Creating Tools in FastMCP

Tools are the core building blocks that allow your LLM to interact with external systems, execute code, and access data.

Basic Tool Definition

Creating a tool is as simple as decorating a Python function:

from fastmcp import FastMCP

mcp = FastMCP(name="CalculatorServer")

@mcp.tool
def add(a: int, b: int) -> int:
    """Adds two integer numbers together."""
    return a + b

FastMCP automatically: - Uses the function name as the tool name - Uses the docstring as the tool description - Generates an input schema from type annotations - Handles parameter validation and error reporting

Decorator Arguments

@mcp.tool(
    name="find_products",
    description="Search the product catalog",
    tags={"catalog", "search"},
    meta={"version": "1.2"}
)
def search_products(query: str, category: str | None = None) -> list[dict]:
    return [{"id": 2, "name": "Product"}]

Type Annotations

Supported types: - Basic: int, float, str, bool - Collections: list[str], dict[str, int], set[int] - Optional: str | None, Optional[str] - Union: str | int - Literal: Literal["A", "B"] - Pydantic models for complex data

Async Support

@mcp.tool
async def fetch_data(url: str) -> dict:
    """Fetch data asynchronously."""
    async with httpx.AsyncClient() as client:
        response = await client.get(url)
        return response.json()

Use async def for I/O-bound operations to keep server responsive.