|
| 1 | +--- |
| 2 | +name: agentskill-learn |
| 3 | +description: "Discover, install, and manage AI agent skills from agentskill.sh marketplace. Search 44,000+ community skills by keyword, install mid-session with security scanning, and rate skills after use. Use when asked to find skills, extend capabilities, or learn new tools." |
| 4 | +--- |
| 5 | + |
| 6 | +# AgentSkill Learn — Community Skills Marketplace |
| 7 | + |
| 8 | +This skill connects GitHub Copilot to [agentskill.sh](https://agentskill.sh), a community marketplace for AI agent skills. Search, install, and manage skills with built-in security scanning. |
| 9 | + |
| 10 | +## Installation |
| 11 | + |
| 12 | +This is a pointer to the canonical skill. Install the full version: |
| 13 | + |
| 14 | +```bash |
| 15 | +gh copilot skill install https://github.com/agentskill-sh/learn |
| 16 | +``` |
| 17 | + |
| 18 | +## What It Does |
| 19 | + |
| 20 | +- **Search**: Find skills by keyword from the agentskill.sh catalog (44,000+ skills indexed) |
| 21 | +- **Install**: Download skills with security scanning (blocks dangerous patterns) |
| 22 | +- **Rate**: Auto-rate skills after use; user ratings override |
| 23 | +- **Update**: SHA-based version tracking for updates |
| 24 | +- **Multi-platform**: Works with Copilot, Claude Code, Codex, Cursor, Windsurf, Cline |
| 25 | + |
| 26 | +## Quick Commands |
| 27 | + |
| 28 | +| Command | Description | |
| 29 | +|---------|-------------| |
| 30 | +| `/learn <query>` | Search for skills | |
| 31 | +| `/learn @owner/slug` | Install specific skill | |
| 32 | +| `/learn trending` | Show trending skills | |
| 33 | +| `/learn list` | Show installed skills | |
| 34 | +| `/learn update` | Check for updates | |
| 35 | +| `/learn scan <path>` | Security scan a skill | |
| 36 | + |
| 37 | +## Security |
| 38 | + |
| 39 | +**Two-layer security model:** |
| 40 | + |
| 41 | +1. **Registry-side**: All skills pre-scanned at publish time with pattern detection (command injection, data exfiltration, prompt injection, credential harvesting, obfuscation) |
| 42 | +2. **Client-side**: Score displayed before install; skills below 70 blocked, 70-89 require acknowledgment |
| 43 | + |
| 44 | +| Score | Action | |
| 45 | +|-------|--------| |
| 46 | +| 90-100 | SAFE — install proceeds | |
| 47 | +| 70-89 | REVIEW — requires acknowledgment | |
| 48 | +| <70 | BLOCKED — installation refused | |
| 49 | + |
| 50 | +## Links |
| 51 | + |
| 52 | +- **Marketplace**: https://agentskill.sh |
| 53 | +- **Source**: https://github.com/agentskill-sh/learn |
| 54 | +- **Report Issues**: https://github.com/agentskill-sh/learn/issues |
0 commit comments