yusufkaraaslan/Skill_Seekers ? reverse-engineered prompt

I want to build a tool that's like a universal data prepper for large language models. Let's call it 'Skill Seekers'.

The main idea is that I can point it at any source of information—a documentation website, a whole GitHub repository, a PDF, or even a YouTube video—and it will automatically scrape and process everything. It should be smart enough to handle code blocks, text, and other content, turning a complex source into a clean, structured knowledge base.

Then, I want to be able to export this knowledge asset for different AI systems. For instance, I should be able to run a command to package it for Claude, or another to prepare it for a custom OpenAI GPT. The final app should be a Python CLI tool that's super easy to use, like skill-seekers create <source-url>. The whole point is to make it incredibly fast to teach an AI about a new library or project without spending days manually copying and pasting docs.