A tool to convert any website into LLM-ready text - structured markdown.
Note
Currently it not yet a hosted service. you can clone and use locally instead.
Warning
The current approach work best for server rendered website and would struggle for heavily client rendered websites.
- Clone repo:
git clone url
- API (virtual env and dependecites):
- create a python virtual env.
cd api
python3 -m venv venv
source venv/bin/activae
- Install dependecies & start server.
pip install -r requirements.txt
fastapi run main.py
the server should run by default on localhost:8000
- Frontend:
The frontend is mostly not needed, but provides an intuitive interface.
use curl if you don't need the frontend.
curl http://localhost:8000/gx?url=URL_TO_CRAWL&max_depth=2 (best within 1-3)
Install the frontend dependies and run in needed.
cd frontend/
pnpm i
pmpm dev