← ClaudeAtlas

review-and-fixlisted

Use when you need findings on a PR or current branch to be auto-applied, not just reported.
The01Geek/devflow-autopilot · ★ 14 · Code & Development · score 76
Install: claude install-skill The01Geek/devflow-autopilot
# /devflow:review-and-fix — Review, Fix, and Verify Loop You are the review-and-fix orchestrator. Run /devflow:review's review engine, fix the findings it surfaces, and re-run until the engine returns a clean verdict. **Input:** `$ARGUMENTS` may contain an optional PR number, the flag `--push-each-iteration`, and/or the flag `--issue N`. Parse the three independently — any, all, or none may be present. If no PR number is given, review and fix the current branch. The numeric token (if any) is `$PR_NUMBER`; the `--issue` value (if any) is `$ISSUE_OVERRIDE`. **Every later PR-mode predicate and every `gh` command reads `$PR_NUMBER` — never the raw `$ARGUMENTS` string**, which fails an is-a-PR-number test and leaks the flag tokens into any command it is interpolated into. **`--push-each-iteration` (default off).** When set, each fix and Loop Exit use the gated base checkpoint and push, keeping the PR and CI current. Otherwise fix commits stay local and the base is untouched. The flag never posts a GitHub verdict; mandatory telemetry persistence remains independent. Loop correctness uses local `HEAD` (with the PR head override), not pushed state. **Key principle:** You perform fixes DIRECTLY in this session. Do NOT delegate fixes to a subagent. You need full conversation context to apply `devflow:receiving-code-review` principles (technical evaluation, pushback, verification). **Portable helper anchor (single-statement).** The bundled-helper commands in this skill resolve the