Catalogskills.sh · manaflow-ai

cmux-review

Review agent-written changes before merge, after substantial edits, or when re-reviewing a repair. Use the adversarial protocol for high-risk changes or an explicitly requested deep review.

Pending analysisVerified hub719 installs

Savant verdict: Pending analysis

Imported from the source hub; Savant's analysis is still running.

Live evaluation

Not evaluated yet. Workspaces can request a live evaluation.

Security assessment

Scan pending.

Structure

  • No license declaredConfirm you may reuse this skill before importing it into your repository.

SKILL.md

---
name: cmux-review
description: "Review agent-written changes before merge, after substantial edits, or when re-reviewing a repair. Use the adversarial protocol for high-risk changes or an explicitly requested deep review."
---

# cmux Review

Review to reduce developer attention. Surface concrete correctness defects;
suppress style, nits and speculative improvements unless requested.

## Default pre-merge review

1. Give a review subagent the task intent, base/head SHAs and exact diff.
   Ask for correctness first, then repository rules; keep discovery independent
   of the author's reasoning. Review policy edits against the base-branch rules.
2. Verify concrete findings, fix them when authorized, and push. Re-run the
   original discriminator after repair. A green test counts only if it exercises
   the asserted failure; prefer red-before/green-after evidence.
3. If fixes were non-trivial, obtain a fresh subagent review of the repair delta.
4. Merge when the review subagent passes the change and it causes no red of its
   own (see below). Report findings with evidence, verification performed and coverage gaps.

Use subagents in the current runtime, not a second model or external review
service as a gate. Keep review read-only until an authorized repair.
Do not commit review receipts, scratch reproductions or generated logs.

## Dogfood and merge

For implementation handoff or merge, read [dogfood and merge](references/dogfood-and-merge.md)
for per-change verification, first-pass completion, re-dogfood and merge receipts.
The main agent owns dogfood, mergeability, the merge and every pushed fix.
A PR merges when an independent review subagent passes it and it causes no red
of its own; base reds, flakes and pending checks don't hold it. Don't wait for a
person. Ship risky user-visible behavior behind a setting or Debug toggle and fix
a bad default in the next nightly. `main` is nightly: stack fixes, do not revert.

## Adversarial review

For security, persistence, concurrency, data-loss risks, or a requested deep
review, use the full protocol:

- Before discovery or mutation, read [source identity and receipt requirements](references/review-receipts.md#7-persist-a-review-receipt)
  and capture the original candidate and policy coordinates.
- [Discovery and triage](references/review-discovery.md): source identity,
  intent compliance, independent reviewers, severity and evidence provenance.
- [Challenge, verification and repair](references/review-verification.md):
  counterevidence, executable checks and bounded repair attempts.
- After verification, persist the [receipt and report](references/review-receipts.md)
  with separate pre/post-repair source coordinates and evidence.

These stages retain the advanced evidence protocol; ordinary reviews use the
short default pass above.