using-toolkit

Compare original and translation side by side

🇺🇸

Original

English

🇨🇳

Translation

Chinese

<EXTREMELY-IMPORTANT> If you think there is even a 1% chance a skill might apply to what you are doing, you ABSOLUTELY MUST invoke the skill.

IF A SKILL APPLIES TO YOUR TASK, YOU DO NOT HAVE A CHOICE. YOU MUST USE IT.

This is not negotiable. This is not optional. You cannot rationalize your way out of this. </EXTREMELY-IMPORTANT>

<EXTREMELY-IMPORTANT> 如果你认为哪怕只有1%的概率某项技能可能适用于你当前正在做的事，你绝对必须调用该技能。

如果某项技能适用于你的任务，你没有选择余地，必须使用它。

这没有协商空间，不是可选项，你不能找任何理由回避这一要求。 </EXTREMELY-IMPORTANT>

Overview

概述

Master catalog and dispatch system for all 64 toolkit skills. Ensures proper skill identification at conversation start.

全部64项工具包技能的主目录与调度系统，确保对话启动时能正确识别适用技能。

How to Access Skills

如何访问技能

In Claude Code: Use the

Skill

tool. When you invoke a skill, its content is loaded and presented to you — follow it directly. Never use the Read tool on skill files.

在Claude Code中： 使用

Skill

工具。当你调用某项技能时，其内容会被加载并展示给你——直接遵循即可。永远不要使用Read工具读取技能文件。

Phase 1: Skill Identification

阶段1：技能识别

Before ANY action: scan catalog, identify ALL matching skills (even 1% relevance), invoke via Skill tool.

在采取任何行动前：扫描目录，识别所有匹配的技能（哪怕只有1%的相关性），通过Skill工具调用。

Phase 2: Skill Prioritization

阶段2：技能优先级排序

Invoke in order: (1) Process (brainstorming, planning) → (2) Specialist (frontend, backend, architect) → (3) Quality (TDD, review) → (4) Docs → (5) Design → (6) Ops → (7) Terminal (verification).

按以下顺序调用：(1) 流程类（头脑风暴、规划）→ (2) 专家类（前端、后端、架构师）→ (3) 质量类（TDD、评审）→ (4) 文档类 → (5) 设计类 → (6) 运维类 → (7) 终端类（验证）。

Phase 3: Skill Execution

阶段3：技能执行

Rigid skills: Follow exactly. Flexible skills: Adapt to context. Always invoke

verification-before-completion

before claiming done.

刚性技能： 严格遵循要求执行。弹性技能： 根据上下文适配调整。在宣称任务完成前，必须始终调用

verification-before-completion

。

Available Skills (64 Total)

可用技能（共64项）

Core Skills (6)

核心技能（6项）

Skill	When to Use
`using-toolkit`	Session start — establishes skill usage
`self-learning`	Starting work on unfamiliar projects, or when corrected
`resilient-execution`	When an approach fails — ensures retry with alternatives
`circuit-breaker`	Autonomous loops, repeated operations, stagnation detection
`auto-improvement`	Self-improving system, tracks effectiveness, learns from errors
`verification-before-completion`	Before claiming ANY task is complete

Skill	使用场景
`using-toolkit`	会话启动时使用——明确技能使用规则
`self-learning`	开始处理不熟悉的项目，或收到纠正反馈时
`resilient-execution`	方案执行失败时——确保尝试其他替代方案重试
`circuit-breaker`	自主循环、重复操作、停滞检测场景
`auto-improvement`	系统自优化，跟踪有效性，从错误中学习
`verification-before-completion`	在宣称任何任务完成前使用

Process & Workflow Skills (9)

流程与工作流技能（9项）

Skill	When to Use
`planning`	Before ANY implementation — forces structured planning
`brainstorming`	Before creative work — exploring ideas, features, designs
`task-management`	Breaking work into tracked steps during implementation
`executing-plans`	Executing approved plan documents step by step
`subagent-driven-development`	Multi-task execution with two-stage review gates
`dispatching-parallel-agents`	Running multiple independent tasks concurrently
`autonomous-loop`	Ralph-style iterative autonomous development loops
`ralph-status`	End of every autonomous loop iteration — structured progress reporting
`task-decomposition`	Hierarchical task breakdown, dependency mapping, parallelization

Skill	使用场景
`planning`	任何实现工作开展前——强制进行结构化规划
`brainstorming`	创造性工作开展前——探索想法、功能、设计
`task-management`	实现过程中将工作拆分为可跟踪的步骤
`executing-plans`	按步骤执行已获批的规划文档
`subagent-driven-development`	带两阶段评审关口的多任务执行
`dispatching-parallel-agents`	并发运行多个独立任务
`autonomous-loop`	Ralph风格的迭代自主开发循环
`ralph-status`	每次自主循环迭代结束时——结构化进度汇报
`task-decomposition`	分层任务拆解、依赖映射、并行化处理

Quality Assurance Skills (17)

质量保障技能（17项）

Skill	When to Use
`code-review`	After completing tasks, before committing
`test-driven-development`	Writing any new code (RED-GREEN-REFACTOR)
`systematic-debugging`	Investigating bugs, errors, unexpected behavior
`testing-strategy`	Choosing testing approach for a project
`security-review`	Reviewing for vulnerabilities, auth, input validation
`performance-optimization`	Optimizing speed, reducing load times
`acceptance-testing`	Validating implementation meets spec acceptance criteria
`llm-as-judge`	Evaluating subjective quality (tone, UX, readability, aesthetics)
`senior-frontend`	React/Next.js/TypeScript specialist with >85% test coverage
`senior-backend`	API design, microservices, event-driven architecture
`senior-architect`	System design, scalability, trade-off analysis, ADRs
`senior-fullstack`	End-to-end development across the full stack
`clean-code`	SOLID, DRY, code smells, refactoring patterns
`react-best-practices`	React hooks, context, suspense, server components
`webapp-testing`	Playwright-based web testing, screenshots, browser logs
`senior-prompt-engineer`	Prompt design, optimization, chain-of-thought
`senior-data-scientist`	ML pipelines, statistical analysis, experiment design

Skill	使用场景
`code-review`	完成任务后，提交代码前
`test-driven-development`	编写任何新代码时（红-绿-重构流程）
`systematic-debugging`	排查Bug、错误、异常行为时
`testing-strategy`	为项目选择测试方案时
`security-review`	排查漏洞、鉴权、输入验证相关问题时
`performance-optimization`	优化速度、减少加载时间时
`acceptance-testing`	验证实现是否符合规格说明的验收标准时
`llm-as-judge`	评估主观质量（语气、UX、可读性、美观度）时
`senior-frontend`	要求测试覆盖率>85%的React/Next.js/TypeScript专业开发
`senior-backend`	API设计、微服务、事件驱动架构相关工作
`senior-architect`	系统设计、可扩展性、权衡分析、架构决策记录相关工作
`senior-fullstack`	全栈端到端开发
`clean-code`	SOLID、DRY、代码异味、重构模式相关工作
`react-best-practices`	React hooks、context、suspense、服务端组件相关工作
`webapp-testing`	基于Playwright的Web测试、截图、浏览器日志相关工作
`senior-prompt-engineer`	提示词设计、优化、思维链相关工作
`senior-data-scientist`	ML pipeline、统计分析、实验设计相关工作

Documentation Skills (5)

文档技能（5项）

Skill	When to Use
`prd-generation`	Generating Product Requirements Documents
`tech-docs-generator`	Generating or updating technical documentation
`writing-skills`	Creating new skills, commands, or agent definitions
`spec-writing`	Writing specifications with JTBD methodology and acceptance criteria
`reverse-engineering-specs`	Generating implementation-free specs from existing codebases

Skill	使用场景
`prd-generation`	生成产品需求文档时
`tech-docs-generator`	生成或更新技术文档时
`writing-skills`	创建新技能、命令或Agent定义时
`spec-writing`	采用JTBD方法编写带验收标准的规格说明时
`reverse-engineering-specs`	从现有代码库生成无实现绑定的规格说明时

Design Skills (3)

设计技能（3项）

Skill	When to Use
`api-design`	Designing API endpoints and generating specs
`frontend-ui-design`	Component architecture, responsive design, accessibility
`database-schema-design`	Data modeling, migrations, indexing

Skill	使用场景
`api-design`	设计API端点并生成规格说明时
`frontend-ui-design`	组件架构、响应式设计、可访问性相关工作
`database-schema-design`	数据建模、迁移、索引相关工作

Operations Skills (7)

运维技能（7项）

Skill	When to Use
`deployment`	Setting up CI/CD pipelines and deploy checklists
`using-git-worktrees`	Creating isolated development environments
`finishing-a-development-branch`	Completing work on a branch, preparing to merge
`git-commit-helper`	Conventional commits, semantic versioning, changelogs
`senior-devops`	CI/CD, Docker, Kubernetes, infrastructure-as-code
`mcp-builder`	MCP server development, tools, resources, transport layers
`agent-development`	Building AI agents, tool use, memory, planning

Skill	使用场景
`deployment`	搭建CI/CD pipeline和部署检查清单时
`using-git-worktrees`	创建隔离开发环境时
`finishing-a-development-branch`	完成分支上的工作，准备合并时
`git-commit-helper`	约定式提交、语义化版本、更新变更日志时
`senior-devops`	CI/CD、Docker、Kubernetes、基础设施即代码相关工作
`mcp-builder`	MCP服务端开发、工具、资源、传输层相关工作
`agent-development`	构建AI Agent、工具使用、记忆、规划相关工作

Creative Skills (6)

创意技能（6项）

Skill	When to Use
`ui-ux-pro-max`	Full UI/UX design intelligence with styles, palettes, fonts, UX guidelines
`ui-design-system`	Design tokens, component libraries, Tailwind CSS, responsive patterns
`canvas-design`	HTML Canvas, SVG, data visualization, generative art
`mobile-design`	React Native, Flutter, SwiftUI, platform HIG compliance
`ux-researcher-designer`	User research, personas, journey maps, usability testing
`artifacts-builder`	Generate standalone artifacts, interactive demos, prototypes

Skill	使用场景
`ui-ux-pro-max`	全栈UI/UX设计，包含样式、配色、字体、UX规范相关工作
`ui-design-system`	设计token、组件库、Tailwind CSS、响应式模式相关工作
`canvas-design`	HTML Canvas、SVG、数据可视化、生成艺术相关工作
`mobile-design`	React Native、Flutter、SwiftUI、符合平台HIG规范的设计工作
`ux-researcher-designer`	用户研究、用户画像、旅程地图、可用性测试相关工作
`artifacts-builder`	生成独立产物、交互演示、原型时

Business Skills (3)

商业技能（3项）

Skill	When to Use
`seo-optimizer`	Technical SEO, meta tags, structured data, Core Web Vitals
`content-research-writer`	Research methodology, long-form content, citations
`content-creator`	Marketing copy, social media, brand voice

Skill	使用场景
`seo-optimizer`	技术SEO、meta标签、结构化数据、核心网页指标相关工作
`content-research-writer`	研究方法论、长文内容、引用标注相关工作
`content-creator`	营销文案、社交媒体内容、品牌调性相关工作

Document Processing Skills (3)

文档处理技能（3项）

Skill	When to Use
`docx-processing`	Word document generation, template filling
`pdf-processing`	PDF generation, form filling, OCR, merge/split
`xlsx-processing`	Excel manipulation, formulas, charts

Skill	使用场景
`docx-processing`	Word文档生成、模板填充时
`pdf-processing`	PDF生成、表单填充、OCR、合并/拆分时
`xlsx-processing`	Excel操作、公式、图表相关工作

Frameworks & Languages Skills (3)

框架与语言技能（3项）

Skill	When to Use
`laravel-specialist`	Laravel development — Eloquent, Blade, Livewire, queues, Pest
`php-specialist`	Modern PHP 8.x — PSR standards, static analysis, Composer
`laravel-boost`	Laravel Boost performance optimization — caching, queries, N+1

Skill	使用场景
`laravel-specialist`	Laravel开发——Eloquent、Blade、Livewire、队列、Pest相关工作
`php-specialist`	现代PHP 8.x开发——PSR规范、静态分析、Composer相关工作
`laravel-boost`	Laravel Boost性能优化——缓存、查询、N+1问题相关工作

Productivity Skills (1)

生产力技能（1项）

Skill	When to Use
`file-organizer`	Project structure, file naming, directory architecture

Skill	使用场景
`file-organizer`	项目结构、文件命名、目录架构相关工作

Communication Skills (1)

沟通技能（1项）

Skill	When to Use
`email-composer`	Professional email drafting, tone adjustment

Skill	使用场景
`email-composer`	专业邮件起草、语气调整时

Decision Table: Choosing the Right Skill

决策表：选择合适的技能

User Request Contains	Primary Skill	Supporting Skills
"build", "implement", "create feature"	`planning`	`brainstorming` , `tdd` , `code-review`
"fix", "bug", "error", "broken"	`systematic-debugging`	`tdd` , `resilient-execution`
"test", "coverage", "spec"	`test-driven-development`	`testing-strategy` , `acceptance-testing`
"review", "check", "audit"	`code-review`	`security-review` , `clean-code`
"plan", "how should we"	`planning`	`brainstorming` , `task-decomposition`
"deploy", "CI/CD", "pipeline"	`deployment`	`senior-devops`
"API", "endpoint", "REST", "GraphQL"	`api-design`	`senior-backend`
"React", "Next.js", "component"	`senior-frontend`	`react-best-practices` , `frontend-ui-design`
"Laravel", "Eloquent", "Blade", "Livewire"	`laravel-specialist`	`php-specialist` , `laravel-boost`
"PHP", "Composer", "PSR"	`php-specialist`	`laravel-specialist`
"database", "schema", "migration"	`database-schema-design`	`senior-backend`
"design", "UI", "UX"	`ui-ux-pro-max`	`ui-design-system` , `frontend-ui-design`
"mobile", "iOS", "Android"	`mobile-design`	`ui-ux-pro-max`
"document", "docs", "README"	`tech-docs-generator`	`prd-generation`
"spec", "requirements", "PRD"	`spec-writing`	`prd-generation`
"autonomous", "loop", "ralph"	`autonomous-loop`	`ralph-status` , `circuit-breaker`
"performance", "optimize", "slow"	`performance-optimization`	`laravel-boost`
"security", "vulnerability", "auth"	`security-review`	`senior-backend`
"SEO", "meta tags", "search engine"	`seo-optimizer`	`content-research-writer`
"email", "draft", "compose"	`email-composer`	`content-creator`
"PDF", "Word", "Excel"	`pdf-processing` / `docx-processing` / `xlsx-processing`	—
"agent", "AI", "tool use"	`agent-development`	`mcp-builder`
"MCP", "server", "transport"	`mcp-builder`	`agent-development`

用户请求包含	核心技能	辅助技能
"build", "implement", "create feature"	`planning`	`brainstorming` , `tdd` , `code-review`
"fix", "bug", "error", "broken"	`systematic-debugging`	`tdd` , `resilient-execution`
"test", "coverage", "spec"	`test-driven-development`	`testing-strategy` , `acceptance-testing`
"review", "check", "audit"	`code-review`	`security-review` , `clean-code`
"plan", "how should we"	`planning`	`brainstorming` , `task-decomposition`
"deploy", "CI/CD", "pipeline"	`deployment`	`senior-devops`
"API", "endpoint", "REST", "GraphQL"	`api-design`	`senior-backend`
"React", "Next.js", "component"	`senior-frontend`	`react-best-practices` , `frontend-ui-design`
"Laravel", "Eloquent", "Blade", "Livewire"	`laravel-specialist`	`php-specialist` , `laravel-boost`
"PHP", "Composer", "PSR"	`php-specialist`	`laravel-specialist`
"database", "schema", "migration"	`database-schema-design`	`senior-backend`
"design", "UI", "UX"	`ui-ux-pro-max`	`ui-design-system` , `frontend-ui-design`
"mobile", "iOS", "Android"	`mobile-design`	`ui-ux-pro-max`
"document", "docs", "README"	`tech-docs-generator`	`prd-generation`
"spec", "requirements", "PRD"	`spec-writing`	`prd-generation`
"autonomous", "loop", "ralph"	`autonomous-loop`	`ralph-status` , `circuit-breaker`
"performance", "optimize", "slow"	`performance-optimization`	`laravel-boost`
"security", "vulnerability", "auth"	`security-review`	`senior-backend`
"SEO", "meta tags", "search engine"	`seo-optimizer`	`content-research-writer`
"email", "draft", "compose"	`email-composer`	`content-creator`
"PDF", "Word", "Excel"	`pdf-processing` / `docx-processing` / `xlsx-processing`	—
"agent", "AI", "tool use"	`agent-development`	`mcp-builder`
"MCP", "server", "transport"	`mcp-builder`	`agent-development`

Workflow Patterns

工作流模式

Pattern	Skill Chain
"Build feature X"	brainstorming -> planning -> executing-plans -> code-review -> verification
"Fix bug Y"	systematic-debugging -> TDD -> code-review -> verification
"Write new code"	test-driven-development (always)
"Run autonomously"	autonomous-loop -> ralph-status -> circuit-breaker
"Write specs"	spec-writing (JTBD methodology)
"Understand legacy code"	reverse-engineering-specs -> spec-writing (audit)
"Check acceptance criteria"	acceptance-testing (backpressure chain)
"Validate subjective quality"	llm-as-judge (rubric-based evaluation)
"Document the API"	tech-docs-generator or api-design
"Create a PRD for Z"	prd-generation
"Set up CI/CD"	deployment
"How should we test?"	testing-strategy
"Design the database"	database-schema-design
"Build a UI component"	frontend-ui-design
"Check for security issues"	security-review
"Make it faster"	performance-optimization
"Done with this branch"	finishing-a-development-branch
"Design a UI"	ui-ux-pro-max -> ui-design-system -> frontend-ui-design
"Build mobile app"	mobile-design -> planning -> tdd
"Optimize SEO"	seo-optimizer
"Write marketing copy"	content-creator
"Process documents"	docx-processing / pdf-processing / xlsx-processing
"Compose email"	email-composer
"Build an AI agent"	agent-development -> planning -> tdd
"Set up infrastructure"	senior-devops -> deployment
"Decompose complex task"	task-decomposition -> dispatching-parallel-agents
"Build Laravel feature"	laravel-specialist -> planning -> tdd -> code-review
"Optimize Laravel app"	laravel-boost -> performance-optimization -> verification
"Write PHP code"	php-specialist -> tdd -> clean-code

模式	技能链
"开发功能X"	brainstorming -> planning -> executing-plans -> code-review -> verification
"修复Bug Y"	systematic-debugging -> TDD -> code-review -> verification
"编写新代码"	test-driven-development (始终执行)
"自主运行"	autonomous-loop -> ralph-status -> circuit-breaker
"编写规格说明"	spec-writing (JTBD方法论)
"理解遗留代码"	reverse-engineering-specs -> spec-writing (审计)
"检查验收标准"	acceptance-testing (背压链)
"验证主观质量"	llm-as-judge (基于评分规则的评估)
"为API编写文档"	tech-docs-generator 或 api-design
"为Z生成PRD"	prd-generation
"搭建CI/CD"	deployment
"我们应该怎么测试？"	testing-strategy
"设计数据库"	database-schema-design
"开发UI组件"	frontend-ui-design
"检查安全问题"	security-review
"优化速度"	performance-optimization
"完成该分支开发"	finishing-a-development-branch
"设计UI"	ui-ux-pro-max -> ui-design-system -> frontend-ui-design
"开发移动应用"	mobile-design -> planning -> tdd
"优化SEO"	seo-optimizer
"编写营销文案"	content-creator
"处理文档"	docx-processing / pdf-processing / xlsx-processing
"撰写邮件"	email-composer
"构建AI Agent"	agent-development -> planning -> tdd
"搭建基础设施"	senior-devops -> deployment
"拆解复杂任务"	task-decomposition -> dispatching-parallel-agents
"开发Laravel功能"	laravel-specialist -> planning -> tdd -> code-review
"优化Laravel应用"	laravel-boost -> performance-optimization -> verification
"编写PHP代码"	php-specialist -> tdd -> clean-code

Anti-Patterns

反模式

Never skip skill check for "simple" tasks — always check catalog
Never invoke skills from memory — use Skill tool for current version
Never read skill files with Read tool — use Skill tool
Never skip verification — always invoke
```
verification-before-completion
```
last
Never respond before checking skills — skill check is ALWAYS first

永远不要因为任务“简单”就跳过技能检查——始终扫描目录
永远不要凭记忆调用技能——使用Skill工具获取当前版本
永远不要使用Read工具读取技能文件——使用Skill工具
永远不要跳过验证——最后必须调用
```
verification-before-completion
```
永远不要在检查技能前给出回复——技能检查永远是第一步

Anti-Rationalization Guards

反合理化防护

If you think "This is just simple / I need context first / Let me explore first / This doesn't need a skill / I remember this skill / The skill is overkill / I'll do one thing first" — STOP. Check for skills FIRST. No exceptions.

如果你产生了“这只是小事/我先需要上下文/我先探索下/这不需要技能/我记得这个技能/这个技能太小题大做了/我先做一件事再说”这类想法——立刻停止。先检查是否有适用技能。没有例外。

Integration Points

集成点

Key integrations:

self-learning

(project context),

verification-before-completion

(terminal checkpoint),

circuit-breaker

(stall safety),

auto-improvement

(effectiveness tracking),

planning

(most common first skill),

resilient-execution

(failure recovery).

关键集成项：

self-learning

（项目上下文）、

verification-before-completion

（终端检查点）、

circuit-breaker

（停滞安全防护）、

auto-improvement

（有效性跟踪）、

planning

（最常用的首项技能）、

resilient-execution

（故障恢复）。

Core Rules

核心规则

Plan before coding. TDD always. Verify completion with evidence. Review before merge. Use subagents for parallel work. Self-learn continuously. Try 3 approaches before escalating. Report RALPH_STATUS in loops. Protect config files. Write specs with acceptance criteria.

编码前先规划。始终遵循TDD。用证据验证完成情况。合并前先评审。使用子代理处理并行工作。持续自主学习。上报前尝试3种解决方案。循环中上报RALPH_STATUS。保护配置文件。编写带验收标准的规格说明。

Skill Types

技能类型

Rigid (TDD, debugging, planning, verification, code-review, autonomous-loop, circuit-breaker, spec-writing, acceptance-testing): Follow exactly. Flexible (brainstorming, tech-docs, api-design, frontend, database, performance, security-review, prd-generation, laravel-specialist, php-specialist): Adapt to context.

刚性（TDD、调试、规划、验证、代码评审、自主循环、熔断、规格编写、验收测试）：严格遵循要求执行。弹性（头脑风暴、技术文档、API设计、前端、数据库、性能、安全评审、PRD生成、laravel-specialist、php-specialist）：根据上下文适配调整。

Find Missing Skills

查找缺失技能

npx skills find [query]

to search,

npx skills add <owner/repo@skill> -g -y

to install. Prefer 1K+ weekly installs from reputable sources.

使用

npx skills find [query]

进行搜索，使用

npx skills add <owner/repo@skill> -g -y

进行安装。优先选择来自可靠来源、周安装量1K+的技能。

Skill Type

技能属性

RIGID — Skill checking is mandatory. Every task begins with a catalog scan. No exceptions. No rationalization.

RIGID —— 技能检查是强制性的。所有任务都从目录扫描开始。没有例外。不接受任何辩解。