skills/review-pr/SKILL.md
Use when you need to review or update the latest open GitLab merge request or GitHub pull request through a local worktree, or when a user provides a specific MR or PR number.
npx skillsauth add minakoto00/vcs-review-flow review-prInstall this skill globally with one command. Works with Claude Code, Cursor, and Windsurf.
3 of 9 scanners reported clean
Some scanners were skipped, did not run, or reported a non-clean status. Review each row below.
Use this skill when the user wants you to pull, review, or update a remote GitLab MR or GitHub PR through a local worktree instead of switching the current checkout.
gitjqgh for GitHub repositoriesglab for GitLab repositoriesgit rev-parse --show-toplevel
bash scripts/detect_platform.sh --repo <repo-root>
Latest open MR/PR:
bash scripts/resolve_review_target.sh --repo <repo-root> --latest
Specific MR/PR number:
bash scripts/resolve_review_target.sh --repo <repo-root> --number <number>
bash scripts/fetch_review_comments.sh --repo <repo-root> --number <number> --platform <github|gitlab> --json
Rules:
code-review comments exist, ask the user whether to include code-review comments in scope before continuingdiscussion comments exist, ask the user whether to include discussion comments in scope after the code-review comment decisionReview full changeset for additional issuesDo not do additional reviewsSpecify otherwisebash scripts/repo_policy.sh --repo <repo-root>
Read both AGENTS.md and CLAUDE.md when present. If they conflict on worktree location or workflow, surface that conflict instead of guessing.
bash scripts/worktree_sync.sh \
--repo <repo-root> \
--source-branch <source-branch> \
--head-sha <head-sha>
Rules:
For approved comments:
1. Review full changeset for additional issues.
2. Do not do additional reviews.
3. Specify otherwise.
Review full changeset for additional issues, run a full code review session from the prepared worktree.If the environment exposes a review skill such as requesting-code-review, invoke it with:
If no dedicated review skill is available, perform the review directly with a code-review mindset across the full changeset.
Keep any new findings separate from the remote-comment verification report.
The change plan must include:
If the user confirmed approved comments as valid, include them in the change plan even when tests do not yet cover them. If the user requested a full changeset review, include any additional findings from that review in the change plan alongside the confirmed remote-comment issues.
1. Implement locally on the original source branch, commit, and push to update the MR/PR.
2. Implement locally on the original source branch, commit there, then merge or cherry-pick into a user-specified branch.
3. Do not implement locally; post the proposed changes as a GitLab/GitHub comment with concrete patch guidance.
gh or glab is missing for the detected platform.Detect platform:
bash scripts/detect_platform.sh --repo <repo-root>
Resolve latest target:
bash scripts/resolve_review_target.sh --repo <repo-root> --latest
Resolve explicit target:
bash scripts/resolve_review_target.sh --repo <repo-root> --number 123
Inspect repo policy:
bash scripts/repo_policy.sh --repo <repo-root>
Sync worktree:
bash scripts/worktree_sync.sh --repo <repo-root> --source-branch feat/example --head-sha <sha>
Fetch review comments:
bash scripts/fetch_review_comments.sh --repo <repo-root> --number 123 --platform github --json
Post comment-only proposal:
bash scripts/post_review_comment.sh --repo <repo-root> --number 123 --body-file /tmp/review-plan.md
testing
Create, edit, improve, or audit AgentSkills. Use when creating a new skill from scratch or when asked to improve, review, audit, tidy up, or clean up an existing skill or SKILL.md file. Also use when editing or restructuring a skill directory (moving files to references/ or scripts/, removing stale content, validating against the AgentSkills spec). Triggers on phrases like "create a skill", "author a skill", "tidy up a skill", "improve this skill", "review the skill", "clean up the skill", "audit the skill".
testing
Host security hardening and risk-tolerance configuration for OpenClaw deployments. Use when a user asks for security audits, firewall/SSH/update hardening, risk posture, exposure review, OpenClaw cron scheduling for periodic checks, or version status checks on a machine running OpenClaw (laptop, workstation, Pi, VPS).
testing
Create, edit, improve, or audit AgentSkills. Use when creating a new skill from scratch or when asked to improve, review, audit, tidy up, or clean up an existing skill or SKILL.md file. Also use when editing or restructuring a skill directory (moving files to references/ or scripts/, removing stale content, validating against the AgentSkills spec). Triggers on phrases like "create a skill", "author a skill", "tidy up a skill", "improve this skill", "review the skill", "clean up the skill", "audit the skill".
testing
Host security hardening and risk-tolerance configuration for OpenClaw deployments. Use when a user asks for security audits, firewall/SSH/update hardening, risk posture, exposure review, OpenClaw cron scheduling for periodic checks, or version status checks on a machine running OpenClaw (laptop, workstation, Pi, VPS).