Readdown Mcp

Gives Claude two tools for turning web content into clean Markdown without spinning up a browser. fetch_markdown grabs a URL via HTTP and returns optimized output with metadata and token counts. convert_html does the same for raw HTML strings you already have. Built on readdown, which replaces the usual Readability plus Turndown stack with a single 5KB package that's tuned for LLM context windows. Useful when you need Claude to ingest article content or documentation pages efficiently, or when you're building workflows that process web content at scale and want to minimize token overhead.