← ClaudeAtlas

mr-fixlisted

Fetch unresolved code review comments from a GitLab MR or GitHub PR, plan the fixes, implement them after approval, verify, self-review the diff in a bounded fix loop, and commit. Use this skill whenever the user mentions fixing, addressing, resolving, or implementing MR comments, PR comments, review findings, review feedback, reviewer suggestions, or code review results — and whenever they type /mr-fix, /mr-fix 298, or /mr-fix MR 298. Also use it when the user pastes review comments and asks to act on them, or says things like "address the review on 298", "fix what the reviewer flagged", or "handle the MR feedback".
rdlugs/ai-skills · ★ 0 · Code & Development · score 60
Install: claude install-skill rdlugs/ai-skills
# mr-fix Turns code review feedback into reviewed, formatted, committed changes. Language- and environment-agnostic: it detects the forge, the container setup, and the toolchain before it touches anything. ## Step 0 — Preflight (detect everything; assume nothing) Run this block first. It answers every environmental question the rest of the skill depends on. Nothing below is hardcoded — if a check comes back empty, the skill adapts rather than failing. ```bash # --- Forge: which platform, which CLI, is it authed? --- git remote get-url origin git branch --show-current git status --porcelain command -v glab && glab auth status 2>&1 | tail -2 command -v gh && gh auth status 2>&1 | tail -2 # --- Runner: is this project containerized? --- ls docker-compose.yml docker-compose.yaml compose.yml compose.yaml 2>/dev/null docker compose ps --services --status running 2>/dev/null # --- Toolchain: what actually exists in this repo? --- ls composer.json package.json vendor/bin/pint vendor/bin/php-cs-fixer 2>/dev/null ls pyproject.toml uv.lock poetry.lock Pipfile requirements.txt setup.py tox.ini 2>/dev/null ls -d .venv venv 2>/dev/null command -v uv poetry pipenv hatch pdm rye conda 2>/dev/null command -v oco ``` If `pyproject.toml` exists, read it — `[tool.ruff]`, `[tool.black]`, `[tool.mypy]`, `[tool.pytest.ini_options]` tell you what the project actually uses, which beats guessing from what happens to be installed globally. Derive four facts, then reuse them everywhere: **