Converts web pages to clean Markdown with 90-95% size reduction, making HTML digestible for AI context windows. Built on trafilatura and BeautifulSoup4, it strips scripts, styles, and navigation while preserving tables, images, and links. The Playwright integration handles JavaScript-heavy SPAs and authenticated pages by executing client-side code and leveraging your browser profile with cookies. Stream processing and configurable size limits (1MB-50MB) keep large pages manageable, while optional caching speeds up repeated conversions. Reach for this when you need to feed web content into Claude but the raw HTML is too bloated or requires JavaScript rendering to display properly.
claude mcp add --transport stdio io.github.sunshad0w-html2md uvx html2md