Skip to content
Conduit

skill

skillio

AI Agent 的能力发现与装配中枢 - Discover and install AI skills with natural language

copyboy/skilliocommunitylistedunreviewed
Use skill

Set up

Skills are installed by placing their folder where your agent loads skills.

Download ZIP from GitHub Open the repository

The download comes from github.com/copyboy/skillio itself, not from our server, so what you get is exactly what its author published.

Install instructions come from the README

Skills do not use an MCP config. Follow the steps in the README below.

README

synced Oct 23, 2026

Skillio

AI Agent 的能力发现与装配中枢

PyPI version License: MIT

What is Skillio?

Skillio is an intent-to-capability matching engine for AI Agents. Instead of knowing which tools exist, you simply describe what you want to do, and Skillio finds the right skill for you.

# Traditional way: You need to know the tool name
$ yt-dlp https://youtube.com/watch?v=xxx

# Skillio way: Just describe your intent
$ skillio search "download YouTube video"
→ Recommended: video-downloader (powered by yt-dlp)

$ skillio install video-downloader
→ Skill installed to ~/.cursor/skills/

Core Philosophy

Old approach (Supply-side): User needs to know "what tools exist"

Skillio approach (Demand-side): User just describes "what they want to do"

Installation

pip install skillio

Quick Start

Search for Skills

# Natural language search
skillio search "I want to convert video to GIF"
skillio search "下载 B 站视频"
skillio search "PDF to Word"

# Keyword search
skillio search --keyword "video download"

Install Skills

# Install a skill
skillio install video-downloader

# Install to specific location
skillio install video-downloader --target ~/.cursor/skills/

Manage Installed Skills

# List installed skills
skillio list

# Show skill details
skillio info video-downloader

# Update a skill
skillio update video-downloader

# Remove a skill
skillio remove video-downloader

Architecture

┌─────────────────────────────────────────────────────┐
│                   User / AI Agent                    │
│              "I want to convert video to GIF"        │
└─────────────────────┬───────────────────────────────┘
                      ▼
┌─────────────────────────────────────────────────────┐
│              Layer 3: Intent Understanding           │
│   - Natural language parsing                         │
│   - Capability tag extraction                        │
└─────────────────────┬───────────────────────────────┘
                      ▼
┌─────────────────────────────────────────────────────┐
│              Layer 2: Capability Matching            │
│   - Reverse index: tags → skills                     │
│   - Ranking: quality, relevance, popularity          │
└─────────────────────┬───────────────────────────────┘
                      ▼
┌─────────────────────────────────────────────────────┐
│              Layer 1: Skill Repository               │
│   - Metadata storage                                 │
│   - Version management                               │
│   - Source tracking (GitHub/Docs/Manual)             │
└─────────────────────────────────────────────────────┘

Skill Schema

Each skill is defined with structured metadata:

name: video-downloader
version: 1.0.0
source:
  type: github
  repo: yt-dlp/yt-dlp
description: Download videos from YouTube, Bilibili, and 1000+ sites
capabilities:
  - video download
  - YouTube
  - Bilibili
  - batch download
scenarios:
  - download video
  - save video for offline
  - extract audio from video
dependencies:
  - python >= 3.8
  - ffmpeg

Roadmap

  • Phase 1: MVP - Intent matching engine (CLI)
  • Phase 2: Platform - Online registry (skillio.dev)
  • Phase 3: Ecosystem - Cross-platform skill standard

Contributing

We welcome contributions! See CONTRIBUTING.md for guidelines.

License

MIT License - see LICENSE for details.

Links

↑↓ navigate openesc close