ray0404/PyScraper-md ? reverse-engineered prompt

Build me a Python tool that takes a web page or a local HTML file and turns it into clean GitHub Flavored Markdown.

I want it to work well for both simple sites and JavaScript heavy pages, and I’d like a normal command line flow where I can save the result to a file or print it out. It should keep useful things like headings, tables, code blocks, links, and images, and let me choose how images and SVGs are handled, like keeping them remote, downloading them, or turning them into inline data.

It would also be great if it could crawl a site a few pages deep, stay within the same section if asked, and save multiple pages into a folder. Please include a small web interface for running scrapes in the browser, and make the core scraper usable as a Python library too. If you need to, look up current docs for Playwright or Flask so the dynamic scraping and web app work smoothly.