Add repo-auditor and pr-ci-fixer review subagents

[BigSimmo]

Summary

• Follow-up to the core six subagents (merged in Add core clinical-RAG review subagents #599). Adds two optional review specialists under .claude/agents/:
  • repo-auditor (Sonnet, read-only) — structural/dead-code/import triage. Findings are candidates only, never an automatic delete list; verifies against routes, scripts/*, migrations, worker/**, and tests before proposing removal.
  • pr-ci-fixer (Sonnet, read-only) — diagnoses CI/check failures and proposes the smallest fix. Reruns, PR comments, pushes, and any GitHub/provider action stay confirmation-required per the AGENTS.md pr-ci-fix routing and the API/provider boundary.
• Both inherit the AGENTS.md review-throttling protocol (docs/codex-review-protocol.md, docs/branch-review-ledger.md) and the provider-confirmation boundary. Tools limited to Read, Grep, Glob, Bash.

Verification

• npm run verify:pr-local

Docs-only change: adds two .claude/agents/*.md definition files. No source, ingestion, retrieval, ranking, answer-generation, privacy, Supabase, or UI runtime code is touched — so the RAG/answer eval gates and check:production-readiness do not apply.

• npm run verify:ui — n/a (no UI/routing/styling change)
• npm run verify:release — n/a (not a release/handoff PR)

Clinical Governance Preflight

n/a — this change does not touch ingestion, answer generation, search/ranking, source rendering, document access, privacy, production env, or clinical output. It only adds Markdown agent definitions that encode the existing governance/provider boundaries as guidance for future reviews.

Notes

• Applied per the merged-PR workflow: since Add core clinical-RAG review subagents #599 merged, claude/subagents-clinical-rag-uqa5aa was restarted from the updated main (which now contains the six core agents) and this work committed on top, then force-pushed with --force-with-lease over the already-merged tip. This is a new PR, not a reuse of Add core clinical-RAG review subagents #599.

🤖 Generated with Claude Code

---

Generated by Claude Code

[supabase]

This pull request has been ignored for the connected project sjrfecxgysukkwxsowpy because there are no changes detected in supabase directory. You can change this behaviour in Project Integrations Settings ↗︎.

---

Preview Branches by Supabase.
Learn more about Supabase Branching ↗︎.

[coderabbitai]

Warning

Review limit reached

@BigSimmo, you've reached your PR review limit, so we couldn't start this review.

Next review available in: 2 minutes

Enable usage-based reviews in Billing to review now. Otherwise, wait until the next included review is available.
You're only billed for reviews past your plan's rate limits ($0.25/file).

How can I continue?

After more reviews become available, a review can be triggered using the @coderabbitai review command as a PR comment. Alternatively, push new commits to this PR.

To avoid repeated limits, reduce automatic review volume by pausing incremental auto-reviews earlier, using label-based review opt-in, excluding WIP or generated PR titles, or requesting reviews manually when the PR is ready. If your team needs uninterrupted high-volume reviews, an organization admin can enable usage-based reviews.

How do review limits work?

CodeRabbit enforces per-developer PR review limits for each organization. Most developers receive the normal plan review availability.

For paid Pro and Pro+ PR reviews, CodeRabbit uses adaptive limits for sustained high-volume activity. When a developer's recent PR review activity reaches the 95th percentile or higher among CodeRabbit users, additional reviews become available more gradually as earlier reviews age out of the rolling window.

Please refer docs for additional details.

Review details

⚙️ Run configuration

Configuration used: Organization UI

Review profile: CHILL

Plan: Pro Plus

Run ID: c01c46fa-5ac9-4f35-9800-873ee69ed667

📥 Commits

Reviewing files that changed from the base of the PR and between 1b5ec7f and 848c15d.

📒 Files selected for processing (2)

• .claude/agents/pr-ci-fixer.md
• .claude/agents/repo-auditor.md

✨ Finishing Touches

🧪 Generate unit tests (beta)

• Create PR with unit tests
• Commit unit tests in branch claude/subagents-clinical-rag-uqa5aa

✨ Simplify code

• Create PR with simplified code
• Commit simplified code in branch claude/subagents-clinical-rag-uqa5aa

---

_{Comment @coderabbitai help to get the list of available commands.}

[chatgpt-codex-connector]

💡 Codex Review

Here are some automated review suggestions for this pull request.

Reviewed commit: 848c15dd3d

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]

Require confirmation for GitHub log reads

When the check log is not already available locally, the diagnose step will naturally use PR/CI tooling such as gh run view --log, but this hard-rule list only requires approval for GitHub/GitLab API writes. That conflicts with the repo boundary in AGENTS.md lines 191-194, which makes any GitHub/GitLab or hosted-CI interaction confirmation-required; make read-only log/check access explicitly confirmation-required before telling the agent to read failing logs.

Useful? React with 👍 / 👎.