[kixelated]

Summary

A transcript audit of moq-dev/moq and moq-dev/moq.pro sessions (396 sessions from Oct 17-19, 2026; about 339k Bash commands from Sep-Oct) found agents re-implementing quest workflow steps as shell loops and jq one-liners. This plans a new m0 epic, quest/m0/helpers/, ranked after the worktree and gate quests and before the Kix decision UI:

• [M] status.md: quest status <pr> (plain and --json): head, checks with requiredness, merge, auto-merge and queue state, reviews current vs stale, reactions, and unresolved threads with IDs. 219 sessions.
• [M] move.md: quest mv / quest rm via git mv / git rm, rewriting or dropping root-absolute links under quest/. 203 sessions.
• [M] wait-checks.md: quest wait emits checks:failed / checks:passed for the head's required checks, so the merge skill drops its separate check watcher. 119 sessions.
• [S] claim.md: quest claim, quest claim --release, and quest unblocked --unclaimed. 105 sessions.
• [S] gate-lock.md: quest gate --lock <name> replaces moq.pro's mkdir merge mutex (79 sessions). Starts after kixelated/quest#128 lands gate.ts.
• [XS] afk.md: an afk argument for /quest-complete and /quest-spawn that asks only about merges and deletes.

Dropped as not generic: replying to and resolving review threads, and parsing open decisions from a PR body. The 1,731 load-average polling commands are recorded in worktree-agents.md. I did not edit local-gate.md, because kixelated/quest#128 rewrites it. Both quests already rank first in m0.

Planning decisions

Goal: agents stop hand-writing gh/jq/sed loops because quest commands cover the routine steps. GitHub code stays in src/cli.

• ✅ Yes, only the really generic stuff (user)
• Narrower

GitHub helpers:

• ✅ CI checks in wait (Recommended)
• ✅ quest status (Recommended)
• ✅ Claim check (Recommended)
• Reply/resolve threads (Recommended)

Local helpers:

• ✅ quest mv / quest rm (Recommended)
• ✅ Named gate, limit 1 (Recommended)
• ✅ Overnight triage mode (Recommended)

Placement:

• ✅ m0 epic, before Kix (Recommended)
• m0 epic, after Kix
• Loose m0 quests

Checks in wait:

• ✅ Default events (Recommended)
• --checks mode
• Both

Status contents:

• ✅ Head, checks, merge (Recommended)
• ✅ Reviews by head (Recommended)
• ✅ Unresolved threads (Recommended)
• Body open decisions

Claims:

• ✅ quest claim + --unclaimed (Recommended)
• Only --unclaimed filter
• quest claims listing

mv / rm:

• Filesystem, quest/ links (Recommended)
• ✅ git mv / git rm
• Rewrite repo-wide

Named gate:

• ✅ quest gate --lock (Recommended)
• --name --limit
• Leave it to repos

Overnight mode:

• ✅ Shared afk argument (Recommended)
• complete only
• Make it the default

Which checks count:

• ✅ All on the head (Recommended). The user asked whether a failing non-required check wakes the wait; see the next decision.
• Required only

Optional check failure:

• ✅ Required only (Recommended): only required checks wake the wait (requiredness from GraphQL isRequired), and a repo with no required checks treats every check as required
• Any failure wakes
• Wake, but tag it

Claim release:

• ✅ Yes, --release (Recommended)
• No

Docs:

• ✅ No, inline only (Recommended)
• Yes, a docs quest

Order:

• ✅ By session count (Recommended)
• Cheapest first

Checks

quest check, just check and just test pass.

🤖 Generated with Claude Code

(Written by Claude Opus 5.5)

[chatgpt-codex-connector]

Codex Review Summary

This comment shows the latest Codex review activity on this pull request.

Review	Status	Commit	Review trigger
📝 Code Review	✅ Completed 2026-10-19T17:42:11.245400Z	8f3c0a2	New commits
🔒 Security Review	⚠️ Failed 2026-10-19T17:45:14.210442Z	8f3c0a2	New commits

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" or "@codex security review".

Codex reacts with 👀 while any review is running, comments if it has suggestions, and reacts with 👍 once all reviews finish with no findings.

[chatgpt-codex-connector]

💡 Codex Review

Here are some automated review suggestions for this pull request.

Reviewed commit: f053efaeca

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".

[chatgpt-codex-connector]

💡 Codex Review

Here are some automated review suggestions for this pull request.

Reviewed commit: 8f3c0a27e2

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".

[chatgpt-codex-connector]

Handle prose links before deleting their target

When a deleted quest is referenced by an inline Markdown link in a Goal or Plan, rm only drops linked list items and this clause leaves the prose reference unchanged. Deleting the target then makes quest check report a dangling link, contradicting the command’s promise to avoid creating broken links; update or remove these links, or refuse the deletion before staging it, rather than merely reporting them.

AGENTS.md reference: assets/AGENTS.md:L137-L140

Useful? React with 👍 / 👎.

[kixelated]

Addressed Codex's two findings in 8f3c0a2: quest mv now refuses to move a claimed quest (claim branch or open PR), and each quest claim commit carries a random nonce so concurrent claimers can't both succeed. The decisions are listed in the PR body. CI passes on this head.

(Written by Claude Opus 5.5)