How yusufkaraaslan/skill_seekers works
Project type and purpose
• This is a Python-based CLI tool with an optional “MCP server” component. It is not a classical web app but a developer tool that automates:
• Scraping HTML/PDF documentation websites
• Organizing and AI-enhancing content
• Packaging it into a Claude AI “skill” zip file
• Optionally uploading it via an API or manual process
• The README confirms it’s a full-featured ETL + AI-pipeline tool, with both CLI scripts (cli/) and an MCP-compatible micro-service (mcp/).
File structure
• Top‐level directories:
• cli/ – all entry-point scripts (scraper, enhancer, packager, uploader, utils)
• mcp/ – a small server implementing Model Context Protocol tools for Claude Code integration
• configs/ – JSON presets for different frameworks (React, Django, Godot, etc.)
• docs/ – extensive user and technical documentation
• tests/ – unit and integration tests
• Supporting files: setup scripts, requirements, GitHub workflows, README, CONTRIBUTING, etc.
• No traditional database: persistent state is file‐system based (output/ cache, checkpoints).
README for additional insights
• Key external dependencies: websites or PDFs, Claude Code locally (via MCP) or Anthropic API, optional OCR tools.
• Blue for core pipeline stages (Scraper, Enhancer, Packager)
• Green for external systems (Websites, AI services)
• Orange for configuration and caching
• Purple for the MCP server