Skill Seekers

Source: yusufkaraaslan/Skill_Seekers
Version: 2.7.0
Author: @yusufkaraaslan
License: MIT

Overview

Convert documentation websites, GitHub repositories, and PDFs into Claude AI skills with automatic conflict detection. Supports multi-LLM platforms.

Key Features

🌐 Documentation Scraping

Scrape any documentation site and convert to skill format.

📄 PDF Support

Extract knowledge from PDF documents.

🐙 GitHub Repository Scraping

Three-stream architecture for efficient repo analysis.

🤖 Multi-LLM Platform Support

Works with: Claude, GPT, Gemini, Anthropic API

🎯 Codebase Analysis & AI Enhancement

Analyze codebases and generate enhanced AI skills.

Installation

PyPI (Recommended)

bash

pip install skill-seekers

uv (Modern Python Tool)

bash

uv add skill-seekers

MCP Integration

Add to MCP config for Claude Code / Antigravity:

json

{
  "mcpServers": {
    "skill-seekers": {
      "command": "uvx",
      "args": ["skill-seekers", "mcp"]
    }
  }
}

Usage Examples

Documentation Scraping

bash

skill-seekers scrape https://docs.example.com --output skills/

PDF Extraction

bash

skill-seekers pdf ./document.pdf --output skills/

GitHub Repository Scraping

bash

skill-seekers github owner/repo --output skills/

Unified Multi-Source

bash

skill-seekers multi \
  --url https://docs.example.com \
  --github owner/repo \
  --pdf ./doc.pdf \
  --output skills/

MCP Tools Available (18 Total)

Tool	Description
`scrape_docs`	Scrape documentation website
`extract_pdf`	Extract from PDF
`scrape_github`	Analyze GitHub repo
`generate_skill`	Generate skill file
`upload_skill`	Upload to Claude
...	And 13 more

Output Structure

code

skills/
├── SKILL.md           # Main skill file
├── knowledge/         # Scraped knowledge
│   ├── category1.md
│   └── category2.md
└── metadata.json      # Skill metadata

Related Skills

•documentation-maintenance

Added to Knowledge Base: 2026-01-22
Credit: @yusufkaraaslan