agent-skills
Production-grade engineering skills for AI coding agents.
Agent Skills is a production-grade skill library that provides AI coding agents with structured engineering workflows, quality gates, and best practices used by senior engineers. The collection includes 24 skills organized across six development phases (Define, Plan, Build, Verify, Review, Ship) plus 4 specialist agent personas and 8 lifecycle slash commands. Each skill encodes specific workflows with anti-rationalization tables, verification gates, and evidence requirements to prevent agents from skipping critical steps like specs, tests, and security reviews. Built on Google engineering culture principles, it works with 70+ AI agents including Claude Code, Cursor, Codex, GitHub Copilot, Windsurf, and Gemini CLI.
Key Features
Use Cases
- 01Enforcing test-driven development and code review quality gates before AI agents merge changes
- 02Running structured requirements interviews to clarify underspecified feature requests
- 03Implementing incremental vertical slices with feature flags and rollback-friendly changes
- 04Conducting five-axis code reviews with staff engineer-level standards before production deploys
- 05Performing OWASP Top 10 security audits and threat modeling on user-facing features
- 06Orchestrating multi-phase development workflows from spec writing through production shipping
Related Skills
View moresuperpowers
An agentic skills framework & software development methodology that works.
skills
Skills for Real Engineers. Straight from my .agents directory.
skills
Public repository for Agent Skills
ponytail
Makes your AI agent think like the laziest senior dev in the room. The best code is the code you never wrote.
agent-skills — FAQ
What is agent-skills?+
Agent Skills is a library of 24 production-grade workflow skills that encode senior engineering practices for AI coding agents. Each skill provides step-by-step processes, verification gates, and anti-rationalization tables to ensure agents follow best practices across the full development lifecycle from spec to production.
How do I install agent-skills?+
The fastest method is using the skills CLI: run 'npx skills add addyosmani/agent-skills' to install all 24 skills into any of 70+ supported agents. You can also install individual skills with --skill flag, or use native integrations for Claude Code (/plugin marketplace add), Cursor (.cursor/skills/), Codex (codex plugin marketplace add), or other clients.
Which AI clients work with agent-skills?+
Agent Skills supports 70+ AI agents including Claude Code, Cursor, Codex, GitHub Copilot, Cline, Windsurf, Gemini CLI, Antigravity CLI, OpenCode, Command Code, and Kiro. Skills are plain Markdown and work with any agent that accepts system prompts or instruction files.
Is agent-skills free to use?+
Yes, agent-skills is open source and released under the MIT license. You can use, modify, and distribute the skills in your projects, teams, and tools without cost.
Do I need API keys or prerequisites?+
No external API keys are required. You only need an AI coding agent that supports skill/plugin installation and Git for cloning or the npx command for CLI installation. Some skills reference browser testing with DevTools MCP, which requires the Chrome DevTools MCP server if you want live runtime data.
What's the difference between skills, commands, and agent personas?+
Skills are structured workflows (24 total) that agents follow for specific tasks. Commands are 8 slash shortcuts (/spec, /plan, /build, etc.) that activate the right skills automatically. Agent personas are 4 pre-configured specialists (code-reviewer, test-engineer, security-auditor, web-performance-auditor) for targeted reviews.
How do I install agent-skills?+
Open the source repository on GitHub and follow its README. agent-skills is a skill — MCP Agents Market links you directly to the official repo.
Is agent-skills free?+
agent-skills is an open-source project hosted on GitHub. Check the repository for its license and any usage requirements.