Record corrected local Claude and staging qualification evidence

2026-10-21 16:27:18 +02:00 · 2026-10-11 00:09:40 -05:00
1 parent 9c678b2d90
commit 42daef1a1a
@@ -2,19 +2,19 @@

 **Experimental and unmerged. Local journeys and partial staging qualification passed on the sources below; final-source staging acceptance remains incomplete.** A successful build, probe, or earlier-source journey does not establish the remaining product conditions.

-## Current source boundary (2026-10-11 04:40 UTC)
+## Current source boundary (2026-10-11 05:06 UTC)

 | Surface | Source and evidence |
 | --- | --- |
-| Combined branch | `b698f54966`: includes reviewed Stop, verification, legacy explicit-continuation and safe startup-stage diagnostics. Local40 served controller `8ccfdd3dd2`, but later inspection disproved the claimed diagnostic artifact selection: saved test-drive config still selected the old remote runner and pack. Local42 corrects those two fixture paths and is testing a fresh task. |
-| Core branch | `28d3fc1b65`; focused tests and independent reviews passed. Prior exact heads core9a31/combinedb68 passed all CI and Greptile5/5; those results exclude the new startup-stage diagnostics. |
-| Serving staging | Deployment 19 `05a1d1f2172261066ca53d3776f798c3073fc49f` verified at 04:34:10.638 UTC. It includes pause, verifier, Stop-race and safe rejection diagnostics; the latest legacy periodic-recovery correction is not yet deployed. |
-| Cancelled candidate | Deployment 17 `ca8ab88c78556c4711d59f71c6293619543a66a1` was cancelled during build before check/deploy. Fresh review found that Stop could leave a retry authorized and that premature terminalization could release the task lock. Corrections passed real cancellation/lifecycle-path tests and independent review; deployment18 of `bd86e22e72` is in progress. |
-| Provider and live state | Both providers recovered with verified snapshots: staging archived at 03:30:20.703, local at 03:34:37.716. Local normal Computer Connect subsequently resumed it for read-only cache diagnosis. On local39, the paused message recovered successfully with its original marker intact. Claude passed verification but failed at provider session opening. Local archived again at 04:04:56.078 with a verified snapshot and zero active owners. No provider stop was forced or cancelled and no recovery gate or database state was manually cleared. |
+| Combined branch | `0421a3b8c027c60af6b754980bf2eef02d895e8f`. Full recursive typecheck and build passed. It includes the reviewed lifecycle and legacy continuation fixes, safe startup diagnostics, deterministic diagnostic test barrier, updated source-bound Copilot profile, and current upstream changes. Latest-head CI remains pending. |
+| Core branch | `18f5f26ca5043793166e54a41d38adbb15f19f8b` (test-only follow-up to `d7d7808302`). Canonical Boat migration 0333 follows unchanged upstream 0332. Previous exact-head CI passes remain historical evidence, not a pass for this source. |
+| Serving staging | Deployment 19 `05a1d1f2172261066ca53d3776f798c3073fc49f` verified at 04:34:10.638 UTC. Native Claude fresh/warm, saved Codex continuation and computer use, direct desktop input, and state-preserving Vite HMR passed. The Boat naturally archived at 04:56:59.592 UTC. |
+| Staging candidate | Deployment 20 of `d50dd0f49befcbe53875ad3acad9a39381f4a89b` is in progress. It preserves applied staging migration history and appends only the missing upstream index. Runtime comparison against the combined branch differs only in two migration filename references. Serving verification and final legacy regression remain pending. |
+| Local retry | Local42 serves controller `7c7c35b1ca` and fresh task BOA19 has the matched `28d3fc1b65` diagnostic pack selected. Its first-use 1166 MiB upload completed at 05:04:11 UTC. The fresh turn succeeded and reread both original proof hashes; an ordinary warm follow-up also succeeded at 05:08:05 UTC with the same saved session, runner, and provider thread. Earlier local40/41 attempts selected old remote artifacts through saved test-drive configuration; the correction preserves the database and historical failed sessions. |

-Latest source-specific live successes are staging warm HMR and corrected native computer use, local detach/reconnect retention, and local saved-session continuation after controller restart with an unchanged one-line marker. Latest native Claude saved-session recovery failed and active sibling cancellation isolation remains unverified. Pending admission and pause/resume passed locally; concurrent Stop is covered by regression tests. A local cancellation during a **failing** stop is narrower evidence.
+Staging19 establishes the main native runtime, desktop, and warm development journeys on its recorded source. The older quarantined local Claude session remains a failure. Live sibling cancellation isolation remains unverified. The full local suite is not green; latest-head CI and final-source staging acceptance are still pending.

-The [consolidated retry receipt](assets/2026-10-10-boat/evidence/retry-qualification-summary.json) preserves exact sources, timestamps, original failures, and receipt hashes. A deployment proves serving source, not product acceptance. Staging retains its previously applied migration prefix; no database reset, history rewrite, or guard bypass was used. Historical observations below retain their original cutoffs.
+The [consolidated retry receipt](assets/2026-10-10-boat/evidence/retry-qualification-summary.json) preserves earlier sources, timestamps, failures, and receipt hashes. [Staging19 receipts](assets/2026-10-10-boat/evidence/staging19-live-evidence.json) contain the newer successes. A deployment proves serving source, not product acceptance. No provider stop was forced or cancelled, no quarantine was cleared, and no database history was reset or rewritten. Dated observations below retain their original cutoffs.

 ## Product outcomes

@@ -196,3 +196,13 @@ The startup-stage diagnostic patch (`28d3fc1b65` / combined `b698f54966`) preser
 **Staging19 passed native Claude fresh and warm jobs, native Codex saved-task continuation, actual computer use, direct browser desktop input, and warm HMR.** Claude retained the recorded runner/session/owner identity, while its provider subprocesses restarted. Codex retained runner, provider, Vite, owner and saved-session identities. The heading changed to `Boat staging warm HMR confirmed 0450` while unsaved input `keep-this-unsaved-draft-boat-0449` and page-session `1791694134833` survived. Proof hashes remained unchanged. The desktop viewer closed normally after inspection. These results do not validate the older quarantined Claude fixture or legacy jobs on deployment19.

 Evidence: [consolidated receipts](assets/2026-10-10-boat/evidence/staging19-live-evidence.json), [Claude warm result](assets/2026-10-10-boat/staging-19-claude-warm-proof.png), [HMR before](assets/2026-10-10-boat/staging-19-hmr-before.png), [HMR after](assets/2026-10-10-boat/staging-19-hmr-after.png), [direct desktop input](assets/2026-10-10-boat/staging-19-direct-desktop-input.png).
+
+## Latest build and deterministic fixture check (2026-10-11 05:08 UTC)
+
+The full recursive typecheck and build passed on combined `0421a3b8c0`. Core `18f5f26ca5` / combined `9c678b2d90` adds only a signoff test sequencing correction: it waits for the actual process run to finish before sending the agent stage decision. The previously failing case passed alone, and all five signoff cases passed, including negative permission checks. The test uses disposable databases, which were removed after verification; the active product fixtures were untouched. No runtime change or staging redeployment is needed for this test-only commit. Latest-head CI and review remain pending.
+
+## Local42 corrected native Claude cold and warm pass (2026-10-11 05:09 UTC)
+
+The corrected fixture passed both fresh and ordinary warm jobs. The running executable hash matched the intended diagnostic binary, and the selected pack digest matched its source receipt. Both turns read the original persistent proof hashes unchanged. The native session, logical runner, provider thread, process identity and computer owner remained stable across warm reuse. Cold completion took 12m02 including the first-use 1166 MiB upload; the warm follow-up took 1m21. This proves the corrected fixture works; it does not explain or erase earlier failures. No saved state, quarantine, or package authority was reset. The configured five-minute idle period is now being observed before a post-idle follow-up.
+
+[Receipt](assets/2026-10-10-boat/evidence/local42-claude-cold-warm-proof.json) · [Visible result](assets/2026-10-10-boat/local-42-claude-warm-proof.png).
@@ -60,3 +60,12 @@ Local38 pending admission and Pause work passed; later normal continuation remai
 [Fresh Claude diagnosis](local40-fresh-claude-diagnosis.json) confirms settled checkpoints and present inputs without identifying the stalled startup step. [CI and review receipt](core9a31-ui-b68-ci-review.json) records both previous exact heads green; it excludes the later startup-stage diagnostic patch.

 **Artifact-selection correction:** the requested local40/41 diagnostic paths were ignored by test-drive; see [corrected provenance](local-artifact-selection-correction.json). Those failures do not establish execution of the new diagnostics. [Staging19 live receipts](staging19-live-evidence.json) establish fresh/warm Claude, saved Codex, desktop and HMR passes on the verified deployed source.
+
+### Latest migration lineage checks
+
+- [Canonical migration 0333](canonical-migration-0333-proof.json): generated Boat schema follows unchanged upstream 0332.
+- [Staging migration 0333](staging-migration-0333-proof.json): applied staging history is preserved; only the missing upstream index is appended.
+
+Both passed numbering, safety, schema drift, and database TypeScript checks. The concurrent-index database test skipped locally because embedded PostgreSQL was unavailable. Official generation pruned the old 0322 metadata snapshot; no applied SQL or journal history was rewritten.
+
+- [Local42 corrected Claude cold/warm pass](local42-claude-cold-warm-proof.json): actual executable hash, selected pack, saved session, runner identity and unchanged file hashes. Post-idle qualification remains pending.
@@ -0,0 +1,29 @@
+{
+  "sqlByteIdentical": true,
+  "sqlSha256": "42787b7bd020429b61551af0e602d749dc763bb678fe9e716c994543aa1abe23",
+  "upstream0332Unchanged": true,
+  "upstreamJournalPrefixUnchanged": true,
+  "newJournal": {
+    "idx": 333,
+    "version": "7",
+    "when": 1791694491443,
+    "tag": "0333_smooth_mordo",
+    "breakpoints": true
+  },
+  "snapshotPrevId": "8dadd21e-8c02-4ada-b99a-55ba133e7f70",
+  "snapshotAddedTables": [
+    "public.computers"
+  ],
+  "questionHistoryIndexPreserved": true,
+  "commit": "d7d7808302f3bd8045ee3d7afff55a066635f463",
+  "worktreeClean": true,
+  "prunedSnapshot": "0322_snapshot.json",
+  "checks": {
+    "officialGenerate": true,
+    "numbering": true,
+    "safety": true,
+    "dbTypecheckDirect": true,
+    "snapshotDrift": "1 passed",
+    "upstreamConcurrentIndexDatabaseTest": "skipped: embedded PostgreSQL unavailable"
+  }
+}
@@ -0,0 +1,67 @@
+{
+  "observedAt": "2026-10-11T05:08:59.461805+00:00",
+  "task": "BOA-19",
+  "source": "7c7c35b1ca2c82d378f49afdbb2d06bb8af5c673",
+  "runs": [
+    {
+      "id": "ab152fcf-ea0a-4f74-b83b-24d0242bbe55",
+      "status": "succeeded",
+      "started_at": "2026-10-11T04:54:12.655Z",
+      "finished_at": "2026-10-11T05:06:15.012Z",
+      "native_session_id": "438f844d-05d6-4ce2-ba3e-edbd97abca66",
+      "runner_instance_id": "13a06b89-735f-45d0-97cb-e41b4ff757cd",
+      "session_id_before": null,
+      "session_id_after": "1ef41bc8-eb7c-4d38-9b24-aa78b87de767",
+      "error_code": null
+    },
+    {
+      "id": "3c08b933-f8c6-455e-aadb-48c1a69c22ba",
+      "status": "succeeded",
+      "started_at": "2026-10-11T05:06:44.746Z",
+      "finished_at": "2026-10-11T05:08:05.635Z",
+      "native_session_id": "438f844d-05d6-4ce2-ba3e-edbd97abca66",
+      "runner_instance_id": "13a06b89-735f-45d0-97cb-e41b4ff757cd",
+      "session_id_before": "1ef41bc8-eb7c-4d38-9b24-aa78b87de767",
+      "session_id_after": "1ef41bc8-eb7c-4d38-9b24-aa78b87de767",
+      "error_code": null
+    }
+  ],
+  "sameNativeSession": true,
+  "sameRunnerInstance": true,
+  "sameProviderThreadAfterWarm": true,
+  "providerPackDigest": "sha256:af875f72dbce8bf9f1bb2cce962e8380d5b23dd72fb98c84d04e949d9665c719",
+  "liveExecutableProof": {
+    "checkedAt": "2026-10-11T05:08:12.695516+00:00",
+    "ownerId": "8a4e0587-1400-423d-90d4-467fc5316e64",
+    "bootId": "d4442d4d-66b7-4225-ab33-b1924fa0e575",
+    "launchGeneration": 1,
+    "pid": 61492,
+    "startTicks": "80999",
+    "runnerSha256": "366a0fe12a39a6be773540369cd577810590d1b0d03491e834b414575087a3c9",
+    "runnerBytes": 37535320,
+    "expectedHashMatches": true,
+    "admissionStages": [],
+    "journalExitCode": 0
+  },
+  "firstUseUpload": {
+    "totalMiB": 1166,
+    "completedAt": "2026-10-11T05:04:11.777Z"
+  },
+  "limitations": [
+    "Per-stage timing not captured on successful admission",
+    "Does not retroactively explain earlier timeouts",
+    "Post-idle archival/resume not yet observed"
+  ],
+  "guestReadsEndedAt": "2026-10-11T05:08:12.695516+00:00",
+  "artifactSource": "28d3fc1b6580ed2088f7c91b7bd61b8348ff1de7",
+  "proofFiles": [
+    {
+      "name": "boat-final-claude.txt",
+      "sha256": "14fa23911a1a95ede22a93f43fc027934b20a1ba497f8e6120abc8769498e124"
+    },
+    {
+      "name": "boat-claude-framing-proof.txt",
+      "sha256": "afaf525f4bf0c38614c5df0cf97140e7d1bf9cad96faa293e2317dc82116af4a"
+    }
+  ]
+}
@@ -0,0 +1,35 @@
+{
+  "stageJournalPrefixByteValuesUnchanged": true,
+  "appliedSqlAndSnapshotFilesUnchanged": true,
+  "prunedHistoricalSnapshot": [
+    "meta/0322_snapshot.json"
+  ],
+  "newJournal": {
+    "idx": 333,
+    "version": "7",
+    "when": 1791694801825,
+    "tag": "0333_furry_butterfly",
+    "breakpoints": true
+  },
+  "newSnapshotPrevId": "d863f835-b38f-4555-ba6c-2661a8ca05e7",
+  "onlySchemaDelta": [
+    "public.agent_wakeup_requests"
+  ],
+  "newIndex": "agent_wakeup_requests_question_response_history_idx",
+  "newSqlByteIdenticalToUpstream": true,
+  "commit": "d50dd0f49befcbe53875ad3acad9a39381f4a89b",
+  "worktreeClean": true,
+  "canonicalUiComparison": "0421a3b8c027c60af6b754980bf2eef02d895e8f",
+  "runtimeDifferencesExcludingMigrationHistory": [
+    "packages/db/src/client.ts: stage migration filename",
+    "packages/db/src/question-response-history-migration.test.ts: stage migration filename"
+  ],
+  "validation": {
+    "officialPnpmGenerate": true,
+    "numberingAndSafety": true,
+    "dbTypecheck": true,
+    "snapshotDrift": "1 passed",
+    "concurrentIndexDatabaseTest": "1 skipped: embedded PostgreSQL unavailable",
+    "noDatabaseApplied": true
+  }
+}