跳到主内容
智客 ZICQ

技能库 智客分类:Agent 工作流 react

React

反应组件模式、钩子最佳做法、国家管理和业绩优化

12 安装量

官方网址:skills.sh

技能介绍

先看中文介绍;官方 description 原文单独保留,不改写 SKILL.md。

做什么

反应组件模式、钩子最佳做法、国家管理和业绩优化

何时用

官方 description 未单独写出 Use when。按规范,代理会在用户任务与这段 description 的关键词匹配时激活本技能。

代理如何加载

按 Agent Skills 渐进披露:启动时只加载 name 与 description(约 100 token);任务匹配后才读入整份 SKILL.md 正文;scripts/、references/、assets/ 仅在需要时再读。 本文件正文结构:React Code Review Rules、Security (Critical)、Hooks Rules (Essential)、State Management (Essential)、Performance (Essential)、Component Design (Essential)。

文件分析

文件分析:这是一份仅含 SKILL.md 的指令型技能,代理激活后整份正文进入上下文。

官方 description(原文)

React component patterns, hooks best practices, state management, and performance optimization

React Code Review RulesSecurity (Critical)Hooks Rules (Essential)State Management (Essential)Performance (Essential)Component Design (Essential)AccessibilityAdvanced PatternsAnti-patternsTestingError HandlingType Safety

来源分类:skills.sh agent-skill skills.sh agent-skill

SKILL.md 与 Agent 调用

官方规范 ↗
name
react
description
React component patterns, hooks best practices, state management, and performance optimization
  1. 发现技能客户端向 Agent 提供名称与描述目录。
  2. 匹配与调用用户指定或任务匹配后,载入 SKILL.md 指令。
  3. 按需加载按步骤读取参考文档、使用脚本与素材。

具体调用语法与可用工具以目标 Agent 客户端为准。 查看调用机制说明 ↗

安装这个技能

Skills CLI ↗

先选择目标 Agent 和安装范围,保留技能包的附属文件,安装后检查客户端能否发现该技能。

交给 Agent 安装

复制安装指令给支持 Agent Skills 的代理,确认其中的目标目录与客户端匹配。

把 Agent Skill「react」安装到我的项目:SKILL.md 原文与官方 description 见 https://zicq.com/zh/skills/skl-86bdd7faec22c794-%E5%8F%8D%E5%BA%94.html
请存为 .cursor/skills/react/SKILL.md 或 .claude/skills/react/SKILL.md,frontmatter 的 name 与 description 保持原样,不要改写。

GitHub 完整包 ↗

终端安装 · Skills CLI

需要 Node.js 与 npx。先查看仓库技能列表,确认实际名称。

npx skills add 'https://github.com/yldgio/codereview-skills' --list

npx skills add 'https://github.com/yldgio/codereview-skills' --skill 'react'

CLI 会交互选择目标 Agent,默认安装到项目;用户级安装使用 -g。先通过查看命令核对仓库内容,再用 npx skills list 检查已安装技能。

阅读排版
--- name: react description: React component patterns, hooks best practices, state management, and performance optimization metadata: version: "1.1.0" --- ## React Code Review Rules ### Security (Critical) - **Input Sanitization**: All user-supplied content must be escaped and validated. Always escape variables in output, especially when inserted into code or comments - **Template Variable Safety**: Never use template variables (`{{ }}`) in code or documentation. All variables must be carefully declared, validated, and typed. Disallow undeclared or untyped variables - **Comment Hygiene**: Never use HTML comments (``) to store instructions or data. Scan for all HTML comment patterns and require output auto-escaping when handling template variables - **Dynamic Output**: Validate and sanitize all user-supplied or dynamic variables before insertion into code, templates, or client output. Use strict allow-lists and escaping functions - Never render user input directly without sanitization (XSS prevention) - Use `dangerouslySetInnerHTML` only when absolutely necessary and with sanitized content - Validate and sanitize all user-provided content before rendering - Never interpolate untrusted user input into component code or instructions - Escape user input when rendering dynamic content ### Hooks Rules (Essential) - Hooks must be called at top level (not inside conditions, loops, or nested functions) - Custom hooks must start with `use` prefix - `useEffect` must have correct dependency array (no missing/extra deps) - `useEffect` cleanup functions must be returned for subscriptions/timers ### State Management (Essential) - State should be as local as possible (don't lift prematurely) - Avoid redundant state (derive values instead of storing) - Use `useReducer` for complex state logic with multiple sub-values - Prefer controlled components over uncontrolled (except file inputs) ### Performance (Essential) - Wrap expensive computations in `useMemo` - Stabilize callbacks with `useCallback` when passed to memoized children - Use `React.memo()` for components that render often with same props - Avoid creating objects/arrays inline in JSX (causes re-renders) ### Component Design (Essential) - Single responsibility: one component, one purpose - Props should be minimal and well-typed - Avoid prop drilling > 2 levels (use Context or composition) - Prefer composition over prop-based conditional rendering ### Accessibility - Interactive elements must be keyboard accessible - Use semantic HTML (`button` not `div onClick`) - Images need `alt` text - Form inputs need associated labels ### Advanced Patterns - Use compound components (parent + stateless children) to model complex, related UI pieces while keeping state in the parent - Prefer controlled abstractions (e.g., `value`/`onChange` pairs) so consumers can own state when needed - Use Context for cross-cutting concerns (theme, auth, feature flags) and co-locate provider logic with domain-specific hooks - Consider render props or headless components when you need to share complex interaction logic while letting consumers control markup and styling - Combine memoization (`React.memo`, `useMemo`, `useCallback`) with clear prop contracts to optimize expensive, frequently rendered components - For deeper exploration of advanced React patterns, see the official [React documentation](https://react.dev) ### Anti-patterns - Avoid `useEffect` for state derivation (compute during render instead) - Avoid `useEffect` on mount for data that could be fetched server-side - Avoid index as key in lists that reorder ### Testing - Write unit tests for components using React Testing Library - Test custom hooks with `@testing-library/react-hooks` - Test user interactions and state changes - Mock external dependencies appropriately - Test error states and edge cases - Ensure tests are maintainable and readable ### Error Handling - Implement error boundaries for components - Provide fallback UIs for errors - Handle async errors appropriately - Log errors for debugging and monitoring - Don't let errors crash the entire app ### Type Safety - Use TypeScript or PropTypes for type checking - Define prop types for all components - Use TypeScript interfaces for complex props - Validate prop types at runtime with PropTypes - Catch type errors early in development ### Styling - Use CSS Modules for scoped styles - Consider styled-components or CSS-in-JS for dynamic styles - Avoid inline styles for static CSS - Follow a consistent styling approach - Prevent global CSS conflicts with scoping ### Dependency Management - Review third-party dependencies before adding - Keep dependencies up to date - Audit packages for security vulnerabilities - Avoid over-reliance on libraries for simple tasks - Consider bundle size impact of dependencies ### Documentation - Add doc comments for complex components and hooks - Write self-explanatory code over excessive comments - Document non-obvious behavior and edge cases - Include usage examples in component docs - Keep comments up to date with code changes

相关技能

Agent 工作流

Skill Creator

创造有效技能指南。 当用户想创造出新的技能(或更新现有的技能),以专业知识,工作流程,或工具集成来扩展克洛德的能力时,应该使用这种技能.

Agent 工作流

Clawdhub

使用ClawdHub CLI搜索,安装,更新并发布从taladhub.com的代理技能. 需要获取苍蝇上的新技能时使用,将安装的技能同步到最新版本或特定版本,或者发布 npm-instainddhub CLI 的新/更新的技能文件夹.

Agent 工作流

Agent Team Orchestration

管弦乐团多代理团队,任务设定周期,交接协议,审查工作流程. 使用时间: (1)建立2+特派员队伍,具有不同专业,(2)确定任务路线和生命周期(收录框_ spec_建设_审查_完成),(3)在特派员之间制定交接协议,(4)建立审查和质量关口,(5)管理特派员之间的交流和文物共享.

Agent 工作流

Superpowers

Spec-first,TDD,子代理驱动的软件开发工作流程. 当:(1)构建任何新功能或应用——触发脑暴_计划_子代理执行回路,(2)调试出一个bug或测试失败——触发系统性的根起过程,(3)用户说"让我们构建","帮助我计划","我想添加X",或"这个被打破",(4)完成一个功…