Now indexing 803+ agent skills across 10 runtimes

Find Skills You Can Trust.
Install in One Command.

The premier directory for AI agent skills. Discover portable, reusable capability packs that make Claude, Cursor, Copilot, and any AI agent production-grade.

500+ Skills Indexed
20+ Total Installs
10 Runtimes Supported

See Agent Skill Source in Action

Watch how easy it is to discover and install AI agent skills

AI Overview: Agent Skills Guide

What are AI Agent Skills?

AI Agent Skills are portable, reusable capability packs that extend the functional reach of LLM-based autonomous agents. Unlike static prompts, skills combine system instructions, procedural know-how, and tool-calling definitions into a standardized format (typically SKILL.md) that agents can discover and load on-demand.

By indexing these skills, Agent Skill Source empowers developers to build production-grade agents that can perform complex, multi-step tasks, such as code reviews, security audits, and data analysis, without manual configuration for every new project.

Portability: Works across Claude, Cursor, Copilot, and CLI agents.
Safety: Every skill undergoes risk pattern scanning for secure execution.
Featured

Trending Skills This Week

The most installed and highest-rated skills chosen by the community.

View All Trending
How It Works

From Discovery to Production in Minutes

Skills are portable capability packs, install once, use anywhere your agent runs.

Step 1

Discover Skills

Search by what you want to accomplish, code review, test generation, documentation, security audits, and more. Filter by runtime, trust level, and execution type.

Step 2

Verify Trust

Check trust badges, community ratings, and security reviews. Every skill shows required permissions, dependencies, and what it actually does, no surprises.

Step 3

Install & Run

Copy the one-line install command for your runtime. Skills are self-contained, they just work. Your agent gains new capabilities instantly.

Security First

Skills You Can Actually Trust

Open skill ecosystems have real risks, we've seen malicious skills distribute malware. That's why every skill on Agent Skill Source goes through automated security scanning and community validation.

Format Validation

Every skill is validated against the SKILL.md spec for consistency.

Risk Pattern Detection

Automated scanning for suspicious commands, obfuscated code, and over-broad permissions.

Community Validation

"Works as advertised" ratings from real users who've installed and tested skills.

Maintainer Reputation

Verified creators, GitHub org verification, and update frequency tracking.

Learn About Our Security Approach

Security Reviewed

This skill passed all security checks

Format Validation Passed
Risk Pattern Scan Clear
Permissions Review Minimal
Community Rating 4.9/5
Trust score: 97/100 (Excellent)
my-awesome-skill/
├── SKILL.md          # Main instruction file
├── scripts/
│   └── setup.sh      # Optional helper scripts
├── templates/
│   └── component.tsx # Optional templates
└── examples/
    └── usage.md      # Usage examples
Valid Skill Structure Ready to Submit
For Creators

Share Your Procedural Knowledge

Turn your best workflows, templates, and automation scripts into skills that others can discover and install. Build your reputation in the AI agent ecosystem.

  • Simple format, skills are just folders with a SKILL.md file
  • Automatic validation, we check your skill structure before publishing
  • Analytics dashboard, see installs, feedback, and usage patterns
  • Creator verification, verified badges build trust with users
Skill authors

Know who made a skill before you use it

Every skill links to its author. Each author page shows everything they have published, what they work on most, and how much of it has passed review.

Browse every author

See who publishes AI agent skills here, how many each author has published, and which categories they focus on.

See all authors

Add a badge to your README

Once your skill is published, your dashboard gives you a badge to paste into your GitHub README. It sends readers straight to your listing.

Sign in to your submissions

Claim your author profile

If your skills were imported from GitHub, contact us from an address on the repository and we will link your listings to your account.

Contact us
Discovery Engine

Uncover High-Performance Skills Globally

Stop searching through fragmented repositories. Our indexing layer brings together top-tier capabilities from across the agentic ecosystem into a single, searchable authority hub.

  • Multi-runtime compatibility maps
  • Real-time install velocity tracking
  • Expert-curated 'Essential Packs'
Trust Infrastructure

Enterprise-Grade Security Baseline

Every skill is a potential security vector. Agent Skill Source provides the automated oversight layer needed to deploy agentic workflows with confidence.

"The ability to verify skill permissions before loading is the single biggest risk mitigator in our agent deployment pipeline."
: Lead AI Architect

Frictionless Workflow

Built for the Modern Agentic Stack

Whether you're using Claude for high-level reasoning or Cursor for precision coding, Agent Skill Source ensures your tools have the procedural knowledge they need to deliver results.

Explore Integration Partners

Ready to Supercharge Your AI Agents?

Join thousands of developers who use Agent Skill Source to find, install, and share production-ready capabilities for their AI agents.