iamsupersocks/supersocks-url-scraper ? reverse-engineered prompt

Build me a small URL scraper in Python that can take a link and return a clean JSON result with the page title, metadata, a readable summary, publication date, content type, and any warnings if the page was only partially extracted.

I want it to work well for normal web pages, articles, PDFs, and images, and to fall back gracefully when a page is messy or blocked. If it can’t get the full content, it should still give me useful metadata and tell me what went wrong. Please include a simple command line command for one off use, and also a local web service with endpoints for health, summarizing, reading, and markdown output.

Make it dependency light for the basic case, but support better extraction when extra tools are available. If you need to look up current docs online for anything browser related, go ahead. Please also make it easy to run locally with Docker and good defaults for cron or automation use.