Claude Code Skills & Plugins is a library of reusable expertise packages for AI coding agents. Each skill is a folder holding a SKILL.md file of structured instructions, workflows, and decision frameworks, optional Python scripts, and reference material such as templates and checklists. The collection covers 388 skills across 20 domains, alongside 118 agents and 150 slash commands.
The same repository targets thirteen coding tools. It works natively as Claude Code plugins and as Codex, Gemini CLI, Hermes Agent, and Mistral Vibe skills, and a conversion script rewrites every skill into the native format of Cursor, Aider, Kilo Code, Windsurf, OpenCode, Augment, and Antigravity, generating per-tool install and verification notes as it goes.
Domains include core and advanced engineering, product, marketing, productivity, academic research, research operations, project management, regulatory and quality management, compliance, C-level advisory, business and growth, business operations, commercial, finance, an agent launcher, a loop library, and a Markdown-to-HTML pipeline. Personas sit above skills as agent identities with curated skill loadouts and communication styles, and an orchestration protocol describes four patterns for combining personas, skills, and agents across domain boundaries.
727 Python command-line tools ship with the skills, all written against the standard library so nothing has to be installed to run them. A separate skill audits any skill before installation, scanning for command injection, code execution, data exfiltration, prompt injection, dependency supply-chain risks, and privilege escalation, and returning a PASS, WARN, or FAIL verdict with remediation notes.
Features
- Skill packages: each skill combines a SKILL.md instruction file with optional scripts, references, and assets
- Thirteen target tools: native support for Claude Code, Codex, Gemini CLI, OpenClaw, Hermes Agent, and Mistral Vibe, plus conversion for seven more
- Format conversion: one script rewrites all skills into Cursor rules, Aider conventions, Windsurf skills, and other tool-native layouts
- Twenty domains: engineering, product, marketing, research, compliance, C-level advisory, finance, and more
- Personas: cross-domain agent identities such as startup CTO, growth marketer, and solo founder, with their own skill loadouts
- Orchestration patterns: solo sprint, domain deep-dive, multi-agent handoff, and skill chain
- Standard-library Python tools: 727 command-line scripts that run wherever Python runs
- Skill security auditor: scans a skill for injection, exfiltration, and supply-chain risks before installation
- Semantic versioning: script arguments, plugin source paths, and SKILL.md structures stay stable within patch releases
Runs in
Aider
Command-line LLM pair programmer that maps your repo, edits files, and auto-commits each change to git
Claude Code
Terminal-based agentic coding tool that reads your codebase and runs tasks through natural-language commands
Codex
Lightweight coding agent that runs in your terminal
Hermes
Self-improving AI agent with a learning loop that creates and refines skills, recalls past sessions, and runs across chat platforms