Skip to content

Record overnight human actions and failure evidence - #1336

Merged
Chris0Jeky merged 9 commits into
mainfrom
issue-1327/overnight-handoff
Jul 13, 2026
Merged

Record overnight human actions and failure evidence#1336
Chris0Jeky merged 9 commits into
mainfrom
issue-1327/overnight-handoff

Conversation

@Chris0Jeky

@Chris0Jeky Chris0Jeky commented Jul 13, 2026

Copy link
Copy Markdown
Owner

Summary

  • add the maintainer-only branch protection, release tag, trademark, and project-token actions discovered during the 2026-07-13 overnight run;
  • keep unresolved CI/test blockers visible in the human-owned checklist;
  • mirror the exact SQLite, Redis, background-worker, Extended-workflow, and GitHub Project failures into the durable failure ledger.

This intentionally does not rewrite the stale archive-direction sections; #1296/#1328 own that direction update.

Verification

  • dotnet test backend/Taskdeck.sln -c Release (not run: docs-only handoff)
  • frontend typecheck/build/Vitest (not run: no frontend change)
  • Playwright (not run: no UI change)
  • python scripts/agent_hooks/render_failure_ledger.py
  • every JSONL line parsed with ConvertFrom-Json
  • node scripts/check-docs-governance.mjs
  • git diff --check

Documentation

  • docs/STATUS.md updated (not applicable; no shipped behavior and direction is maintainer-owned)
  • docs/IMPLEMENTATION_MASTERPLAN.md updated (not applicable; no roadmap change)
  • testing guides updated (not applicable; failures are recorded, not reclassified as expected behavior)
  • OUTSTANDING_TASKS.md and docs/agentic/FAILURE_LEDGER.md updated

Tracking

Refs #1327
Related: #1173, #1282, #1299, #1303, #1330, #1332, #1335

Project Priority/status could not be audited because the current token lacks read:project/project write scope; that limitation is itself recorded as a human action.

CI Workflow Validation

  • No workflow/deploy/script/project file changed; CI Extended does not apply.

Risk Notes

  • Security impact: none; no credentials or policy are changed.
  • Behavior/regression risk: none; docs-only.
  • Follow-up tasks: only the maintainer checks off OUTSTANDING_TASKS.md; this PR is not eligible for autonomous merge under the overnight grant.

Review fixes

Copilot AI review requested due to automatic review settings July 13, 2026 03:15

Copilot AI left a comment

Copy link
Copy Markdown
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Copilot was unable to review this pull request because the user who requested the review has reached their quota limit.

Copy link
Copy Markdown
Owner Author

Self-review — head 3544c138

Reviewed the complete docs-only diff against the overnight authority, existing human-owned checklist rules, and failure-capture classifications. Existing PR comments/reviews/threads: none.

Findings: none.

Checks:

  • no existing item was checked off;
  • no stale direction section was rewritten;
  • every new deferral links to a durable issue/tracker;
  • narrow reruns are explicitly not described as clearing failed full gates;
  • docs governance and git diff --check pass;
  • PR body has no closing keyword and GraphQL confirms zero closing-issue references.

This PR remains open for the maintainer; it is outside the overnight autonomous-merge set.

@gemini-code-assist gemini-code-assist Bot left a comment

Copy link
Copy Markdown

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Code Review

This pull request updates OUTSTANDING_TASKS.md and docs/agentic/FAILURE_LEDGER.md to document several overnight deferrals and pre-existing test/CI failures identified on 2026-07-13. The feedback suggests prefixing the 'Future fix' column for the github/project-sync entry in the failure ledger with '#1327:' to maintain consistency with the other newly added entries.

Important

The consumer version of Gemini Code Assist on GitHub is being sunset. Starting June 18, 2026, new organization installations will be blocked, and all code review activity will officially cease on July 17, 2026.
For more details on the timeline and next steps, please review the Help Documentation.

Comment thread docs/agentic/FAILURE_LEDGER.md Outdated

@chatgpt-codex-connector chatgpt-codex-connector Bot left a comment

Copy link
Copy Markdown

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

💡 Codex Review

Here are some automated review suggestions for this pull request.

Reviewed commit: 3544c138cd

ℹ️ About Codex in GitHub

Your team has set up Codex to review pull requests in this repo. Reviews are triggered when you

  • Open a pull request for review
  • Mark a draft as ready
  • Comment "@codex review".

If Codex has suggestions, it will comment; otherwise it will react with 👍.

Codex can also answer questions or update the PR. Try commenting "@codex address that feedback".

Comment thread docs/agentic/FAILURE_LEDGER.md Outdated

Copy link
Copy Markdown
Owner Author

Review fix evidence on head c8ff7a58: both findings are fixed. The JSONL source is now committed and the Markdown is renderer-derived; the #1327 prefix is present in source/output. JSONL parse, renderer, docs governance, and diff check all pass. Both inline threads were replied to and resolved. This docs PR remains outside the autonomous-merge set.

@Chris0Jeky

Copy link
Copy Markdown
Owner Author

Added the newly discovered HTTP MCP security deferral in f0e6856e: OUTSTANDING now links #1338 and states the exact same-route/auth prerequisite before HTTP transport may be documented or exposed. Docs governance and git diff --check pass. No existing item was checked off.

@codex review

@chatgpt-codex-connector

Copy link
Copy Markdown

Codex Review: Didn't find any major issues. You're on a roll.

Reviewed commit: f0e6856e00

ℹ️ About Codex in GitHub

Your team has set up Codex to review pull requests in this repo. Reviews are triggered when you

  • Open a pull request for review
  • Mark a draft as ready
  • Comment "@codex review".

If Codex has suggestions, it will comment; otherwise it will react with 👍.

Codex can also answer questions or update the PR. Try commenting "@codex address that feedback".

@Chris0Jeky

Copy link
Copy Markdown
Owner Author

Added the explicit #1323 prompt-rail remainder in db818702: the hostile fixtures in PR #1340 must be bound to PR #1312's effective prompt/parser path before GEN-04, with grounded-task-or-empty and deterministic-fallback proof. Docs governance and diff check pass; no item was checked off.

@codex review

@chatgpt-codex-connector

Copy link
Copy Markdown

Codex Review: Didn't find any major issues. Chef's kiss.

Reviewed commit: db818702fa

ℹ️ About Codex in GitHub

Your team has set up Codex to review pull requests in this repo. Reviews are triggered when you

  • Open a pull request for review
  • Mark a draft as ready
  • Comment "@codex review".

If Codex has suggestions, it will comment; otherwise it will react with 👍.

Codex can also answer questions or update the PR. Try commenting "@codex address that feedback".

@Chris0Jeky

Copy link
Copy Markdown
Owner Author

Durable failure capture update — 9548c18c

The required E2E failure observed on auth PR #1334 was investigated rather than dismissed:

Verification:

  • all 11 JSONL lines parse;
  • python scripts/agent_hooks/render_failure_ledger.py reproduced the Markdown;
  • docs governance, golden-principles governance, and GitHub-operations governance passed;
  • git diff --check passed.

This PR still closes no issue and must not be autonomously merged. Fresh review requested for exact head 9548c18c.

@Chris0Jeky

Copy link
Copy Markdown
Owner Author

@codex review exact head 9548c18c: please review the narrow #1343 OUTSTANDING/failure-ledger update. Do not merge.

@chatgpt-codex-connector

Copy link
Copy Markdown

Codex Review: Didn't find any major issues. 🚀

Reviewed commit: 9548c18c69

ℹ️ About Codex in GitHub

Your team has set up Codex to review pull requests in this repo. Reviews are triggered when you

  • Open a pull request for review
  • Mark a draft as ready
  • Comment "@codex review".

If Codex has suggestions, it will comment; otherwise it will react with 👍.

Codex can also answer questions or update the PR. Try commenting "@codex address that feedback".

@Chris0Jeky

Copy link
Copy Markdown
Owner Author

Required Windows API failure investigation

The docs-only branch did not change API/runtime code. Windows API Integration failed because TryClaimProcessingAsync_ShouldRefreshTrackedEntity could no longer find the queued row after the long-lived full test host ran; Ubuntu passed. The exact Windows test then passed 5/5 locally on this unchanged head.

Evidence is added to #1335: #1335. This is not dismissed as flaky; #1335 owns the hosted-worker/test-lifecycle repair. I am rerunning the failed job after investigation. The E2E check is separately skipped by the docs-only path filter, so this PR remains ineligible under tonight's zero-skipped-check merge rule even if the rerun passes; it was never authorized for autonomous merge.

@Chris0Jeky

Copy link
Copy Markdown
Owner Author

The unchanged-head Windows API rerun passed: https://github.com/Chris0Jeky/Taskdeck/actions/runs/29225697410/job/86740437756. The original failure remains mapped to #1335 with exact 5/5 local evidence. This clears the failed API job, but the docs-only E2E check was skipped by path filtering and CI Extended remains red under #1330; #1336 is still never-automerge.

@Chris0Jeky

Copy link
Copy Markdown
Owner Author

Final overnight deferral update (1a060f63)

Verification: JSONL parse passed; renderer completed; docs, golden-principles, and GitHub-ops governance passed; git diff --check passed. No existing task was checked off and this PR retains zero closing references.

Fresh CI/review may continue after wrap-up; the PR remains unmerged.

@chatgpt-codex-connector chatgpt-codex-connector Bot left a comment

Copy link
Copy Markdown

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

💡 Codex Review

Here are some automated review suggestions for this pull request.

Reviewed commit: 1a060f63ab

ℹ️ About Codex in GitHub

Your team has set up Codex to review pull requests in this repo. Reviews are triggered when you

  • Open a pull request for review
  • Mark a draft as ready
  • Comment "@codex review".

If Codex has suggestions, it will comment; otherwise it will react with 👍.

Codex can also answer questions or update the PR. Try commenting "@codex address that feedback".

Comment thread OUTSTANDING_TASKS.md Outdated

Copy link
Copy Markdown
Owner Author

Review fix evidence — 567bd89f

Verification passed: docs governance, golden-principles governance, GitHub-operations governance, and git diff --check. No maintainer-owned item was checked off. Fresh exact-head CI/review is now required.

Copy link
Copy Markdown
Owner Author

@codex review

Please run a fresh independent docs-only review of exact head 567bd89f, with special attention to whether the pushed origin/issue-1274/paper-e2e-axe pointer is remotely recoverable and whether any maintainer-owned checklist item was accidentally completed. Inspect all existing comments/threads; do not merge.

@chatgpt-codex-connector

Copy link
Copy Markdown

Codex Review: Didn't find any major issues. 👍

Reviewed commit: 567bd89fe6

ℹ️ About Codex in GitHub

Your team has set up Codex to review pull requests in this repo. Reviews are triggered when you

  • Open a pull request for review
  • Mark a draft as ready
  • Comment "@codex review".

If Codex has suggestions, it will comment; otherwise it will react with 👍.

Codex can also answer questions or update the PR. Try commenting "@codex address that feedback".

Copy link
Copy Markdown
Owner Author

Required E2E failure investigation — run 29277660448

The exact-head docs-only run failed 1/166 Playwright cases: error-recovery.spec.ts:330, expected the locally selected workspace mode workbench after a failed save but received stale guided for the full 8-second assertion window. The other 157 tests passed and 8 were intentionally skipped.

This is the same concrete late-response ordering defect tracked in #1343; it is not a change in 567bd89f, which only makes #1274's handoff pointer remotely durable. I am not dismissing the failure: the exact assertion and run are now added to #1343, and I am rerunning only the failed E2E job to gather unchanged-head evidence. #1336 remains merge-ineligible unless the rerun is green, and #1343 remains open regardless.

Copy link
Copy Markdown
Owner Author

The unchanged-head failed-job rerun passed: CI run 29277660448, E2E job 86914762911. Required CI now concludes SUCCESS on exact head 567bd89f.

The original workbench → stale guided assertion failure remains recorded on #1343 and is not erased by the rerun. This establishes that the one-line docs/handoff repair is green while preserving the separate product race as open work.

@chatgpt-codex-connector chatgpt-codex-connector Bot left a comment

Copy link
Copy Markdown

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

💡 Codex Review

Here are some automated review suggestions for this pull request.

Reviewed commit: 67c86f1777

ℹ️ About Codex in GitHub

Your team has set up Codex to review pull requests in this repo. Reviews are triggered when you

  • Open a pull request for review
  • Mark a draft as ready
  • Comment "@codex review".

If Codex has suggestions, it will comment; otherwise it will react with 👍.

Codex can also answer questions or update the PR. Try commenting "@codex address that feedback".

Comment thread OUTSTANDING_TASKS.md
- [ ] **#1303** — Push the `v0.1.0` release tag only after the maintainer accepts the release deck and its pipeline evidence. Agents may repair the pipeline but must not push the tag.
- [ ] **#1299 trademark check** — Complete the human-only Taskdeck name/trademark clearance step; the licensing/DCO implementation does not substitute for legal-name clearance.
- [ ] **#1327 project sync access** — Run `gh auth refresh -s project`, then rerun `scripts/github/Sync-TaskdeckProjectPriority.ps1` in audit and apply modes and verify no issue/PR project item has an empty Priority.
- [ ] **#1330** — Repair the CI Extended reusable-workflow permission contract; affected PRs remain ineligible while the workflow startup failure is red.

Copy link
Copy Markdown

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

P2 Badge Remove the stale #1330 blocker

Because this commit is based on cf08c69, the #1330 fix is already present in the parent: .github/workflows/ci-extended.yml now guards the Gitleaks reusable job to PR events and grants contents: read plus pull-requests: read on the job. Keeping this new OUTSTANDING entry as an open repair item that says affected PRs remain ineligible will make the start-of-session checklist tell future agents to keep blocking PRs for a workflow startup failure that has already been repaired; please drop it or rewrite it as verification/historical ledger evidence, including the matching open ledger source row.

Useful? React with 👍 / 👎.

Comment thread OUTSTANDING_TASKS.md
- [ ] **#1282 / #1332 / #1335** — Stabilize the required full-suite harness: SQLite concurrency, Redis connect/dispose, and hosted-worker/presence lifecycle isolation. Narrow passes do not clear the recorded full-suite failures.
- [ ] **#1338** — Map HTTP MCP and its API-key middleware to the same `/mcp` route before documenting or exposing HTTP transport; the current parameterless mapping serves MCP at an unintended unauthenticated root route.
- [ ] **#1343** — Prevent late Home/Today summary responses from overwriting a newer local workspace-mode choice; add deterministic ordering tests and repeated error-recovery E2E proof.
- [ ] **#1345** — Resolve or formally mitigate GHSA-2m69-gcr7-jv3q in the transitive `SQLitePCLRaw.lib.e_sqlite3` package once a patched release exists or a safe provider change is proven; current published releases through 2.1.11 are affected.

Copy link
Copy Markdown

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

P2 Badge Reframe the SQLite advisory task around available releases

This durable task tells agents to wait until a patched release exists, but the current package metadata no longer matches that framing: the GitHub advisory marks SQLitePCLRaw.lib.e_sqlite3 affected only through <= 2.1.11, while NuGet already lists stable 3.53.3 published on July 6, 2026. Leaving this as a wait-for-upstream item can delay the security mitigation; please rewrite it to evaluate/prove the available 3.x override or another runtime mitigation rather than waiting for a nonexistent future trigger, including the matching ledger source row.

Useful? React with 👍 / 👎.

Comment thread OUTSTANDING_TASKS.md
## E. Revival overnight deferrals (added 2026-07-13)

- [ ] **#1173** — Apply and verify the required branch-protection settings in GitHub. This is repository-settings work and remains maintainer-only.
- [ ] **#1303** — Push the `v0.1.0` release tag only after the maintainer accepts the release deck and its pipeline evidence. Agents may repair the pipeline but must not push the tag.

Copy link
Copy Markdown

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

P2 Badge Gate the release tag on the staging-trigger fix

When this new checklist item is followed and v0.1.0 is pushed, .github/workflows/release-desktop.yml runs on v* tags and creates a GitHub Release, while .github/workflows/cd-staging-gate.yml still triggers on release: published and waits on the production environment. Since the repo already tracks that auto-trigger hang as #1228, this handoff should name fixing/gating that workflow as a prerequisite to the tag; otherwise the durable checklist can send the maintainer into a known-hanging release side effect.

Useful? React with 👍 / 👎.

@Chris0Jeky
Chris0Jeky merged commit 9c938d9 into main Jul 13, 2026
20 checks passed
@github-project-automation github-project-automation Bot moved this from Pending to Done in Taskdeck Execution Jul 13, 2026
Chris0Jeky added a commit that referenced this pull request Jul 13, 2026
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

Status: Done

Development

Successfully merging this pull request may close these issues.

2 participants