MCP Fetch is a server that enables Claude Desktop and other MCP clients to retrieve and process web content. Built by kazuph, it fetches URLs, extracts content as markdown, and handles images intelligently - automatically processing them for optimal viewing (merging vertically with size limits) and preparing them for clipboard operations. The implementation uses Node.js with libraries like Sharp for image processing, JSDOM and Readability for content extraction, and includes features like robots.txt compliance and animated GIF handling. Particularly useful for research workflows where users need to bring external web content into their AI conversations without manual copying and formatting.
JeremyNixon