Token导航 LogoToken导航TokenDH.com
研究检索需要联网github未标认证来源可访问clear审计未展示

review审查

Agent Skill

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

517

周安装

22

GitHub Stars

公开资料未说明

下载量

181
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:review(审查)
来源仓库:https://github.com/duong/dotfiles
仓库路径:skills/review
安装命令:
npx skills add duong/dotfiles --skill "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 duong/dotfiles --skill "review"

简介

用于查找和筛选相关信息。适用宿主包括 Codex、Claude、Cursor、Gemini CLI,接入前应确认版本、权限和运行环境要求。

  • 适合基于关键词或场景定位内容。
  • 通过 npx 从 GitHub 安装,支持多个 AI 宿主。
  • 安装前建议确认权限及是否触发联网或文件操作。
  • review 属于研究检索类 Skill,可作为该场景下的辅助能力补充。

SKILL.md

name
review
description
Conduct thorough code reviews focusing on code quality, security, best practices, and project standards. Use when asked to review code or PRs.

🔬 Code Review

Act as a top-tier principal software engineer to conduct a thorough code review focusing on code quality, best practices, and adherence to requirements, plan, and project standards.

When to Use

  • When asked to review code changes
  • When asked to review a PR or diff
  • When asked to audit code quality

Constraints

  • Don't make changes - Review-only. Output serves as input for planning.
  • Avoid unfounded assumptions - If unsure, note and ask in the review response.
  • Do ONE THING at a time - Get user approval before moving on.
  • Before using APIs - Look at documentation in installed module README or use web search if necessary.

Thinking Process

For each step, show your work:

🎯 restate |> 💡 ideate |> 🪞 reflectCritically |> 🔭 expandOrthogonally |> ⚖️ scoreRankEvaluate |> 💬 respond

Review Criteria

  • JavaScript/TypeScript - Code quality and best practices
  • TDD - Test coverage and test quality assessment
  • Architecture - NextJS + React/Redux + Shadcn UI patterns
  • UI/UX - Design and component quality
  • Redux - State management patterns and Autodux usage
  • Network Effects - Side effect handling
  • Commit Messages - Quality and conventional commit format
  • Security - OWASP top 10 violations, timing-safe comparisons for secrets/tokens
  • Documentation - Code comments comply with style guides, minimal docblocks for public APIs
  • Dead Code - No unused stray files or dead code

Review Process

  1. Discover & Read Project Rules - Run discovery commands, then read the contents of each file found:

- AGENTS.md / AGENT.md files in repository root and subdirectories - .cursor/rules/*.mdc files for project-specific conventions - .cursorrules file (legacy format) in repository root - Key rules to look for (read if present): - *-best-practices.mdc - Testing and coding patterns - *-conventions.mdc - Code conventions - *-i18n*.mdc - Internationalization rules - *-ui-components.mdc - UI component standards - *pr-review*.mdc - PR review guidelines

  1. Analyze code structure and organization
  2. Check adherence to project-specific rules discovered in step 1
  3. Evaluate test coverage and quality
  4. Assess performance considerations
  5. Deep scan for security vulnerabilities, visible keys, etc.
  6. Review UI/UX implementation and accessibility
  7. Validate architectural patterns and design decisions
  8. Check documentation and commit message quality
  9. Compare to requirements - Verify completed work meets functional requirements
  10. Compare to task plan - Check tasks/ directory for planned work and ensure adherence
  11. Provide actionable feedback citing specific rule violations

Auto-Discovery Commands

Run these from the repository to find project rules, then read each discovered file:

# Get repo root
REPO_ROOT=$(git rev-parse --show-toplevel 2>/dev/null || pwd)

# Find AGENTS.md files
find "$REPO_ROOT" -name "AGENTS.md" -o -name "AGENT.md" 2>/dev/null

# Find cursor rules (.mdc and legacy .cursorrules)
find "$REPO_ROOT" -path "*/.cursor/rules/*.mdc" 2>/dev/null
find "$REPO_ROOT" -name ".cursorrules" 2>/dev/null

Security Checklist

Explicitly review for each OWASP Top 10:

  1. Broken Access Control - Unauthorized access to resources
  2. Cryptographic Failures - Weak crypto, exposed secrets
  3. Injection - SQL, NoSQL, OS, LDAP injection
  4. Insecure Design - Missing security controls
  5. Security Misconfiguration - Default configs, verbose errors
  6. Vulnerable Components - Outdated dependencies
  7. Auth Failures - Weak auth, session management
  8. Data Integrity Failures - Insecure deserialization
  9. Logging Failures - Missing audit trails
  10. SSRF - Server-side request forgery

CRITICAL: Flag crypto.timingSafeEqual or raw token comparisons for CSRF, API keys, sessions.

Philosophy

Simplicity is removing the obvious and adding the meaningful. Perfection is attained not when there is nothing more to add, but when there is nothing more to remove.

Dig deep. Look for: redundancies, forgotten files (d.ts, etc), things that should have been moved or deleted that were not.

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

29.35%
按下载量换算53

windsurf

20.64%
按下载量换算37

OpenCode

18.89%
按下载量换算34

Cursor

11.59%
按下载量换算21

Codex

6.8%
按下载量换算12

Antigravity

3.31%
按下载量换算6

安全审计

暂无安全审计结果可展示。

权限和风险

需要联网

该 Skill 可能需要联网访问来源站点、仓库或外部 API;具体网络访问范围需要结合源码和 README 复核。

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