菜单
Review automated PR feedback, discuss with the user, fix what they approve, respond to comments, and push
用 Codex 或 Claude 帮你安装 复制这段 Prompt，粘贴到 Codex、Claude 或其他助手里，让它检查 Skill 页面并帮你完成安装。
基于 SOC 职业分类
| name | address-pr-feedback |
| description | Review automated PR feedback, discuss with the user, fix what they approve, respond to comments, and push |
| disable-model-invocation | false |
| argument-hint | [pr-number] |
Address automated review feedback on PR #$ARGUMENTS.
Fetch review comments using /pr-comments first, then supplement with both GitHub API endpoints to ensure full coverage:
gh api repos/{owner}/{repo}/pulls/{pr}/comments — inline review comments on specific linesgh api repos/{owner}/{repo}/issues/{pr}/comments — general conversation comments (where automated review bots like the CI code review post)Combine results from all sources, deduplicating and ignoring bot comments that aren't code review feedback (e.g., Vercel deployment status, coverage reports).
Summarize each suggestion with your assessment of whether it's worth fixing. Present as a numbered table with columns: suggestion, your take (fix/skip + rationale). Be opinionated.
Discuss with the user. Wait for them to tell you which suggestions to act on and which to skip. They may have additional context or disagree with your assessment.
Implement the agreed fixes. Make the code changes, run type-check and tests to verify.
Respond to the review comments on the PR via gh api repos/{owner}/{repo}/issues/{pr}/comments. Mention that you're Claude speaking on @thavelick's behalf. Address each numbered suggestion: what was done and why, or why it was skipped.
Commit and push the changes.
Review the changes since a fixed point (commit, branch, tag, or merge-base) along two axes — Standards (does the code follow this repo's documented coding standards?) and Spec (does the code match what the originating issue/PRD asked for?). Runs both reviews in parallel sub-agents and reports them side by side. Use when the user wants to review a branch, a PR, work-in-progress changes, or asks to "review since X".
Shared vocabulary for designing deep modules. Use when the user wants to design or improve a module's interface, find deepening opportunities, decide where a seam goes, make code more testable or AI-navigable, or when another skill needs the deep-module vocabulary.
Diagnosis loop for hard bugs and performance regressions. Use when the user says "diagnose"/"debug this", or reports something broken/throwing/failing/slow.
Build and sharpen a project's domain model. Use when the user wants to pin down domain terminology or a ubiquitous language, record an architectural decision, or when another skill needs to maintain the domain model.
Set up Claude Code hooks to block dangerous git commands (push, reset --hard, clean, branch -D, etc.) before they execute. Use when user wants to prevent destructive git operations, add git safety hooks, or block git push/reset in Claude Code.
Grill the user relentlessly about a plan, decision, or idea. Use when the user wants to stress-test their thinking, or uses any 'grill' trigger phrases.