Fetch
Downloads a URL and converts HTML into Markdown so the model reads text instead of markup.
What it does
Downloads a URL and converts HTML into Markdown so the model reads text instead of markup. Handles chunking and start-index paging for long documents. It respects robots.txt by default, which is the right default but does mean some sites simply will not load.
Install
Terminal
uvx mcp-server-fetch
Client configuration
Paste this into your MCP client config (Claude Desktop, Cursor, or any client that reads mcpServers), then restart the client.
mcp.json
{
"mcpServers": {
"fetch": {
"command": "uvx",
"args": [
"mcp-server-fetch"
]
}
}
}
When to use it
Reach for this when the task needs web inside the agent loop. Skip it if a plain shell command or a one-off script is enough — every server you add costs context and adds a failure mode.