Agent Skills

Active
GitHub JavaScript MIT

Description

Agent Skills is a curated collection of production-grade engineering skills for AI coding agents, maintained by Addy Osmani, providing battle-tested best practices and operational conventions.

Key Features

  • 23 engineering skills — Covers full dev lifecycle: spec definition, planning, building, testing, review, simplification, and shipping
  • 7 slash commands — /spec, /plan, /build, /test, /review, /code-simplify, /ship for different dev phases
  • Multi-tool compatibility — Supports Claude Code, Cursor, Gemini CLI, Windsurf, OpenCode, GitHub Copilot, and more
  • Structured workflows — Each skill includes steps, verification gates, and anti-rationalization tables for code quality
  • Auto skill activation — Automatically selects appropriate skills based on current activity (API design, frontend engineering, etc.)
  • Meta skill system — using-agent-skills meta-skill maps work to the correct skill workflow automatically

Use Cases

💡 Providing standardized engineering practices and quality gates for AI coding assistants
💡 Establishing consistent code review and release processes across teams
💡 Guiding AI to follow best practices in test-driven development
💡 Ensuring AI-generated code meets production-grade engineering standards

Strengths & Limitations

Strengths

  • Actively maintained, recent updates
  • High community interest (91.7k stars)
  • Permissive open-source license (MIT)
  • Responsive to issues, low backlog

Quick Start

# Option 1: Install via Claude Code Marketplace
/plugin marketplace add addyosmani/agent-skills
/plugin install agent-skills@addy-agent-skills

# Option 2: Clone the repository
 git clone https://github.com/addyosmani/agent-skills.git

# Load in Claude Code
claude --plugin-dir /path/to/agent-skills

# Option 3: For Cursor
# Copy SKILL.md files to .cursor/rules/ directory

# Usage examples:
# /spec - Define what to build
# /plan - Plan how to build it
# /build - Build incrementally
# /test - Write tests
# /review - Code review
# /ship - Ship to production

Related Projects

Claude Code

143.8k · Python
Active A

Claude Code is an agentic coding tool by Anthropic that lives in your terminal, understands your codebase, and helps you code faster through natural language commands.

coding-agentcliterminal +2
  • · Terminal-native experience — Runs directly in your terminal with full codebase context and structure understanding
  • · Natural language coding — Execute coding, refactoring, and debugging tasks through natural language commands
  • · Git workflow integration — Handles branching, commits, PR creation, and other Git operations

GitHub Copilot CLI

11.1k · Shell
Active B

GitHub Copilot CLI brings the power of Copilot coding agent directly to your terminal. Supports code generation, command suggestions, error fixing and more.

copilotcligithub +2
  • · Terminal-native development — Copilot coding agent runs directly in your CLI without context switching
  • · Deep GitHub integration — access repos, issues, and PRs with natural language
  • · Agentic capabilities — build, edit, debug, and refactor code with complex task planning and execution

jcode

19.0k · Rust
Active A

A next-generation coding-agent harness built for multi-session workflows, with a focus on extreme performance, low memory, and deep MCP/tool customization.

coding-agentrustterminal +2
  • · Extreme footprint - 27.8 MB per single session and 117 MB for 10 sessions, 5-15x lighter than peers
  • · Sub-millisecond startup - 14 ms to first frame and 49 ms to first input, 200x+ faster than Claude Code
  • · Multi-session swarm - coordinates multiple agents in the same repo with conflict detection and broadcast messaging