원클릭으로
devpilot-pr-review
Use when the user asks to review a pull request, merge request, or a diff — "review this PR", "review PR
Codex 또는 Claude로 설치 이 Prompt를 복사해 Codex, Claude 또는 다른 어시스턴트에 붙여 넣으면 Skill 페이지를 검토하고 설치를 진행할 수 있습니다.
메뉴
Use when the user asks to review a pull request, merge request, or a diff — "review this PR", "review PR
Codex 또는 Claude로 설치 이 Prompt를 복사해 Codex, Claude 또는 다른 어시스턴트에 붙여 넣으면 Skill 페이지를 검토하고 설치를 진행할 수 있습니다.
SOC 직업 분류 기준
Use when the agent keeps making the same mistake across multiple PRs, when lint/test failures recur on the same rule, when PR review comments reveal a missing guide or unenforced convention, or when you want to proactively evolve the harness. Triggers on "self-evolve", "optimize harness", "update harness docs", "agent keeps making the same mistake", "harness self evolution", "harness 自我进化", "优化 harness", "更新 harness 文档".
Turn a single article, document, or web page into a standalone HTML learning artifact: either a bilingual digest or a chaptered Chinese study guide with source summaries. MUST use this skill — not raw summarization — whenever the user asks to summarize, digest, get key points from, extract takeaways from, or turn a URL, PDF, .docx, .md, .txt, paper, report, or blog post into notes or study material. Trigger on: summarize this, key points, takeaways, digest, highlights, TL;DR, what does it say, overview, 学习资料, 学习笔记, 复习资料, 总结, 摘要, 要点, 重点, 提取, 整理, 讲了什么. NOT for: writing new content, translating, comparing two documents, building apps, multi-source news, code review, or data/CSV analysis.
Use proactively when harness documentation needs updating — new module/package added, skill created or removed, CLI surface changed, exec plan completed, new convention established, or CLAUDE.md exceeds ~100 lines. Also use when asked to set up AGENTS.md/skills/sub-agents, fix agent drift, or retrofit guardrails.
Use when setting up a repository for autonomous coding agents, adding guardrails, context files, or automation so agents ship reliably without constant review. Triggers on "make this repo agent-friendly", "agents keep drifting", "set up AGENTS.md / skills / sub-agents", "harness engineering", architectural drift with agent-authored code, or retrofitting guardrails after output quality decayed.
Use when the user wants to resolve, fix, work through, or burn down open GitHub issues in a repository — "fix all the issues", "resolve these tickets", "work through the repo-scan issues", "clear the backlog", "fix issue
Use when the user asks to scan, audit, or sweep an entire GitHub repository for issues and file them as tickets — "scan this repo", "audit the codebase", "find bugs/security holes/missing tests", "check the docs are still accurate", "/repo-scan", "open issues for all the problems you find". Scans security, edge cases, testing coverage, and doc/code drift (CLAUDE.md, AGENTS.md, README.md and the docs they link to) without assuming business logic. Do NOT use for reviewing a single PR (use devpilot-pr-review) or language-specific style review (use devpilot-google-go-style).
| name | devpilot-pr-review |
| description | Use when the user asks to review a pull request, merge request, or a diff — "review this PR", "review PR |
Most PR review fails by staying inside an already-narrow option set: naming, formatting, "LGTM". This skill pushes the review onto the behavior the PR introduces into the system, including behavior neither the author nor the reviewer has noticed yet.
<coverage_first_findings>
Report every finding you reach after tracing the code, including ones you are uncertain about or judge low-severity. Your job at this stage is coverage, not filtering. Each finding carries its own Confidence and Severity so the author and any downstream reviewer can rank and filter. A finding that later gets filtered out is fine; a finding that was silently dropped because it felt minor is not.
</coverage_first_findings>
<investigate_before_asserting>
State how the code behaves only after opening and reading the relevant files. When a finding depends on a caller or test you have not located, mark it Confidence: low and record the gap in Open Questions rather than speculating.
</investigate_before_asserting>
gh pr view <url> --json title,body,files,baseRefName,author + gh pr diff <url>; or git diff <base>...HEAD for a local branch; or read a pasted patch directly. A PR with no stated intent is itself a finding.references/unknown-unknowns.md (also covers the paired Behavior Trace step).references/template.md (includes the severity rubric and version-resolution rules). Apply tone, stance, and language rules from references/style.md. Calibrate against references/example-review.md on first use.references/posting.md.Before posting, walk the self-check in references/rationalizations.md.
devpilot-clean-code-principles.devpilot-google-go-style.| File | What's in it |
|---|---|
references/unknown-unknowns.md | The five blind-spot questions, per-change-class pitfall table, paired Behavior Trace step, output format. |
references/template.md | The review skeleton, per-field rules, version resolution, severity rubric. |
references/style.md | Language / tone / stance rules for the posted review. |
references/posting.md | gh / glab commands, severity-to-mode mapping, skip conditions, inline-comment API. |
references/example-review.md | Fully-filled worked example for calibration. |
references/rationalizations.md | Common shortcuts with rebuttals, plus the pre-post self-check list. |