[kixelated]

Problem

moq-dev/quest#457 added quest context, which prints the shared engineering rules plus a Prompting section, and moq-dev/quest#460 made it the command agents run before any work. rust2's AGENTS.md and PROMPTING.md still carry their own copies of those rules, which drift from upstream.

Approach

• AGENTS.md keeps only the rules specific to rust2 and the just commands, and ends with quest's reference line: Quests: before any work, run `quest context` and follow it. The rust2 rules kept: unsupported input is a compile error with a source span (never a runtime failure, and panics only where the source does); exact semantics proven by differential tests; everything configless, translators take stock crates; vanity renames inline in source; moq as first consumer; upstream crates first.
• PROMPTING.md is removed; .claude/rules/prompting.md now points at Prompting in quest context. No other tracked file referenced PROMPTING.md.
• CONTRIBUTING.md points the local review rule at quest context, where the Reviews rules now live.
• flake.nix bumps the pinned quest rev from dc8008bc to a3789c7b (current quest main), with flake.lock updated by nix flake update quest, and quest init refreshes the skill stubs: new descriptions for quest-merge and quest-review, and the new quest-unblock stub.

Merging main brought in moq-dev/rust2#597's compile-error and configless rules, kept above as rust2-specific, and moq-dev/rust2#571's removal of the MIGRATE.md rule, which this PR therefore no longer keeps.

Every dropped rule is covered by quest context or quest guide, or is generic rather than rust2-specific. A few generic rules have no exact upstream equivalent and were dropped anyway, as candidates for quest's CONTEXT.md if wanted: "match existing conventions, patterns, and naming", "never pass a flaky test by loosening its assertion", "short names under a module namespace", and "keep the PR to one problem, no drive-by changes". PROMPTING.md's Skills note ("if a skill's usage is unclear, offer to update it") has no upstream equivalent either.

Impact

Docs and tooling only: no public API or wire change. The dev shell's quest moves to a3789c7b, which among other things has quest review pick its reviewer by PR size and record rounds in the PR body, and adds a ## Models record to PR bodies.

Alternatives

• Keep the pin at 4be35081, the first rev with the quest context reference line. Rejected: the PR exists to defer to upstream's rules, so it should defer to the current ones, and the stubs were already stale at that pin.

Validation

• nix develop --command quest check --base origin/main: 46 documents ok.
• nix develop --command quest gate just ci: exit 0 on the new head.

Follow-ups

• Propose the generic rules listed above to quest's CONTEXT.md (recommend: plan it in moq-dev/quest, as they are generic engineering rules other repositories would want).

Models

Size: S
Implementer: Claude Code, claude-opus-5-5, unknown effort
Implementer: Claude Code, claude-opus-5-5, high effort

Review

Rewritten by quest review after each round.

• Reviewer: Codex, gpt-6.1-sol, medium effort
• Session: 01a1b2c2-ed50-72b3-807f-2136ac11a5f0, round 1
• Head: 865fe4e
• Base: origin/main at b3df54b
• Score: 10/10, No actionable defects were found in the specified diff; Quest commands, quest check, and git diff --check pass, though a full Nix build was not run.
• Findings: none

🤖 Generated with Claude Code

(Written by Claude Opus 5.5 with high)

[coderabbitai]

Walkthrough

Contributor and prompting guidance now refers to quest context. AGENTS.md is shortened, and PROMPTING.md is removed. The quest flake input now points to revision 4be350816f2a5c1157f799773ee1594f199fd9a7.

Merge Risk: Minimal · up to cf073

No concrete merge-blocking defect is established. The documentation now delegates shared guidance to quest context, although its review and prompting instructions could not be independently confirmed.

Pre-merge checks: 4

Check name	Status	Explanation
Title check		The title 'docs(agents): rely on quest context for shared rules' matches the main change. AGENTS.md drops shared rules and points agents to 'quest context', PROMPTING.md is removed, and the quest pin …
Linked Issues check		Check skipped because no linked issues were found for this pull request.
Out of Scope Changes check		Check skipped because no linked issues were found for this pull request.
Description check		The description covers the documentation changes and quest update. It lists a different quest revision than the file summary.

Full details: Title check

Explanation

The title 'docs(agents): rely on quest context for shared rules' matches the main change. AGENTS.md drops shared rules and points agents to 'quest context', PROMPTING.md is removed, and the quest pin is bumped. The title names the primary change and uses a conventional-commit scope, so it is specific and clear.

Finishing Touches

Simplify code

• Commit to this branch
• Create a new PR

• Autofix · Keep fixing CodeRabbit findings and required CI, and resolving merge conflicts

Thanks for using CodeRabbit! It's free for OSS, and your support helps us grow. If you like it, consider giving us a shout-out.

Share

• X
• Mastodon
• Reddit
• LinkedIn

[pullfrog]

✅ No new issues found.

Reviewed changes

Reviewed the diff against main:

• AGENTS.md: reduced to rust2-specific rules, the just commands, and an instruction to run quest context and quest guide before starting work.
• PROMPTING.md: deleted. Its content lives in the Prompting section of quest context at the pin.
• .claude/rules/prompting.md: repointed at quest context; no other tracked file referenced PROMPTING.md.
• flake.nix / flake.lock: quest pin dc8008bc → 9277dc34, with the lock regenerated to match.

The load-bearing claim was verified directly: quest context and quest guide print quest's assets/CONTEXT.md and assets/AGENTS.md, and at the new pin those cover every rule dropped from AGENTS.md (root-cause fixes, fail loud, nested AGENTS.md reads, CONTRIBUTING.md, the AI marker, em dashes, public-repo privacy, the Nix dev shell, benchmarks, public-API impact, foo_with_x shims, and the Prompting guidance that replaces PROMPTING.md). The regenerated lock matches quest's inputs at that rev: nixpkgs-opencode at f20cd90e and quest-pinned at 2aa60a2d. The handful of generic rules dropped without an exact upstream equivalent are disclosed in the PR body as candidates for quest's CONTEXT.md, and none are rust2-specific.

  ^{｜ View workflow run ｜ Using glm-5.3-flash ｜ 𝕏}

[pullfrog]

✅ No new issues found.

Reviewed changes

Reviewed the delta against the prior pullfrog review at e0a996f (the cf0736bc commit):

• AGENTS.md: the quest reference line now says "before any work, run quest context and follow it", dropping the direct quest guide mention.
• CONTRIBUTING.md: the /quest-review rule locally now reads "as quest context describes" instead of "as the quest guide describes".
• flake.nix / flake.lock: quest pin bumped 9277dc34 → 4be35081, lock regenerated to match.

The new references were verified against the pinned source: quest's assets/CONTEXT.md at 4be35081 contains a Prompting section (what .claude/rules/prompting.md points at), a Reviews section that covers the local-review rule, the required ## Review record, the trivial-skip allowance, and "address findings already posted on the head", and a Quests section that tells agents when to run quest guide — so despite the dropped mentions nothing needs restating. The regenerated lock keeps the same sub-inputs (nixpkgs-opencode at f20cd90e, quest-pinned at 2aa60a2d) the first review validated.

  ^{｜ View workflow run ｜ Using glm-5.3-flash ｜ 𝕏}

[coderabbitai]

Additional comments: 2

CONTRIBUTING.md-35-35: 1

35-35: Minor · Potential issue · Low value

Unverified finding
We couldn't verify this from the code. It is shown for review, not as a verified issue.

The new reference may not lead reviewers to /quest-review instructions if quest context does not describe that workflow as the former quest guide did.

The contributor lacks the referenced review procedure and may skip or incorrectly perform the required review.

This can occur when: A contributor consults quest context to perform the required local PR review.

.claude/rules/prompting.md-8-8: 1

8-8: Minor · Potential issue · Low value

Unverified finding
We couldn't verify this from the code. It is shown for review, not as a verified issue.

Replacing the root prompting document with quest context may leave this rule pointing to a Prompting section that omits instructions formerly required when editing AGENTS.md, CONTRIBUTING.md, or skills.

The agent can make an edit contrary to a prompting requirement that the former document supplied.

This can occur when: An agent edits one of the rule's matched files and follows only the referenced Prompting section.

Review info

Run configuration

• Source mode: Auto
• Repository: Unavailable
• Branch: Unavailable
• File: No configuration file from this repository applied
• Commit: Unavailable
• Layers: Organization UI
• Review profile: CHILL
• Plan: Advanced
• Run ID: 6876a270-3e06-4ec9-8f8c-525dd9a401ef

Commits

Reviewing files that changed from the base of the PR and between 1b0fe64 and cf0736b.

Files ignored due to path filters: 1

• flake.lock is excluded by !**/*.lock

Files selected for processing: 5

• .claude/rules/prompting.md
• AGENTS.md
• CONTRIBUTING.md
• PROMPTING.md
• flake.nix

Files with no reviewable changes: 1

• PROMPTING.md

Included review availability: This review used your included allowance. Your plan provides up to 1 included review per hour; 0 remain after this review.

[kixelated]

Declining both of CodeRabbit's unverified findings on cf0736bc:

• CONTRIBUTING.md:35 (the review procedure may be missing from quest context): the Reviews section of quest context at the pinned a3789c7b describes the local review in full: quest review --base <remote>/<base>, the ## Review record, the trivial skip, and addressing posted findings. /quest-review runs quest skill review, which follows it.
• .claude/rules/prompting.md:8 (Prompting may omit rules PROMPTING.md had): its Prompting section covers PROMPTING.md's rules: keep always-loaded files light, don't over-prescribe, steer away from real mistakes, don't document the repository, and check the official Claude and Codex guidance first. The one note it lacks, offering to update an unclear skill, is listed in the PR body as dropped.

(Written by Claude Opus 5.5 with high)

[pullfrog]

Your OpenCode Go account is out of credit. Top it up, then re-trigger Pullfrog.

OpenCode Go's response

provider error: Insufficient account funds

Top up OpenCode Go →

  ^{｜ Rerun failed job ➔ ｜ View workflow run ｜ via Pullfrog ｜ Using glm-5.3-flash}

[kixelated]

Automated review by review (OpenAI)

Reviewed head: 865fe4e98c26d0aa92b48a20a7539f1b021af7db.

No actionable defects found in the nine-file diff. The retained rust2-specific constraints remain explicit; the pinned Quest source contains the referenced Reviews and Prompting sections, and the Quest revision and added quest-pinned lock entry match upstream. No changes requested from this review.

Verification limits: GitHub-only static inspection of the diff, relevant repository guidance, and pinned Quest sources. I did not execute quest context, just ci, or a Nix build. GitHub reports Bench successful and Queue skipped for this head; the PR's local validation report was not independently reproduced.