  191. package/dist/scripts/ledger-workflow-crash.js +10 -20
  192. package/dist/scripts/managed-dsh-cleanup-diagnostic-proof.js +184 -0
  193. package/dist/scripts/managed-dsh-close-failure-proof.js +172 -0
  194. package/dist/scripts/managed-dsh-run-port-proof.js +153 -0
  195. package/dist/scripts/managed-dsh-terminal-proof.js +252 -0
  196. package/dist/scripts/map-tools-vendor-package-proof.js +365 -0
  197. package/dist/scripts/memory-capture-tests.js +242 -61
  198. package/dist/scripts/memory-decay-tests.js +18 -52
  199. package/dist/scripts/memory-lifecycle-tests.js +657 -0
  200. package/dist/scripts/memory-lifecycle.js +196 -0
  201. package/dist/scripts/memory-metrics.js +24 -12
  202. package/dist/scripts/memory-value-report.js +631 -0
  203. package/dist/scripts/memory-value-tests.js +328 -0
  204. package/dist/scripts/metrics-report-tests.js +151 -0
  205. package/dist/scripts/model-override-e2e.js +137 -0
  206. package/dist/scripts/nightly-evidence-tests.js +161 -0
  207. package/dist/scripts/nightly-evidence.js +286 -0
  208. package/dist/scripts/nudge-digest.js +39 -28
  209. package/dist/scripts/observability-tests.js +381 -0
  210. package/dist/scripts/onboarding-tests.js +161 -0
  211. package/dist/scripts/opensky-check.js +18 -13
  212. package/dist/scripts/opensky-tests.js +28 -28
  213. package/dist/scripts/persona-surface-guard-tests.js +49 -0
  214. package/dist/scripts/pnpm-dsh-closure-proof.js +17 -0
  215. package/dist/scripts/policy-anchor-359-registered-e2e.js +214 -0
  216. package/dist/scripts/price-drift-tests.js +340 -0
  217. package/dist/scripts/price-drift-watch.js +447 -0
  218. package/dist/scripts/probe-poi-tests.js +53 -45
  219. package/dist/scripts/product-metrics.js +483 -0
  220. package/dist/scripts/publish-preverify.js +110 -0
  221. package/dist/scripts/realtime-pricing-tests.js +76 -0
  222. package/dist/scripts/recall-tests.js +513 -0
  223. package/dist/scripts/replay-async.js +20 -18
  224. package/dist/scripts/replay-real.js +18 -18
  225. package/dist/scripts/replay.js +52 -82
  226. package/dist/scripts/session-attach-diagnose.js +20 -20
  227. package/dist/scripts/session-attach-poc.js +34 -37
  228. package/dist/scripts/session-benchmark.js +180 -0
  229. package/dist/scripts/session-dida-live-e2e.js +199 -0
  230. package/dist/scripts/session-extract-tests.js +14 -35
  231. package/dist/scripts/session-link-tests.js +478 -0
  232. package/dist/scripts/session-login.js +27 -46
  233. package/dist/scripts/session-tests.js +796 -96
  234. package/dist/scripts/session-zones-tests.js +358 -0
  235. package/dist/scripts/sf-live-benchmark.js +367 -0
  236. package/dist/scripts/sf-live-challenge-stop-tests.js +453 -0
  237. package/dist/scripts/sf-live-cli-tests.js +92 -0
  238. package/dist/scripts/sf-live-options.js +46 -0
  239. package/dist/scripts/sf-manifest-tests.js +30 -0
  240. package/dist/scripts/sf-manifest.js +33 -0
  241. package/dist/scripts/sf-soft-score-tests.js +67 -0
  242. package/dist/scripts/sf-summary-tests.js +324 -0
  243. package/dist/scripts/sf-summary.js +525 -0
  244. package/dist/scripts/share-tests.js +432 -0
  245. package/dist/scripts/skeleton-check.js +16 -14
  246. package/dist/scripts/skeleton-integration-test.js +28 -11
  247. package/dist/scripts/skills-contract-tests.js +51 -46
  248. package/dist/scripts/smoke-session-gate-tests.js +21 -0
  249. package/dist/scripts/smoke.js +690 -246
  250. package/dist/scripts/spawn-bounded-abort-tests.js +241 -0
  251. package/dist/scripts/sponsor-reuse-tests.js +262 -0
  252. package/dist/scripts/state-cli-tests.js +226 -101
  253. package/dist/scripts/state-cli.js +550 -199
  254. package/dist/scripts/state-repair-cli.js +66 -0
  255. package/dist/scripts/state-repair-tests.js +140 -0
  256. package/dist/scripts/static-golden-tests.js +165 -0
  257. package/dist/scripts/time-eval-tests.js +210 -192
  258. package/dist/scripts/travel-timeline-tests.js +29 -93
  259. package/dist/scripts/turn-handoff-collect-tests.js +169 -0
  260. package/dist/scripts/turn-handoff-collect.js +161 -0
  261. package/dist/scripts/turn-policy-tests.js +55 -0
  262. package/dist/scripts/typed-contract-canary.js +176 -0
  263. package/dist/scripts/tz-resolver-sanity.js +50 -0
  264. package/dist/scripts/tz-resolver-tests.js +584 -0
  265. package/dist/scripts/unified-tests.js +23 -19
  266. package/dist/scripts/visa-policy-tests.js +62 -0
  267. package/dist/scripts/weather-tests.js +276 -41
  268. package/dist/scripts/wish-channel-gate-tests.js +54 -0
  269. package/dist/scripts/write-gate-tests.js +522 -0
  270. package/dist/scripts/z3-lifecycle-fault-tests.js +188 -0
  271. package/dist/scripts/z3-lifecycle-tests.js +170 -0
  272. package/dist/scripts/z3-race-repeat-tests.js +39 -0
  273. package/dist/scripts/z3-race-tests.js +41 -0
  274. package/dist/src/artifact-gate.js +829 -0
  275. package/dist/src/backend/kernel.js +91 -0
  276. package/dist/src/backend/modules/booking-copilot.js +92 -0
  277. package/dist/src/backend/modules/booking-executor.js +124 -0
  278. package/dist/src/backend/modules/session-search.js +551 -0
  279. package/dist/src/benchmark-agent-conformance.js +578 -0
  280. package/dist/src/benchmark-environment-bridge.js +515 -0
  281. package/dist/src/benchmark-headless-child-diagnostics.js +166 -0
  282. package/dist/src/benchmark-tool-isolation.js +151 -0
  283. package/dist/src/bookable-facts.js +533 -0
  284. package/dist/src/booking-saga.js +97 -0
  285. package/dist/src/booking-surface/availability-policy.js +697 -0
  286. package/dist/src/booking-surface/booking-intent.js +43 -0
  287. package/dist/src/booking-surface/cancel-refund-commission.js +379 -0
  288. package/dist/src/booking-surface/canonical-schema.js +116 -0
  289. package/dist/src/booking-surface/contracts.js +75 -0
  290. package/dist/src/booking-surface/dsh-planner.js +1854 -0
  291. package/dist/src/booking-surface/dsh-plugin.js +378 -0
  292. package/dist/src/booking-surface/dsh-worker.js +98 -0
  293. package/dist/src/booking-surface/error-codes.js +231 -0
  294. package/dist/src/booking-surface/index.js +9 -0
  295. package/dist/src/booking-surface/managed-dsh-cleanup-diagnostic.js +97 -0
  296. package/dist/src/booking-surface/managed-dsh-run-port.js +244 -0
  297. package/dist/src/booking-surface/profile.js +24 -0
  298. package/dist/src/booking-surface/recovery-chain.js +367 -0
  299. package/dist/src/booking-surface/runtime.js +1968 -0
  300. package/dist/src/booking-surface/server.js +674 -0
  301. package/dist/src/booking-surface/startup.js +168 -0
  302. package/dist/src/booking-surface/validation.js +261 -0
  303. package/dist/src/bridge.js +20 -13
  304. package/dist/src/companions.js +48 -74
  305. package/dist/src/contracts.js +21 -43
  306. package/dist/src/dsh-llm.js +97 -111
  307. package/dist/src/engine.js +276 -193
  308. package/dist/src/evaluation-cadence.js +97 -0
  309. package/dist/src/evaluation-contracts.js +283 -0
  310. package/dist/src/flight-pack-adapter.js +54 -0
  311. package/dist/src/flight-pack-contract.js +2 -0
  312. package/dist/src/flyai-setup-tool.js +70 -0
  313. package/dist/src/gotry-backend.js +125 -0
  314. package/dist/src/hotel-date-gate.js +118 -0
  315. package/dist/src/i18n.js +88 -0
  316. package/dist/src/index.js +2227 -882
  317. package/dist/src/itinerary-deck.js +139 -0
  318. package/dist/src/itinerary-doc-shared.js +661 -0
  319. package/dist/src/itinerary-html.js +131 -0
  320. package/dist/src/journey.js +155 -108
  321. package/dist/src/lavish-tools.js +400 -0
  322. package/dist/src/ledger-repair-apply.js +538 -0
  323. package/dist/src/ledger-repair-plan.js +463 -0
  324. package/dist/src/loop.js +319 -160
  325. package/dist/src/memory-capture.js +180 -32
  326. package/dist/src/memory-decay.js +30 -14
  327. package/dist/src/memory-lifecycle.js +1310 -0
  328. package/dist/src/memory-utility.js +27 -40
  329. package/dist/src/mock-llm.js +36 -61
  330. package/dist/src/model.js +81 -36
  331. package/dist/src/realtime-pricing.js +103 -0
  332. package/dist/src/recall/card.js +74 -0
  333. package/dist/src/recall/evaluator.js +140 -0
  334. package/dist/src/recall/integration.js +17 -0
  335. package/dist/src/recall/tick.js +166 -0
  336. package/dist/src/session-link/plan-it-action.js +48 -0
  337. package/dist/src/session-link/session-link.js +248 -0
  338. package/dist/src/session-zones.js +507 -0
  339. package/dist/src/share/adapters.js +51 -0
  340. package/dist/src/share/share-consent.js +76 -0
  341. package/dist/src/share/share-deck.js +58 -0
  342. package/dist/src/share/share-token.js +146 -0
  343. package/dist/src/slot-spec.js +84 -106
  344. package/dist/src/state-ledger.js +435 -516
  345. package/dist/src/state-repair.js +299 -0
  346. package/dist/src/subagent-job-id-guard.js +58 -0
  347. package/dist/src/time-anchor.js +231 -33
  348. package/dist/src/tool-packet.js +20 -8
  349. package/dist/src/travel-slots.js +58 -35
  350. package/dist/src/travel-timeline.js +44 -63
  351. package/dist/src/turn-deadline.js +287 -0
  352. package/dist/src/turn-policy.js +118 -0
  353. package/dist/src/tz-resolver.js +143 -0
  354. package/dist/src/unified.js +535 -359
  355. package/dist/src/wish-pool.js +42 -12
  356. package/dist/src/write-gate.js +495 -0
  357. package/dist/src/z3-shared.js +284 -0
  358. package/extension/README.md +21 -0
  359. package/extension/README.zh-CN.md +21 -0
  360. package/extension/background.js +463 -0
  361. package/extension/content-bridge.js +270 -0
  362. package/extension/content-main.js +111 -0
  363. package/extension/manifest.json +60 -0
  364. package/package.json +407 -14
  365. package/schemas/booking.intent.schema.json +193 -0
  366. package/schemas/booking.surface.schema.json +2622 -0
  367. package/ts/capabilities/agent-reach-deep.ts +114 -0
  368. package/ts/capabilities/agent-reach.ts +198 -0
  369. package/ts/capabilities/anything.ts +255 -0
  370. package/ts/capabilities/artifacts.ts +428 -0
  371. package/ts/capabilities/fact-log.ts +54 -0
  372. package/ts/capabilities/flyai-config.ts +388 -0
  373. package/ts/capabilities/flyai.ts +1274 -0
  374. package/ts/capabilities/hbcli.ts +253 -71
  375. package/ts/capabilities/incident-log.ts +88 -59
  376. package/ts/capabilities/itinerary-artifact.ts +214 -0
  377. package/ts/capabilities/itinerary-deck-artifact.ts +218 -0
  378. package/ts/capabilities/itinerary-deck-export.ts +378 -0
  379. package/ts/capabilities/opensky.ts +121 -0
  380. package/ts/capabilities/session/action-cache.ts +148 -0
  381. package/ts/capabilities/session/adapters/ctrip-flight.ts +181 -0
  382. package/ts/capabilities/session/adapters/ctrip-hotel.ts +238 -0
  383. package/ts/capabilities/session/adapters/dida-portal.ts +338 -0
  384. package/ts/capabilities/session/adapters/meituan-local.ts +70 -0
  385. package/ts/capabilities/session/adapters/rail-12306.ts +392 -0
  386. package/ts/capabilities/session/benchmark.ts +276 -0
  387. package/ts/capabilities/session/bridge-ledger.ts +333 -0
  388. package/ts/capabilities/session/extension-bridge.ts +794 -0
  389. package/ts/capabilities/session/extension-channel.ts +144 -0
  390. package/ts/capabilities/session/extension-distribution.ts +266 -0
  391. package/ts/capabilities/session/extract.ts +54 -0
  392. package/ts/capabilities/session/golden-score.ts +139 -0
  393. package/ts/capabilities/session/health-watch.ts +264 -0
  394. package/ts/capabilities/session/read-guard.ts +141 -0
  395. package/ts/capabilities/session/static-flight-golden.ts +209 -0
  396. package/ts/capabilities/session/transport.ts +146 -0
  397. package/ts/capabilities/session/wizard.ts +140 -0
  398. package/ts/capabilities/session-consent.ts +211 -0
  399. package/ts/capabilities/session-login.ts +208 -0
  400. package/ts/capabilities/session-search.ts +893 -0
  401. package/ts/capabilities/weather.ts +330 -0
  402. package/ts/dsh-runtime/vendor/README.md +68 -0
  403. package/ts/dsh-runtime/vendor/dsh-map-tools/LICENSE +21 -0
  404. package/ts/dsh-runtime/vendor/dsh-map-tools/README.en.md +405 -0
  405. package/ts/dsh-runtime/vendor/dsh-map-tools/README.md +404 -0
  406. package/ts/dsh-runtime/vendor/dsh-map-tools/client/client.js +1584 -0
  407. package/ts/dsh-runtime/vendor/dsh-map-tools/cordis.patch.yml +5 -0
  408. package/ts/dsh-runtime/vendor/dsh-map-tools/icon.svg +5 -0
  409. package/ts/dsh-runtime/vendor/dsh-map-tools/lib/clients/amap.js +595 -0
  410. package/ts/dsh-runtime/vendor/dsh-map-tools/lib/clients/nominatim.js +63 -0
  411. package/ts/dsh-runtime/vendor/dsh-map-tools/lib/clients/osrm.js +58 -0
  412. package/ts/dsh-runtime/vendor/dsh-map-tools/lib/clients/photon.js +55 -0
  413. package/ts/dsh-runtime/vendor/dsh-map-tools/lib/config-file.js +87 -0
  414. package/ts/dsh-runtime/vendor/dsh-map-tools/lib/config-route.js +131 -0
  415. package/ts/dsh-runtime/vendor/dsh-map-tools/lib/config.js +16 -0
  416. package/ts/dsh-runtime/vendor/dsh-map-tools/lib/geo.js +232 -0
  417. package/ts/dsh-runtime/vendor/dsh-map-tools/lib/index.js +120 -0
  418. package/ts/dsh-runtime/vendor/dsh-map-tools/lib/settings-ns.js +40 -0
  419. package/ts/dsh-runtime/vendor/dsh-map-tools/lib/staticmap-route.js +115 -0
  420. package/ts/dsh-runtime/vendor/dsh-map-tools/lib/tools/geocode.js +127 -0
  421. package/ts/dsh-runtime/vendor/dsh-map-tools/lib/tools/poi.js +84 -0
  422. package/ts/dsh-runtime/vendor/dsh-map-tools/lib/tools/routes.js +267 -0
  423. package/ts/dsh-runtime/vendor/dsh-map-tools/lib/types/clients/amap.d.ts +90 -0
  424. package/ts/dsh-runtime/vendor/dsh-map-tools/lib/types/clients/nominatim.d.ts +17 -0
  425. package/ts/dsh-runtime/vendor/dsh-map-tools/lib/types/clients/osrm.d.ts +15 -0
  426. package/ts/dsh-runtime/vendor/dsh-map-tools/lib/types/clients/photon.d.ts +22 -0
  427. package/ts/dsh-runtime/vendor/dsh-map-tools/lib/types/config-file.d.ts +33 -0
  428. package/ts/dsh-runtime/vendor/dsh-map-tools/lib/types/config-route.d.ts +23 -0
  429. package/ts/dsh-runtime/vendor/dsh-map-tools/lib/types/config.d.ts +23 -0
  430. package/ts/dsh-runtime/vendor/dsh-map-tools/lib/types/geo.d.ts +61 -0
  431. package/ts/dsh-runtime/vendor/dsh-map-tools/lib/types/index.d.ts +18 -0
  432. package/ts/dsh-runtime/vendor/dsh-map-tools/lib/types/settings-ns.d.ts +14 -0
  433. package/ts/dsh-runtime/vendor/dsh-map-tools/lib/types/staticmap-route.d.ts +31 -0
  434. package/ts/dsh-runtime/vendor/dsh-map-tools/lib/types/tools/geocode.d.ts +11 -0
  435. package/ts/dsh-runtime/vendor/dsh-map-tools/lib/types/tools/poi.d.ts +9 -0
  436. package/ts/dsh-runtime/vendor/dsh-map-tools/lib/types/tools/routes.d.ts +75 -0
  437. package/ts/dsh-runtime/vendor/dsh-map-tools/lib/types/types.d.ts +68 -0
  438. package/ts/dsh-runtime/vendor/dsh-map-tools/lib/types.js +16 -0
  439. package/ts/dsh-runtime/vendor/dsh-map-tools/package.json +97 -0
  440. package/ts/package.json +293 -3
  441. package/ts/scripts/async-collect.ts +91 -0
  442. package/ts/scripts/map-tools-vendor-package-proof.ts +404 -0
  443. package/ts/scripts/skeleton-check.ts +1 -1
  444. package/ts/scripts/skeleton-integration-test.ts +23 -1
  445. package/ts/scripts/state-cli.ts +568 -0
  446. package/ts/scripts/turn-handoff-collect.ts +178 -0
  447. package/ts/src/artifact-gate.ts +991 -0
  448. package/ts/src/benchmark-agent-conformance.ts +619 -0
  449. package/ts/src/benchmark-environment-bridge.ts +538 -0
  450. package/ts/src/benchmark-headless-child-diagnostics.ts +184 -0
  451. package/ts/src/benchmark-tool-isolation.ts +166 -0
  452. package/ts/src/bookable-facts.ts +837 -0
  453. package/ts/src/booking-saga.ts +133 -0
  454. package/ts/src/booking-surface/availability-policy.ts +522 -0
  455. package/ts/src/booking-surface/booking-intent.ts +86 -0
  456. package/ts/src/booking-surface/cancel-refund-commission.ts +488 -0
  457. package/ts/src/booking-surface/canonical-schema.js +116 -0
  458. package/ts/src/booking-surface/contracts.ts +475 -0
  459. package/ts/src/booking-surface/dsh-planner.ts +1949 -0
  460. package/ts/src/booking-surface/dsh-plugin.js +378 -0
  461. package/ts/src/booking-surface/dsh-worker.ts +110 -0
  462. package/ts/src/booking-surface/error-codes.ts +241 -0
  463. package/ts/src/booking-surface/index.ts +9 -0
  464. package/ts/src/booking-surface/managed-dsh-cleanup-diagnostic.ts +151 -0
  465. package/ts/src/booking-surface/managed-dsh-run-port.ts +273 -0
  466. package/ts/src/booking-surface/profile.ts +42 -0
  467. package/ts/src/booking-surface/recovery-chain.ts +409 -0
  468. package/ts/src/booking-surface/runtime.ts +1793 -0
  469. package/ts/src/booking-surface/server.ts +733 -0
  470. package/ts/src/booking-surface/startup.ts +201 -0
  471. package/ts/src/booking-surface/validation.ts +217 -0
  472. package/ts/src/companions.ts +113 -0
  473. package/ts/src/contracts.ts +13 -3
  474. package/ts/src/dsh-llm.ts +44 -11
  475. package/ts/src/flyai-setup-tool.ts +77 -0
  476. package/ts/src/index.ts +1948 -198
  477. package/ts/src/itinerary-deck.ts +189 -0
  478. package/ts/src/itinerary-doc-shared.ts +741 -0
  479. package/ts/src/itinerary-html.ts +171 -0
  480. package/ts/src/loop.ts +223 -25
  481. package/ts/src/memory-capture.ts +283 -0
  482. package/ts/src/memory-decay.ts +55 -0
  483. package/ts/src/memory-utility.ts +70 -0
  484. package/ts/src/mock-llm.ts +4 -8
  485. package/ts/src/model.ts +88 -2
  486. package/ts/src/recall/card.ts +106 -0
  487. package/ts/src/recall/evaluator.ts +170 -0
  488. package/ts/src/recall/integration.ts +35 -0
  489. package/ts/src/recall/tick.ts +213 -0
  490. package/ts/src/session-link/plan-it-action.ts +78 -0
  491. package/ts/src/session-link/session-link.ts +275 -0
  492. package/ts/src/share/adapters.ts +93 -0
  493. package/ts/src/share/share-consent.ts +105 -0
  494. package/ts/src/share/share-deck.ts +85 -0
  495. package/ts/src/share/share-token.ts +166 -0
  496. package/ts/src/slot-spec.ts +172 -0
  497. package/ts/src/state-ledger.ts +912 -0
  498. package/ts/src/subagent-job-id-guard.ts +72 -0
  499. package/ts/src/time-anchor.ts +338 -0
  500. package/ts/src/tool-packet.ts +32 -0
  501. package/ts/src/travel-slots.ts +219 -0
  502. package/ts/src/travel-timeline.ts +83 -0
  503. package/ts/src/turn-deadline.ts +366 -0
  504. package/ts/src/turn-policy.ts +154 -0
  505. package/ts/src/unified.ts +354 -88
  506. package/ts/src/wish-pool.ts +79 -0
  507. package/dist/scripts/session-attach-wait.js +0 -41
@@ -0,0 +1,367 @@
1
+ /**
2
+ * 真实 sf-01..08 双源 e2e 跑批(issue #21 验收清单,goal 2 sf-live-benchmark)
3
+ * ——pluggable golden 版:
4
+ *
5
+ * 之前版本绑 flyaiSearch 失败(Trial limit reached);本版改 pluggable:
6
+ * - 默认 comparator = 手工 golden(data/sf-golden-manifest.json 时间/价格带,零网络零 vendor)
7
+ * - 可选 --golden=flyai / --golden=static 显式切换;未知 vendor 在网络调用前 fail-closed
8
+ * - static = OpenFlights 固定 route/carrier 快照 + manual 时间/价格带;估算字段与回退原因显式落证据
9
+ * - 字段评分改软命中:硬字段(query_id/from/to/currency/source/verdict)必中,软字段
10
+ * (transport_number/departure_at/arrival_at/price)落在窗口内算 correct
11
+ *
12
+ * 与 ts/scripts/session-benchmark.ts 的关系:session-benchmark.ts 是离线 fixture 自测
13
+ * (13/18 字段黄金 + 双源合同纯函数断言);sf-live-benchmark.ts 是真网络真扩展端到端。
14
+ *
15
+ * evidence 落盘:--evidence-root PATH/sf-XX/<ISO ts>.json;默认 ~/.gotry/evidence/session。
16
+ * 自动化测试必须显式传入隔离目录;--help 查看用法(issue #503)。
17
+ *
18
+ * 挑战红线(RFC §3.5,issue #411):任一 query 的 session verdict=challenged
19
+ * (或双源 state=challenge_stop/guard_violation)即停止本批后续查询——不重试、
20
+ * 不绕过,session 与 comparator 都不再调用;批次以部分结果落盘,stop_reason 标注
21
+ * challenge_stop/guard_violation,并给出 attempted/not_attempted 清单。sf-summary
22
+ * 侧缺条批次一律 fail_closed,不得标为完整或有效校准。
23
+ *
24
+ * 退出码:0 = 正常结束(跑满 8 条,或按红线提前停止;stop_reason 区分;
25
+ * 即便部分 query miss 也是 0);1 = 参数/加载异常。人类评审 evidence 决定 issue 是否可关
26
+ *
27
+ * 测试节律:同 query 间隔固定 35_000ms(§3.4 ≥30s)。离线 harness 通过子进程
28
+ * timer preload 加速,不向生产暴露节律逃生开关。
29
+ */
30
+ import { mkdirSync, writeFileSync, readFileSync } from 'node:fs';
31
+ import { dirname, join } from 'node:path';
32
+ import { parseSfLiveOptions, SF_LIVE_HELP } from './sf-live-options.js';
33
+ import { validateSfManifest } from './sf-manifest.js';
34
+ import { evaluateDoubleSource, SESSION_FIELD_ACCURACY_THRESHOLD, } from '../capabilities/session/benchmark.js';
35
+ import { sessionFlightSearch } from '../capabilities/session-search.js';
36
+ import { flyaiSearch } from '../capabilities/flyai.js';
37
+ import { loadStaticFlightSnapshot, resolveStaticGolden, } from '../capabilities/session/static-flight-golden.js';
38
+ import { scoreSessionAgainstGoldenBand, } from '../capabilities/session/golden-score.js';
39
+ /** Read the source manifest before validating the frozen eight-case slice. */
40
+ function loadFlightQueries() {
41
+ return JSON.parse(readFileSync(join(import.meta.dirname, '..', 'data', 'session-golden-20.json'), 'utf8'));
42
+ }
43
+ /** 从 sf-golden-manifest.json 拿手工 golden(默认 official 来源) */
44
+ function loadManualGolden() {
45
+ return JSON.parse(readFileSync(join(import.meta.dirname, '..', 'data', 'sf-golden-manifest.json'), 'utf8'));
46
+ }
47
+ /** 手工 golden 的单条 → SessionComparableRecord */
48
+ function manualToRecord(m, sample) {
49
+ return {
50
+ query_id: m.query_id,
51
+ route_segments: [{
52
+ from: m.from,
53
+ to: m.to,
54
+ departure_at: `${m.date}T${sample.depTime}:00+08:00`,
55
+ arrival_at: `${m.date}T${sample.arrTime}:00+08:00`,
56
+ transport_number: sample.flightNo,
57
+ }],
58
+ journey_type: 'direct',
59
+ currency: 'CNY',
60
+ price: sample.price,
61
+ source: 'manual-golden',
62
+ fetched_at: new Date().toISOString(),
63
+ verdict: 'hit',
64
+ latency_ms: 0,
65
+ read_guard_blocked: 0,
66
+ };
67
+ }
68
+ /** FlyAI vendor: 真打飞猪官方(可能 trial limit,落 verdict=error) */
69
+ async function flyaiOfficial(q) {
70
+ const started = Date.now();
71
+ try {
72
+ const r = await flyaiSearch({ kind: 'flight', origin: q.from, destination: q.to, depDate: q.date, timeoutMs: 30_000 });
73
+ const latencyMs = Date.now() - started;
74
+ if (r.ok && r.options && r.options.length > 0) {
75
+ const sorted = r.options.slice().sort((a, b) => Date.parse(a.depDateTime) - Date.parse(b.depDateTime));
76
+ const opt = sorted[0];
77
+ const rec = {
78
+ query_id: q.id,
79
+ route_segments: [{
80
+ from: q.from,
81
+ to: q.to,
82
+ departure_at: opt.depDateTime,
83
+ arrival_at: opt.arrDateTime,
84
+ transport_number: opt.no,
85
+ }],
86
+ journey_type: 'direct',
87
+ currency: 'CNY',
88
+ price: opt.price,
89
+ source: 'flyai',
90
+ fetched_at: new Date().toISOString(),
91
+ verdict: 'hit',
92
+ latency_ms: latencyMs,
93
+ read_guard_blocked: 0,
94
+ };
95
+ return { record: rec, latencyMs };
96
+ }
97
+ return { record: { ...emptyOfficialRecord(q, 'flyai'), latency_ms: latencyMs, verdict: 'error' }, latencyMs, error: r.error ?? 'no options' };
98
+ }
99
+ catch (e) {
100
+ return { record: emptyOfficialRecord(q, 'flyai'), latencyMs: Date.now() - started, error: e instanceof Error ? e.message : String(e) };
101
+ }
102
+ }
103
+ /** 手工 golden vendor: 不打网络,从 manifest 取首个 known_flight 的时刻窗口算中位时间 + 价格带中位 */
104
+ function manualOfficial(q, manifest) {
105
+ const started = Date.now();
106
+ const m = manifest.matches.find((x) => x.query_id === q.id);
107
+ if (!m)
108
+ return { record: emptyOfficialRecord(q, 'manual-golden'), latencyMs: Date.now() - started };
109
+ // 中位时间(早 + 晚 ÷ 2);中位价格(带中位)
110
+ const depMedian = hhmmMid(m.window_dep_local.earliest, m.window_dep_local.latest);
111
+ const arrMedian = hhmmMid(m.window_arr_local.earliest, m.window_arr_local.latest);
112
+ const priceMedian = Math.round((m.price_cny.min + m.price_cny.max) / 2);
113
+ return {
114
+ record: manualToRecord(m, { depTime: depMedian, arrTime: arrMedian, flightNo: m.known_flights[0] ?? '', price: priceMedian }),
115
+ latencyMs: Date.now() - started,
116
+ };
117
+ }
118
+ function hhmmMid(a, b) {
119
+ const aMin = parseInt(a.split(':')[0], 10) * 60 + parseInt(a.split(':')[1], 10);
120
+ const bMin = parseInt(b.split(':')[0], 10) * 60 + parseInt(b.split(':')[1], 10);
121
+ const mid = Math.round((aMin + bMin) / 2);
122
+ return `${Math.floor(mid / 60).toString().padStart(2, '0')}:${(mid % 60).toString().padStart(2, '0')}`;
123
+ }
124
+ function emptyOfficialRecord(q, source) {
125
+ return {
126
+ query_id: q.id,
127
+ route_segments: [],
128
+ journey_type: 'direct',
129
+ currency: '',
130
+ price: 0,
131
+ source,
132
+ fetched_at: new Date().toISOString(),
133
+ verdict: 'error',
134
+ latency_ms: 0,
135
+ read_guard_blocked: 0,
136
+ };
137
+ }
138
+ /** 生产节律(§3.4 同站点 ≥30s)。测试只可在隔离子进程 mock timer。 */
139
+ const DEFAULT_INTER_QUERY_DELAY_MS = 35_000;
140
+ async function runOne(q, requestedSource, manifest, staticContext) {
141
+ const startedAt = new Date().toISOString();
142
+ let official = null;
143
+ let officialLatencyMs = null;
144
+ let effectiveSource = requestedSource === 'manual' ? 'manual-golden' : requestedSource;
145
+ let fallbackReason = null;
146
+ let estimatedFields = [];
147
+ let officialProvenance = null;
148
+ if (requestedSource === 'manual') {
149
+ const m = manualOfficial(q, manifest);
150
+ official = m.record;
151
+ officialLatencyMs = m.latencyMs;
152
+ }
153
+ else if (requestedSource === 'flyai') {
154
+ const r = await flyaiOfficial(q);
155
+ official = r.record;
156
+ officialLatencyMs = r.latencyMs;
157
+ }
158
+ else {
159
+ const started = Date.now();
160
+ const manual = manualOfficial(q, manifest);
161
+ const resolution = resolveStaticGolden({
162
+ query: q,
163
+ manualRecord: manual.record ?? emptyOfficialRecord(q, 'manual-golden'),
164
+ snapshot: staticContext.snapshot,
165
+ snapshotError: staticContext.error,
166
+ warn: (message) => console.error(message),
167
+ });
168
+ official = resolution.record;
169
+ officialLatencyMs = Date.now() - started;
170
+ effectiveSource = resolution.effective_source;
171
+ fallbackReason = resolution.fallback_reason ?? null;
172
+ estimatedFields = resolution.estimated_fields;
173
+ officialProvenance = resolution.provenance;
174
+ }
175
+ // session = sessionFlightSearch(节律闸 ≥30s 由 caller sleep 35s 处理)
176
+ const sessStarted = Date.now();
177
+ const sessRes = await sessionFlightSearch({ from: q.from, to: q.to, date: q.date, timeoutMs: 25_000 });
178
+ const sessionLatencyMs = Date.now() - sessStarted;
179
+ const sessionVerdict = sessRes.verdict;
180
+ const sessionError = sessRes.error;
181
+ let session = null;
182
+ if (sessRes.verdict === 'hit' && sessRes.options && sessRes.options.length > 0) {
183
+ const sorted = sessRes.options.slice().sort((a, b) => Date.parse(a.depDateTime) - Date.parse(b.depDateTime));
184
+ const opt = sorted[0];
185
+ session = {
186
+ query_id: q.id,
187
+ route_segments: [{
188
+ from: q.from,
189
+ to: q.to,
190
+ departure_at: opt.depDateTime,
191
+ arrival_at: opt.arrDateTime,
192
+ transport_number: opt.flightNo,
193
+ }],
194
+ journey_type: 'direct',
195
+ currency: 'CNY',
196
+ price: opt.price,
197
+ source: 'ctrip-flight',
198
+ fetched_at: new Date().toISOString(),
199
+ verdict: 'hit',
200
+ latency_ms: sessionLatencyMs,
201
+ read_guard_blocked: 0,
202
+ };
203
+ }
204
+ else {
205
+ // challenged 原样保留(issue #411):SessionVerdict 与 SessionEvidenceVerdict 同构,
206
+ // 改写为 error 会让 evaluateDoubleSource 丢失 challenge_stop 语义变成 source_unavailable
207
+ session = {
208
+ query_id: q.id,
209
+ route_segments: [],
210
+ journey_type: 'direct',
211
+ currency: '',
212
+ price: 0,
213
+ source: 'ctrip-flight',
214
+ fetched_at: new Date().toISOString(),
215
+ verdict: sessRes.verdict,
216
+ latency_ms: sessionLatencyMs,
217
+ read_guard_blocked: 0,
218
+ };
219
+ }
220
+ const doubleSource = evaluateDoubleSource({ official: official ?? undefined, session: session ?? undefined });
221
+ // 字段准确率始终对照同一 manual band;provider 只改变 official 来源与 provenance。
222
+ let softResult = null;
223
+ const m = manifest.matches.find((x) => x.query_id === q.id);
224
+ if (session?.verdict === 'hit' && m) {
225
+ softResult = scoreSessionAgainstGoldenBand(session, m, effectiveSource);
226
+ }
227
+ return {
228
+ query_id: q.id,
229
+ expected: { from: q.from, to: q.to, date: q.date, kind: q.kind },
230
+ requested_source: requestedSource,
231
+ effective_source: effectiveSource,
232
+ fallback_reason: fallbackReason,
233
+ comparator_evidence_scope: requestedSource === 'flyai' ? 'live_provider' : 'deterministic_only',
234
+ estimated_fields: estimatedFields,
235
+ official_provenance: officialProvenance,
236
+ official,
237
+ official_source: effectiveSource,
238
+ session,
239
+ doubleSource,
240
+ softScore: softResult,
241
+ officialLatencyMs,
242
+ sessionLatencyMs,
243
+ sessionVerdict: String(sessionVerdict),
244
+ sessionTransportShape: sessRes.transportShape ?? 'not_observed',
245
+ sessionError,
246
+ started_at: startedAt,
247
+ };
248
+ }
249
+ async function main() {
250
+ const args = process.argv.slice(2);
251
+ const { evidenceRoot, source: requestedSource, help } = parseSfLiveOptions(args);
252
+ if (help) {
253
+ console.log(SF_LIVE_HELP);
254
+ return;
255
+ }
256
+ const manifest = loadManualGolden();
257
+ const queries = validateSfManifest(loadFlightQueries(), manifest);
258
+ // Fail before supplier calls if the destination cannot be created (e.g. a file).
259
+ mkdirSync(evidenceRoot, { recursive: true });
260
+ const staticContext = {};
261
+ if (requestedSource === 'static') {
262
+ try {
263
+ staticContext.snapshot = loadStaticFlightSnapshot();
264
+ }
265
+ catch (error) {
266
+ staticContext.error = error instanceof Error ? error.message : String(error);
267
+ }
268
+ }
269
+ const sourceDescription = requestedSource === 'manual'
270
+ ? '手工 golden (sf-golden-manifest.json,零网络零 vendor)'
271
+ : requestedSource === 'flyai'
272
+ ? 'flyai (FlyAI 官方)'
273
+ : 'static (OpenFlights route/carrier + manual time/price band,估算字段显式标注)';
274
+ console.log(`[sf-live-benchmark] requested official source = ${sourceDescription}`);
275
+ console.log(`[sf-live-benchmark] evidence root = ${evidenceRoot}`);
276
+ console.log(`[sf-live-benchmark] 阈值:字段准确率 ≥${SESSION_FIELD_ACCURACY_THRESHOLD * 100}%(软命中)`);
277
+ const runStartedAt = new Date().toISOString();
278
+ const runStartedMs = Date.now();
279
+ const records = [];
280
+ const attemptedQueryIds = [];
281
+ let stopReason = 'completed';
282
+ for (const q of queries) {
283
+ if (records.length > 0)
284
+ await new Promise((r) => setTimeout(r, DEFAULT_INTER_QUERY_DELAY_MS));
285
+ console.log(`\n[sf-live-benchmark] ${q.id} ${q.from}→${q.to} ${q.date}`);
286
+ const rec = await runOne(q, requestedSource, manifest, staticContext);
287
+ records.push(rec);
288
+ attemptedQueryIds.push(q.id);
289
+ const evPath = join(evidenceRoot, q.id, `${runStartedAt.replace(/[:.]/g, '-')}.json`);
290
+ mkdirSync(dirname(evPath), { recursive: true });
291
+ writeFileSync(evPath, JSON.stringify(rec, null, 2));
292
+ console.log(` official(requested=${rec.requested_source},effective=${rec.effective_source}): verdict=${rec.official?.verdict ?? '-'} latency=${rec.officialLatencyMs ?? 0}ms${rec.fallback_reason ? ` fallback=${rec.fallback_reason}` : ''}`);
293
+ console.log(` session(ctrip): verdict=${rec.sessionVerdict} latency=${rec.sessionLatencyMs ?? 0}ms${rec.sessionError ? ` err=${rec.sessionError.slice(0, 80)}` : ''}`);
294
+ console.log(` doubleSource: state=${rec.doubleSource.state} quota=${rec.doubleSource.quota_disposition} price_delta=${rec.doubleSource.price_delta ?? '-'} mismatches=${rec.doubleSource.mismatches.length}`);
295
+ if (rec.softScore) {
296
+ console.log(` soft score: ${(rec.softScore.accuracy * 100).toFixed(1)}% (${rec.softScore.correct}/${rec.softScore.total}) ${rec.softScore.pass ? '✅' : '❌'} missing=${JSON.stringify(rec.softScore.missing)} incorrect=${JSON.stringify(rec.softScore.incorrect)}`);
297
+ }
298
+ // RFC §3.5 红线:风控/验证码或读守卫违例即停止本批后续查询——不重试、不绕过;
299
+ // session 与 comparator 都不再被调用,余下 query 记入 not_attempted(issue #411)
300
+ if (rec.sessionVerdict === 'challenged' || rec.doubleSource.state === 'challenge_stop') {
301
+ stopReason = 'challenge_stop';
302
+ console.log(` [sf-live-benchmark] 挑战触发(verdict=${rec.sessionVerdict},state=${rec.doubleSource.state}):停止本批后续查询(RFC §3.5;不重试不绕过)`);
303
+ break;
304
+ }
305
+ if (rec.doubleSource.state === 'guard_violation') {
306
+ stopReason = 'guard_violation';
307
+ console.log(` [sf-live-benchmark] 读守卫违例(state=guard_violation):停止本批后续查询(RFC §3.5;不重试不绕过)`);
308
+ break;
309
+ }
310
+ }
311
+ const attemptedSet = new Set(attemptedQueryIds);
312
+ const notAttemptedQueryIds = queries.filter((q) => !attemptedSet.has(q.id)).map((q) => q.id);
313
+ const accuracy_pass = records.filter((r) => r.softScore?.pass === true).length;
314
+ const comparable = records.filter((r) => r.doubleSource.state === 'comparable').length;
315
+ const liveProviderComparable = records.filter((r) => r.comparator_evidence_scope === 'live_provider' && r.doubleSource.state === 'comparable').length;
316
+ const hit = records.filter((r) => r.sessionVerdict === 'hit').length;
317
+ const challenge = records.filter((r) => r.sessionVerdict === 'challenged' || r.doubleSource.state === 'challenge_stop' || r.doubleSource.state === 'guard_violation').length;
318
+ const no_data = records.filter((r) => r.official?.verdict !== 'hit' || (r.sessionVerdict !== 'hit' && r.sessionVerdict !== 'challenged')).length;
319
+ const live_under_15s = records.filter((r) => (r.sessionLatencyMs !== null && r.sessionVerdict === 'hit' && r.sessionLatencyMs < 15_000)).length;
320
+ const effectiveSources = Array.from(new Set(records.map((record) => record.effective_source)));
321
+ const fallbackCount = records.filter((record) => record.fallback_reason !== null).length;
322
+ const summary = {
323
+ started_at: runStartedAt,
324
+ duration_ms: Date.now() - runStartedMs,
325
+ threshold: SESSION_FIELD_ACCURACY_THRESHOLD,
326
+ total: records.length,
327
+ accuracy_pass,
328
+ comparable,
329
+ live_provider_comparable: liveProviderComparable,
330
+ hit,
331
+ challenge,
332
+ no_data,
333
+ live_under_15s,
334
+ requested_source: requestedSource,
335
+ effective_sources: effectiveSources,
336
+ fallback_count: fallbackCount,
337
+ golden_source: requestedSource === 'manual' ? 'manual-golden' : requestedSource,
338
+ comparator_evidence_scope: requestedSource === 'flyai' ? 'live_provider' : 'deterministic_only',
339
+ batch_complete: stopReason === 'completed' && records.length === queries.length,
340
+ stop_reason: stopReason,
341
+ attempted_query_ids: attemptedQueryIds,
342
+ not_attempted_query_ids: notAttemptedQueryIds,
343
+ records,
344
+ };
345
+ const summaryPath = join(evidenceRoot, 'sf-summary', `${runStartedAt.replace(/[:.]/g, '-')}.json`);
346
+ mkdirSync(dirname(summaryPath), { recursive: true });
347
+ writeFileSync(summaryPath, JSON.stringify(summary, null, 2));
348
+ console.log(`\n──── sf-live-benchmark 汇总 ────`);
349
+ console.log(`golden requested = ${summary.requested_source}`);
350
+ console.log(`comparator evidence scope = ${summary.comparator_evidence_scope}${summary.comparator_evidence_scope === 'deterministic_only' ? ' (not live provider inventory evidence)' : ''}`);
351
+ console.log(`golden effective = ${summary.effective_sources.join(',')};fallback=${summary.fallback_count}`);
352
+ console.log(`跑批 query 数: ${records.length}`);
353
+ console.log(`stop reason: ${summary.stop_reason}${summary.stop_reason !== 'completed' ? `;attempted=[${summary.attempted_query_ids.join(',')}] not_attempted=[${summary.not_attempted_query_ids.join(',')}]` : ''}`);
354
+ console.log(`verdict=hit: ${hit}/${records.length}`);
355
+ console.log(`双源合同=comparable: ${comparable}/${records.length}`);
356
+ console.log(`live-provider comparable: ${liveProviderComparable}/${records.length}`);
357
+ console.log(`字段准确率 ≥${SESSION_FIELD_ACCURACY_THRESHOLD * 100}% (软命中): ${accuracy_pass}/${records.filter((r) => r.softScore !== null).length}`);
358
+ console.log(`live <15s (hit): ${live_under_15s}/${hit || 0}`);
359
+ console.log(`challenge/guard 触发: ${challenge}`);
360
+ console.log(`总耗时: ${summary.duration_ms}ms`);
361
+ process.exit(0);
362
+ }
363
+ main().catch((e) => {
364
+ console.error(`[sf-live-benchmark] 异常: ${e instanceof Error ? e.message : String(e)}`);
365
+ console.error(e instanceof Error ? e.stack : '');
366
+ process.exit(1);
367
+ });