fastmcp

Install fastmcp with pip, create a FastMCP instance, and decorate functions with @mcp.tool to expose them as tools. Start from a template like an API wrapper or database server, then validate with fastmcp inspect and fastmcp call before installing into a client.

Run fastmcp inspect server.py:mcp to verify the server loads, then fastmcp list to enumerate tools and fastmcp call to execute each tool with real arguments. For HTTP transport, start with fastmcp run --transport http and call the local /mcp endpoint.

Yes, FastMCP supports direct installation with commands like fastmcp install claude-code server.py, fastmcp install claude-desktop server.py, and fastmcp install cursor server.py -e . Declare dependencies explicitly since client installs run in isolated environments.

Use a read-only connection via SQLite URI mode, reject any SQL that does not start with SELECT, validate table names against a strict regex, and cap result rows. The database template implements list_tables, describe_table, and a constrained query tool with these safeguards.

Inspection fails when the file has import-time side effects that crash, the FastMCP object name does not match the file:object reference, or optional template dependencies like httpx are missing. Fix imports and install dependencies, then rerun fastmcp inspect.

Use FastMCP when creating a new server or wrapping an API, database, or file workflow as MCP tools. If the server already exists and only needs client configuration, use a connection-focused approach instead of building one.