Converts web pages to clean, LLM-optimized Markdown without needing a browser.
Readdown-mcp is an MCP server designed to efficiently transform web content into a clean, LLM-optimized Markdown format. It processes web pages to extract only the most relevant information, converting HTML into structured Markdown while minimizing token count for more cost-effective API calls. This lightweight tool operates server-side without requiring a full browser, making it an ideal solution for preparing web data for consumption by Large Language Models, complete with built-in token estimation and metadata extraction.
主な機能
01Converts web pages to clean, LLM-optimized Markdown
02Extracts metadata (title, author, date)
03Token-efficient output for cheaper API calls
04Built-in token estimation
05Server-side operation without a browser
060 GitHub stars
ユースケース
01Efficiently extracting structured Markdown from web pages
02Integrating with AI code assistants like Claude Desktop or Cursor
03Preparing web content for Large Language Model consumption
