agentic-awesome-skills

Why we included this project

Teams running Claude Code, Codex, Cursor, or Gemini CLI often hit the same wall: the agent ships with a few built-in skills, but turning that into a repeatable workflow means finding and wiring up the right playbooks yourself. This project gathers more than two thousand reusable SKILL.md playbooks into one local catalog and adds a small control plane on top. Your coding agent searches the full set, picks the exact skills, and validates the stack in memory before anything touches your project; AAS itself does not rank or recommend, so the choice stays with the agent. The read-only MCP server and CLI then produce an immutable plan for human review, which is what you want when agent behavior has to be reproducible and auditable across a team. If you are standardizing how your coding agents work rather than experimenting with one-off prompts, this is a practical way to browse the catalog and lock down a stack your team can reuse.