Repository navigation
quest: unblock the dart tag push, OSS-Fuzz application, and the nixpkgs review - #8517
Conversation
…gs review Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
|
No actionable comments were generated in the recent review. WalkthroughThe pull request updates quest documentation. It adds notes about a test stall, records the Dart 0.1.1 publication, changes review-fix preparation instructions, and records OSS-Fuzz application decisions. It also removes an earlier Dart release-planning document. Merge Risk:
|
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
There was a problem hiding this comment.
Important
The Dart publish landed between this branch's first commit and its head: the moq-dart-v0.1.1 tag was pushed at 21:59Z and release-dart.yml failed to publish, yet quest/m0/ffi-shutdown-backport.md still records the tag push as the next step. One inline comment covers the line to fix.
Reviewed changes
/quest-unblock output for 2026-11-11: four quest-text files, no code, API, or wire change.
- Dart wait cleared:
quest/m0/ffi-shutdown-backport.mddrops theWaitingitem on #8378 (verified merged 2026-11-11T10:08:11Z intorelease) and notes the tag push is ready. - Test-hang follow-up added:
quest/m0/test-lints-hang.mdis a new[XS]m0 quest forsh/rs/test-lints.test.pystalling after its first test. Both claims check out:just checkdoes reach it throughtest:gh) run just gh testinsh/dispatch.sh, and the script is in.github/justfile'stestrecipe. - Nixpkgs decision recorded:
quest/m2/cluster-routing/routes/playwright-pin.mdnow says an agent drafts NixOS/nixpkgs#585789's review fixes with anAssisted-by:trailer, vetted by the maintainer before anything is pushed or posted. - OSS-Fuzz wait settled:
quest/m4/oss-fuzz.mdreplaces the maintainer item with the 2026-11-11 decision;security@moq.devmatchesSECURITY.md:6.
ℹ️ The PR body's remaining OSS-Fuzz blocker is stale
The Description says the quest "stays blocked only on quest/m1/session-fuzz.md (#8506, queued)", but that path does not exist on main or on this head — the session targets the quest means live in rs/moq-net/fuzz/fuzz_targets/session_lite.rs and session_ietf.rs, and no quest file tracks them. #8506 also merged 2026-11-11T20:05:26Z, nearly three hours before this PR's head commit, so "queued" no longer holds either.
Technical details
# PR body cites a quest path and a PR state that no longer hold
## Affected sites
- PR Description, "Settled" bullet — names `quest/m1/session-fuzz.md`; `git ls-tree -r origin/main quest/m1/` has no such file, and neither does the head tree.
- Same bullet — says #8506 is "queued"; `gh pr view 8506` reports merged 2026-11-11T20:05:26Z.
## Required outcome
- The description names where the whole-session fuzz work is actually tracked (or says it landed) instead of a nonexistent quest path, so the next unblock run does not follow a dead link.deepseek-v4.1-flash-1017 (free via Pullfrog for OSS) | 𝕏
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
There was a problem hiding this comment.
ℹ️ No critical issues — one factual nit inline.
Reviewed changes
Re-reviewed the two commits since the prior pullfrog review (1df96ef..26b777a), which turn the earlier findings into records of what actually happened.
- Dart publish outcome recorded:
quest/m0/ffi-shutdown-backport.mdnow records thatmoq-dart-v0.1.1was pushed on #8378's merge andrelease-dart.ymlpublished nothing, and itsWaitingitem became the maintainer enabling pub.dev automated publishing — addressing the prior inline finding. - Trusted-publishing question answered:
quest/m3/release/dart-publish.md's "Confirm trusted publishing" step became "Configure trusted publishing", citing themoq-dart-v0.1.1tag as the evidence it is not enabled formoq-dev/moq. - Nixpkgs reply disclosure widened:
quest/m2/cluster-routing/routes/playwright-pin.md's decision now requires a separate AI-disclosure line in the posted reply alongside theAssisted-by:commit trailer — addressing the prior inline finding. - Hang diagnosis recorded:
quest/m0/test-lints-hang.mdnarrows the title to a busy host and blames the fail-fast test's ungatedjust rs testcalls, which reachsh/local.sh→quest gate. I checked the mechanism end to end:just rs testrunssh/rs/nextest.sh, which execssh/local.sh, and withoutLOCAL_SUPERVISED=1that ends insh/quest.sh gate;test_clippy_warning_fails_test_buildsets it and the fail-fast test does not.just checkreaches the script throughtest:gh) run just gh test, which runspython3 sh/rs/test-lints.test.pydirectly, so nothing upstream supplies the variable either.
deepseek-v4.1-flash-1017 (free via Pullfrog for OSS) | 𝕏
There was a problem hiding this comment.
ℹ️ No new commits since the last review — one carried-over factual nit still stands inline.
Reviewed changes
No commits landed since the prior Pullfrog review at 26b777a (git log 26b777a..HEAD is empty), so this run re-checked the five quest files at head and the single finding still open against them rather than a new delta.
- Dart publish record:
quest/m0/ffi-shutdown-backport.mdstill records the failedmoq-dart-v0.1.1publish (tag points at #8378's merge, pub.dev lists only 0.1.0) and waits on the maintainer enabling automated publishing. - Hang diagnosis:
quest/m0/test-lints-hang.mdstill attributes the stall totest_fail_fast_environment_reaches_nextestrunningjust rs testwithoutLOCAL_SUPERVISED=1, whichtest_clippy_warning_fails_test_builddoes set (sh/rs/test-lints.test.py:113). - Nixpkgs reply disclosure:
quest/m2/cluster-routing/routes/playwright-pin.mdstill requires both theAssisted-by:trailer and a separate AI-disclosure line in the posted reply. - OSS-Fuzz decision:
quest/m4/oss-fuzz.mdis unchanged at head;security@moq.devmatchesSECURITY.md:6.
deepseek-v4.1-flash-1017 (free via Pullfrog for OSS) | 𝕏
…d-run count Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
There was a problem hiding this comment.
ℹ️ No critical issues — one stale sentence in a quest intro that the new record now contradicts.
Reviewed changes
Re-reviewed the single commit since the prior pullfrog review (26b777a..eedcc3a), which closes the Dart wait and corrects the gated-run count.
- Dart publish completed: deleted
quest/m0/ffi-shutdown-backport.mdafter verifying pub.dev now listsmoq0.1.1 (published 2026-11-11T23:26:03Z) andrelease-dart.ymlrun 43107848443 succeeded on its rerun;grep -rn ffi-shutdown-backport quest/is clean andquest/m0/README.mdindexes no children, so the deletion leaves no dangling links. - Trusted-publishing fix recorded:
quest/m3/release/dart-publish.mdreplaces the "Configure trusted publishing" step with the swappedmoq/moq-devrepository field, the correctedmoq-dev/moqconfig, and the successful rerun. The recorded tag pattern,pub.devenvironment, and push trigger all match.github/workflows/release-dart.yml. - Gated-run count corrected:
quest/m0/test-lints-hang.mdnow says 20 runs (10 viajust rs test, 10 viash/rs/nextest.sh), matching the test's nine-case matrix plus--no-runper command. This closes the prior inline thread.
deepseek-v4.1-flash-1017 (free via Pullfrog for OSS) | 𝕏
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
There was a problem hiding this comment.
✅ No new issues found.
Reviewed changes
Re-reviewed the single commit since the prior pullfrog review (eedcc3a..ff3ab99), which fixes the one open finding.
- Plan intro refreshed:
quest/m3/release/dart-publish.mdnow saysmoqhas both 0.1.0 and 0.1.1 on pub.dev, describes 0.1.1 as published by a tag-workflow rerun, and narrows the open question to whether a tag publishes unattended with no rerun. This matches the bullet below it, andgh run view 43107848443still reportsconclusion: successon tagmoq-dart-v0.1.1viapush.
deepseek-v4.1-flash-1017 (free via Pullfrog for OSS) | 𝕏

Summary
/quest-unblockon 2026-11-11: records the maintainer's answers and clears a wait whose condition resolved.quest/m0/ffi-shutdown-backport.mdwaited on moq-dev/moq#8378 merging intorelease. It merged 2026-11-11 andmoq-dart-v0.1.1was pushed, but itsrelease-dart.ymlrun failed: pub.dev's GitHub Actions publishing named the repositorymoq/moq-dev(owner and name swapped). With the maintainer's approval, the coordinator set it tomoq-dev/moqin the maintainer's signed-in browser and reran the run, which published 0.1.1 (pub.dev lists it). The quest is deleted, andquest/m3/release/dart-publish.mdrecords the fix.quest/m4/oss-fuzz.md's maintainer item. The quest stays blocked only onquest/m1/session-fuzz.md(moq-dev/moq#8506, queued).quest/m2/cluster-routing/routes/playwright-pin.mdstill waits on nixpkgs, and now records how its review gets answered.Decisions
security@moq.dev, fromSECURITY.md(rejected: a personal address).Assisted-by:trailer and a separate AI-disclosure line in the reply, as nixpkgs' policy requires for review comments, and the maintainer reviews the diff and reply before anything is pushed or posted (rejected: the maintainer answers it; leave it).Checked and still waiting
cachix-retire,mbx-r2-verify(both date-gated: 2026-11-24 and 11-26),crates-trusted-publisher,demo-sockets-deploy,qmux-js-retire,qmux-rs-retire(all after the next cut).cargo miri teststill unsupported).runner-client-bumpandinterop-runner-approvalstill wait.nightly/macos.md's websocket test passed nightly since 11-06 and needs the week to finish (about 11-13); hakari's benchmark-tests job was cancelled again on 11-10 and 11-11.Impact
Public API: none. Wire: none. Quest text only.
Validation
quest check --base origin/mainpasses.just checkwas killed after 80 minutes intest gh:python3 sh/rs/test-lints.test.pyhangs after its first test, andtimeout 300 python3 sh/rs/test-lints.test.pyhangs the same way onorigin/main(f190d90), so this hang was already onmainand this PR did not cause it.Follow-ups
Models
Size: XS
Implementer: Claude Code, claude-opus-5-5, low effort (iterate, XS)
🤖 Generated with Claude Code
(Written by Claude Opus 5.5)
Review
Rewritten by
quest reviewafter each round.01a1ccf0-3274-7111-9004-cb34af76eca0, round 1The scoped diff changes only quest documentation and records maintainer decisions, with no actionable defects or blocking documentation issues.(Written by Claude Opus 5.5)