About readdown-mcp
readdown-mcp is an MCP server that converts web pages or supplied HTML into clean Markdown for LLM use. It fetches pages over HTTP without a browser and includes metadata and token estimates.
Key features
- 01Fetch URLs and convert pages to Markdown
- 02Convert supplied HTML to Markdown
- 03Extract title, author, and date
- 04Estimate output token count
Use cases
- 01Prepare a web page for LLM processing
- 02Convert existing HTML content to Markdown
- 03Check page metadata and estimated token use
MCP support
It exposes URL fetching and HTML-to-Markdown conversion as tools for MCP clients.
Configuration
Local server config is the same for every client. Copy it into your client’s MCP config file.
docker run -i --rm ghcr.io/zcag/readdown-mcp:1.0.0Local servers need a runtime on your machine (e.g. Node.js 18+). Config file locations differ by client; see your client’s docs.
ghcr.io/zcag/readdown-mcp:1.0.0stdioTools
Local servers expose tools only once running on your machine, and the maintainer has not provided a tool list.
Details
From the project READMEreaddown-mcp
MCP server for converting web pages to clean, LLM-optimized Markdown.
Lightweight — uses HTTP fetch, no browser needed. Powered by readdown.
Tools
fetch_markdown
Fetch a URL and convert it to clean Markdown with metadata and token count.
convert_html
Convert an HTML string to Markdown. Use when you already have the HTML.
Install
Claude Desktop
Add to claude_desktop_config.json:
{
"mcpServers": {
"readdown": {
"command": "npx",
"args": ["-y", "github:zcag/readdown-mcp"]
}
}
}
Cursor
Add to .cursor/mcp.json:
{
"mcpServers": {
"readdown": {
"command": "npx",
"args": ["-y", "github:zcag/readdown-mcp"]
}
}
}
Why readdown?
Most web-fetching MCP servers use @mozilla/readability + turndown (two packages, 65KB).
readdown replaces both in a single 5KB package with better LLM optimization:
- Token-efficient output (fewer tokens = cheaper API calls)
- Built-in token estimation
- Metadata extraction (title, author, date)
- Works server-side with linkedom (no browser DOM needed)
License
MIT
Comments
New to MCP?
From what the protocol is to installing your first server, these answers cover most first questions.
An open protocol that defines how AI applications connect to external tools and data in a uniform way. Clients talk to the model; MCP servers provide tools, resources and prompt templates. They exchange standard messages, so one server can be reused by many clients.