SiteToMarkdown

Hey everyone, I built SiteToMarkdown to fix a problem I kept running into with LLMs: pasting a single docs URL never gives enough context. Most products spread their documentation across dozens or hundreds of pages, and tools like NotebookLM, ChatGPT, or RAG systems only see a thin slice of that when you feed them one link at a time. SiteToMarkdown lets you paste the root of your docs or help center (for example, /docs/ or /help/) and automatically crawls everything underneath. It renders JavaScript, handles SPAs, and works around generic bot blocking so it can see the same content a real user would. Then it strips out HTML, nav, and ads and merges just the main content into one clean Markdown file. You can drop that file into NotebookLM as a single source, upload it to ChatGPT or Claude, keep it in your repo for GitHub Copilot, or plug it into your own RAG and agent pipelines. The goal is to make full-site, docs-grounded context as simple as downloading one Markdown file—no subscriptions, no custom scrapers, and no endless copy‑paste.