Token导航 LogoToken导航TokenDH.com
研究检索只读github未标认证来源可访问clear审计未展示

pr-review公关审查

Agent Skill

pr-review 用于查找、检索和筛选相关信息,适合在 Codex、Claude、Cursor、Gemini CLI 中需要根据关键词、任务场景或来源线索快速定位候选结果时使用。可结合来源仓库、安装命令和原始 README 继续核验具体用法。安装前建议确认权限范围、维护状态,以及是否会触发联网、命令执行或文件读写。

总安装

441

周安装

18

GitHub Stars

公开资料未说明

下载量

143
CodexClaudeCursorGemini CLI

安装说明

本站只整理中文说明和来源信息,不托管安装包,也不代用户安装。

GitHub

来源数

2

许可证

MIT

最后核验

2026-05-01

来源状态

来源可访问

安装方式

通过对话安装

复制提示词发给支持本地命令或 Skills 的 AI 助手,先确认命令和权限,再让它执行。

请帮我安装这个 Agent Skill:pr-review(公关审查)
来源仓库:https://github.com/saturate/claude
仓库路径:skills/pr-review
安装命令:
npx skills add saturate/claude --skill "pr-review"
安装前请先检查当前环境是否支持对应 CLI,并向我确认将要执行的命令、安装目录、联网范围和文件读写权限;确认后再执行。

命令行安装

复制命令到本机终端执行。该命令会通过 npx skills 从第三方来源获取 Skill;本站只展示命令,不托管安装包,也不自动执行。

AgentSkills.tonpx skills
npx skills add saturate/claude --skill "pr-review"

简介

pr-review 用于查找、检索和筛选相关信息, 适合在 Codex、Claude、Cursor、Gemini CLI 中需要根据关键词、任务场景或来源线索快速定位候选结果时使用。

  • 可结合来源仓库、安装命令和原始 README 继续核验具体用法。
  • 安装前建议确认权限范围、维护状态, 以及是否会触发联网、命令执行或文件读写。

SKILL.md

name
pr-review
description
Performs comprehensive code reviews checking for bugs, security issues, performance problems, testing gaps, and code quality. Use when reviewing PRs, code changes, or when asked to review code.
compatibility
Basic tools only - grep, file reading
allowed-tools
Read Grep Glob Bash
metadata
author
Saturate
version
1.0

Code Review

Review code like a senior engineer - thorough but practical. Focus on things that actually matter. Don't waste time on style nitpicks a linter should catch.

Review Checklist

Use this checklist to guide your review. Need examples of what to look for? Check out references/common-issues.md for code patterns.

Security (Critical)

  • [ ] Input validation and sanitization
  • [ ] SQL injection, XSS, command injection risks
  • [ ] Auth checks in place and correct
  • [ ] Sensitive data handling (passwords, tokens, PII)
  • [ ] Dependency vulnerabilities

Bugs & Logic (Critical)

  • [ ] Null/undefined handling
  • [ ] Edge cases (empty arrays, null values, boundaries)
  • [ ] Error handling in place
  • [ ] Race conditions or concurrency issues
  • [ ] State management issues

Performance (Important)

  • [ ] Algorithm complexity (watch for O(n²) where O(n) exists)
  • [ ] N+1 query problems
  • [ ] Memory leaks (listeners, subscriptions, closures)
  • [ ] Blocking operations that should be async

Testing (Important)

  • [ ] Changes covered by tests
  • [ ] Tests verify actual behavior
  • [ ] Edge cases tested
  • [ ] Error conditions tested

Code Quality

  • [ ] Code is understandable
  • [ ] No unnecessary complexity or clever code
  • [ ] Duplication worth extracting
  • [ ] Names match what they do

Architecture

  • [ ] Fits existing patterns (or has good reason not to)
  • [ ] No breaking changes without migration
  • [ ] Avoids unnecessary coupling

Output Format

Structure your review like this (see references/review-template.md for detailed examples):

  • Summary: One line verdict (Good to merge / Has issues / Needs work)
  • Critical: Security, data loss, crashes - must fix before merge
  • Important: Bugs, performance, missing tests - should fix
  • Minor: Quality improvements - nice to have
  • Questions: Things to clarify with the author
  • Prevent This: Suggest tooling/config to catch these issues automatically in the future
  • Positive Notes: Briefly acknowledge what's done well

Guidelines

  • Be specific: file:line, what's wrong, why it matters, how to fix
  • Skip style issues that linters catch
  • Explain impact, not just "this is wrong"
  • Consider trade-offs - sometimes simple is better than perfect
  • Briefly note if something is done well, but keep it short

Suggesting Future Mitigations

Only suggest mitigations for recurring patterns or critical issues. Don't suggest tools for one-off mistakes. Focus on automatable checks, not process changes.

TypeScript configuration:

  • Type safety issues (any, implicit types) → Suggest strict: true, noImplicitAny, strictNullChecks in tsconfig.json
  • Missing null checks → Suggest strictNullChecks: true

Linting rules:

  • Code quality patterns ESLint could catch → Suggest specific ESLint rules
  • Framework-specific issues → Suggest framework ESLint plugins (react-hooks, vue, etc.)
  • Formatting inconsistencies → Suggest Prettier in pre-commit hook

Pre-commit hooks:

  • Secrets/credentials committed → Suggest trufflehog or git-secrets
  • Test failures → Suggest running tests before commit
  • Type errors → Suggest tsc --noEmit check

CI/CD checks:

  • Security vulnerabilities → Suggest npm audit / dependency scanning
  • Missing tests → Suggest coverage thresholds
  • Build errors → Ensure build runs in CI

References

Need more guidance? Check these out:

  • Review Template - What your review output should look like, with severity categories and example issues
  • Common Issues - Quick reference of problems that come up often in reviews, with good/bad code examples

适合场景

01

用户想查找某类 Agent Skill 时

02

需要根据任务场景推荐可安装能力包时

03

需要对比不同来源的安装命令和来源信息时

04

需要参考平台分布和安装热度时

能力概览

能力 1

按任务关键词查找相关 Skills

能力 2

展示可复制的安装命令

能力 3

保留来源站点、仓库和原始说明,方便继续核验

能力 4

补充不同宿主或平台的使用分布数据

安装后应在对应宿主中按原始 README 的触发条件使用;具体调用方式请以来源页面和 README 为准。

平台分布

Claude Code

27.19%
按下载量换算39

windsurf

21.81%
按下载量换算31

trae

15.95%
按下载量换算23

OpenCode

11.41%
按下载量换算16

weavefox

8.55%
按下载量换算12

Codex

3.35%
按下载量换算5

安全审计

暂无安全审计结果可展示。

权限和风险

只读

该 Skill 主要提供规则、说明或参考内容,本身偏只读;真正读写文件、联网或执行命令仍取决于宿主 Agent 的任务。

安装前确认

本站仅展示第三方公开信息,不托管安装包,不提供自动安装或运行环境。安装前应自行审查源码、依赖和命令行为。当前只有一个来源,正式发布前建议补源仓库或其他目录站核验。

来源信息

继续浏览同类 Skills