Custom Docker Build Fails: FastMCP + Pydantic Conflict (“cannot specify both default and default_factory”)

[baxteross]

Hi,
I’m trying to build a custom Agent Zero Docker image from my Github fork using:
FROM agent0ai/agent-zero-base:latest

The build succeeds, but the container fails to start due to a runtime error inside FastMCP:

File "/a0/python/helpers/mcp_server.py", line 6, in
from fastmcp import FastMCP
...
File "/opt/venv-a0/lib/python3.12/site-packages/pydantic/fields.py", line 252, in init
raise TypeError('cannot specify both default and default_factory')

TypeError: cannot specify both default and default_factory

Has anyone else experienced this problem?
Thanks

[pUrGe12]

You probably have the wrong version of pydantic installed. They are using fastmcp version 2.3.4 which is not the latest one so that supports pydantic2.11.7. I think this issue was fixed with the latest version of fastmcp but that will break other code inside mcp_server.py for example (the create_sse_app implementation). Maybe try downgrading to pydantic==2.11.7 manually.

[relder251]

Try this:

Root Cause:

This error happens when a Pydantic model field is declared using both default= and default_factory= parameters at the same time, e.g.:

my_field: set = Field(default=set(), default_factory=set)
This was ignored or handled leniently in Pydantic 1.x, but in Pydantic 2.x it's a hard error.

Why it happens:

Either your own project, or a dependency (such as an older/unpatched version of FastMCP or another library), defines a model with BOTH default and default_factory.
With Pydantic 2.x, this is now forbidden.
This happens frequently if an ecosystem package hasn't updated yet for Pydantic 2.x compatibility!
How to find & fix:

Search for both parameters used together:

grep -r 'default=' . | grep 'default_factory'
Fix any line that has both – keep only one.

Dependency versions:

If a third-party library (like FastMCP) causes the error, check their docs/issues for a Pydantic 2.x compatibility release.
If they're not compatible yet, either:
Pin Pydantic to v1.x using your requirements.txt or pip:
pip install 'pydantic<2.0.0'
or Upgrade the offending library to a recent release that supports Pydantic 2.x.