What happened?
When a tool parameter is annotated as str, passing a string value that happens to contain valid JSON causes validation to fail. Instead of arriving as a Python string, the value is treated as a dict/list, and Pydantic rejects it with a string-type validation error. I expected a string argument to remain a string regardless of its contents.
If the bug reproduces, the call should fail with a validation error indicating content was treated as a dict instead of a str.
Example Code
import asyncio
import json
from fastmcp import FastMCP, Client
mcp = FastMCP("repro")
@mcp.tool()
def echo_type(content: str) -> str:
return f"type={type(content).__name__} value={content}"
async def demo():
payload = json.dumps({
"account_full_name": "Test Data",
"as_of_date": "2026-04-13",
})
async with Client(mcp) as client:
result = await client.call_tool("echo_type", {"content": payload})
print(result)
if __name__ == "__main__":
asyncio.run(demo())Version Information
fastmcp: 2.10.5
python: 3.13.2 (tags/v3.13.2:4f8bb39, Feb 4 2025, 15:23:48) [MSC v.1942 64 bit (AMD64)]
os: Windows-11-10.0.22631-SP0
What happened?
When a tool parameter is annotated as str, passing a string value that happens to contain valid JSON causes validation to fail. Instead of arriving as a Python string, the value is treated as a dict/list, and Pydantic rejects it with a string-type validation error. I expected a string argument to remain a string regardless of its contents.
If the bug reproduces, the call should fail with a validation error indicating content was treated as a dict instead of a str.
Example Code
Version Information