This is a web scraping toolkit that turns messy HTML into structured data. You get 11 tools split across fetching (with custom headers and auth support), conversion (HTML to Markdown, JSON to Markdown, text extraction), and extraction (article content via readability, metadata, links, images, JSON-LD structured data). Runs over stdio, so it plugs into Claude Desktop or any MCP client. Useful when you need to pull clean content from web pages without writing your own scraper, whether that's grabbing article text, converting documentation to Markdown, or extracting SEO metadata. Built on Python 3.11+ and ships with automated publishing scripts for PyPI.
claude mcp add --transport stdio huoshuiai42-huoshui-fetch -- uvx huoshui-fetch