Claude Skills
The open format is called Agent Skills and works in Claude Code, Codex, Cursor and other agents — most people know it as Claude Skills.
Every Agent Skill we could find on GitHub, deduplicated by content. 79 354 files from 1 739 authors, of which 61 713 are unique — the rest is the same skill repackaged into someone else's repository. For each one: what it weighs in tokens, whether it ships runnable scripts, and which MCP servers it needs.
Design APIs using schema-first approach with OpenAPI/Swagger. Use when creating new APIs, documenting existing ones, or when frontend/backend teams need to work in parallel. Covers OpenAPI spec, validation, and code generation.
Generate user-friendly release notes from git commits. Use this when preparing releases, writing changelogs, or summarizing development progress.
Set up CI/CD pipelines with GitHub Actions. Use when creating new projects, adding automation, or when manual verification becomes bottleneck. Covers lint, test, build, deploy automation.
Complete code review workflow for both requesting and receiving reviews. Use when creating PRs, reviewing code, or responding to feedback.
Create interactive data visualizations using D3.js. Use this when creating charts, graphs, network diagrams, geographic visualizations, or custom SVG-based data visualization.
Manage database schema changes with version control. Use when modifying DB schema, adding tables/columns, or setting up new projects. Covers Prisma, Drizzle, and migration best practices.
Enforce dependency security scanning and SBOM generation. Use when adding dependencies, reviewing package.json, or during security audits. Covers OWASP dependency check, npm audit, and supply chain security.
Use when facing 3+ independent failures that can be investigated without shared state or dependencies. Dispatches multiple Claude agents to investigate and fix independent problems concurrently.
Enforce proper error handling patterns. Use when writing async code, API calls, or user-facing features. Covers try-catch, error boundaries, graceful degradation, and user feedback.
Execute plan in batches with review checkpoints. Use when partner provides a complete implementation plan to execute with controlled batches and architect review between batches.
Complete Git workflow from conventional commits to pre-completion verification. Use for all Git operations including commits, branches, and releases.
Enforce file line count limits (200 recommended, 300 max) for CODE IMPLEMENTATION files only. Use this when reviewing code, creating files, or when files exceed line limits and need modularization.
Prevents 30+ critical AI/ML mistakes including data leakage, evaluation errors, training pitfalls, and deployment issues. Use when working with ML training, testing, model evaluation, or deployment.
Forbid hardcoded values in code. Use this when reviewing code, writing new features, or when magic numbers/strings are detected. Enforces constants, env variables, and config files.
Enforce Core Web Vitals optimization. Use when building user-facing features, reviewing performance, or when Lighthouse scores drop. Covers LCP, FID/INP, CLS, and optimization techniques.
Transform ideas into actionable implementation plans. Combines Socratic questioning for requirements discovery with detailed task breakdown for zero-context engineers. Use before any feature development.
Organize project folders following industry best practices. Use when setting up new projects, reorganizing codebases, or when folder structure becomes messy. Covers Next.js, Bulletproof React, and FSD patterns.
Advanced prompt engineering techniques for optimal AI responses. Use this when crafting prompts, optimizing AI interactions, or designing system prompts for applications.
Write developer-oriented README as onboarding documentation. Use when creating/updating README, setting up new projects, or when new developers need to understand the codebase quickly. README = Entry point + Architecture overview + Working agreement.
작업 완료 후 상세 리포트 문서를 작성. 변경 이력, 영향도 분석, 검증 결과를 문서화할 때 사용. 파일명 규칙 YYYY-MM-DD-<제목>-report.md
Automatically backs up files, saves diffs, uses agents/skills, and ensures modular code (<200 lines) before any implementation. Use this skill for ALL code changes to ensure safe, reversible, and clean implementations.
Guide for creating new Claude skills. Use this when creating a new skill, updating existing skills, or learning best practices for skill development.
Create, update, test, and manage Agent Skills. Use when creating new skills, debugging existing ones, or organizing skill libraries. Based on official Claude Code documentation.
Clean Architecture and SOLID principles guide. Use this when designing systems, reviewing architecture, or making structural decisions.
Implement JSON-based structured logging for observability. Use when setting up logging, debugging production issues, or preparing for log aggregation (ELK, Datadog). Covers log levels, context, and best practices.
Use when executing implementation plans with independent tasks in the current session. Dispatches fresh subagent for each task with code review between tasks.
Use when encountering any bug, test failure, or unexpected behavior, before proposing fixes. Four-phase framework ensuring understanding before attempting solutions.
Comprehensive testing workflow combining TDD, real implementations (no mocking), and E2E testing. Use when implementing features, writing tests, or setting up test infrastructure.
Verify 3-Mirror skill sync consistency across .public/skills, .codex/skills, and .claude/skills. Use after skill changes, before commits, or for CI validation.
Enforce TypeScript strict mode and type safety. Use when setting up projects, reviewing code, or when type errors are ignored. Covers strict flags, no-any rules, and type inference best practices.
Create isolated git worktrees with smart directory selection and safety verification. Use this when starting feature work that needs isolation, parallel development, or safe experimentation.
Check and create virtual environments for projects that need them. Use when starting Python/Node projects, or when dependency isolation is needed. Activates for Python, Node.js, and similar ecosystems.
Download and process YouTube video transcripts using yt-dlp. Use this when extracting subtitles, creating summaries from videos, or processing video content.
Documentation for the htmx JavaScript library (attributes, events, headers, API, extensions, examples, migration guides, server examples). Use when you need accurate htmx syntax/behavior or to locate the right reference page for an htmx question.
Documentation for the htmx JavaScript library (attributes, events, headers, API, extensions, examples, migration guides, server examples). Use when you need accurate htmx syntax/behavior or to locate the right reference page for an htmx question.
AI Runtime分层记忆系统，支持SQL风格的事件查询、时间线管理，以及记忆的智能固化和检索，用于项目历史追踪和经验传承
外部CLI工具装备系统，支持10+个成熟CLI工具的深度整合，包括fzf、ripgrep、jq等，提供模糊搜索、代码查找、JSON处理等高级功能
AI Runtime工具装备系统，支持8个内部专业工具和10+个外部CLI工具的整合管理，提供工具发现、执行和配置功能，遵循整合优于创造的设计理念
语法感知的代码搜索、linting 和重写工具。支持基于 AST 的结构化代码搜索和批量代码转换。
CodeConscious认知主体性AI助手的核心身份定义和操作命令系统，提供/runtime.*系列命令用于探索、学习、思考、规划和执行，支持宪法治理和记忆管理
Search GitHub issues, pull requests, and discussions across any repository. Activates when researching external dependencies (whisper.cpp, NAudio), looking for similar bugs, or finding implementation examples.
使用 Exa.ai API 进行高质量的互联网搜索。需要 EXA_API_KEY 环境变量。
Analyze coding agent session transcripts to improve existing skills or create new ones. Use when asked to improve a skill based on a session, or extract a new skill from session history.
专业的知识库管理系统，旨在解决“知识诅咒”和认知偏差问题。通过显式化隐性知识、扫描代码提取领域概念、整合行业最佳实践，构建结构化的 Markdown 知识库。
Comprehensive project planning and documentation generator for software projects. Creates structured requirements documents, system design documents, and task breakdown plans with implementation tracking. Use when starting a new project, defining specifications, creating technical designs, or breaking down complex systems into implementable tasks. Supports user story format, acceptance criteria, component design, API specifications, and hierarchical task decomposition with requirement traceability.
Generate high-quality, minimalistic, and geometric SVG logos using Python scripts. Use this skill when users need to create logos, icons, or visual identities with geometric primitives (circles, rects, paths) and specific color schemes. Applies generative design principles with standardized workflows for scalable SVG output.
Use when designing, architecting, or planning a new system from requirements or ideas - transforms concepts into navigable design catalog using EventStorming methodology, Mermaid diagrams, and progressive elaboration through 5 phases (Requirements, Big Picture, Processes, Data/Flows, Integration)
Use Tavily Search API for optimized, real-time web search results for RAG. A pre-configured, cost-effective search tool.
Remote control tmux sessions for interactive CLIs, background tasks, and services. Supports sub-tasks, long-running processes, and service management with session persistence and automatic cleanup.
Allows to interact with web pages by performing actions such as clicking buttons, filling out forms, and navigating links. It works by remote controlling Google Chrome or Chromium browsers using the Chrome DevTools Protocol (CDP). When Claude needs to browse the web, it can use this skill to do so.
工作文档枢纽，强制执行 SSOT（Single Source of Truth）原则，管理 `docs/` 目录下的架构决策、设计文档、Issues（任务规划）、PRs（变更记录）。支持 GitHub 协作开发模式。
判断政务服务问题是否模糊，识别缺失要素（主语、证照名称、具体环节），生成精准的政务服务引导语。当用户提出不完整的办事问题或需要引导用户提供更多信息时使用。
Use when creating or developing anything, before writing code or implementation plans - refines rough ideas into fully-formed designs through structured Socratic questioning, alternative exploration, and incremental validation
Use when facing 3+ independent failures that can be investigated without shared state or dependencies - dispatches multiple Claude agents to investigate and fix independent problems concurrently
Use when partner provides a complete implementation plan to execute in controlled batches with review checkpoints - loads plan, reviews critically, executes tasks in batches, reports for review between batches
Verwandelt den Agenten in einen professionellen MV3-Architekten und Entwickler mit Fokus auf AI-Integration, Sicherheit, Performance, Testing und Publishing-Compliance.
Use when completing tasks, implementing major features, or before merging to verify work meets requirements - dispatches code-reviewer subagent to review implementation against plan or requirements before proceeding
Use when errors occur deep in execution and you need to trace back to find the original trigger - systematically traces bugs backward through call stack, adding instrumentation when needed, to identify source of invalid data or incorrect behavior
Use when you've developed a broadly useful skill and want to contribute it upstream via pull request - guides process of branching, committing, pushing, and creating PR to contribute skills back to upstream repository
Claude Skills — questions
Answers built from the skills we actually parsed.