pydantic-mcp
- other
Model Context Protocol tool calling support for Pydantic AI
About
What is pydantic-mcp?
pydantic-mcp is a Python package that enables Model Context Protocol (MCP) tool calling support for Pydantic AI. It converts an MCP client session into a list of pydantic_ai.Tool instances for use with Pydantic AI agents. The package is now superseded by built-in MCP support in PydanticAI.
How to use pydantic-mcp?
Create an mcp.ClientSession and call tools = await mcptools(session) to get a list of pydantic_ai.Tool instances for the supported MCP tools. The package is runnable as a demo against Groq’s llama-3.1-8b-instant model using an environment variable GROQ_API_KEY.
Key features of pydantic-mcp
- Converts MCP tools to Pydantic AI Tool instances
- Works with any MCP-compliant server (e.g., filesystem)
- Simple one-call integration: mcptools(session)
- Provides a ready-to-run demo using Groq LLM
Use cases of pydantic-mcp
- Connect an LLM agent to MCP‑enabled tools (e.g., file system, databases)
- Read and summarize files via a secure MCP filesystem server
- Leverage Pydantic AI’s tool‑calling with any MCP server
FAQ from pydantic-mcp
Is pydantic-mcp still maintained?
No. PydanticAI now includes native support for MCP servers. Users should migrate to ai.pydantic.dev/mcp/ instead.
What models does the demo work with?
The demo is shown with Groq’s llama-3.1-8b-instant model; any model supported by Pydantic AI can be used with the Tool instances obtained.
How do I install pydantic-mcp?
The README does not include explicit installation instructions, but it uses uv run tests/demo.py and assumes a GROQ_API_KEY.
Does it support any MCP server?
Yes. The package converts any mcp.ClientSession — the demo uses a secure filesystem server, but any MCP server can be used.
Is there a pricing or license?
The package is distributed under the MIT License (as indicated in the demo summary of the LICENSE file).
Details
- Author
- rectalogic
- GitHub stars
- 9
- Category
- other
- Repository
- rectalogic/pydantic-mcp
pydantic-mcp
PydanticAI now includes support for MCP servers, use that instead
Model Context Protocol tool calling support for Pydantic AI.
Just create an mcp.ClientSession and call tools = await mcptools(session)
to get a list of pydantic_ai.Tool instances for the supported tools.
Example:
https://github.com/rectalogic/pydantic-mcp/blob/98c4e9abf31837cb48ebeb3eae6cb55b16c841e8/tests/demo.py#L15-L25
Demo
You can run the demo against Groq llama-3.1-8b-instant:
``sh-session``
$ export GROQ_API_KEY=xxx
$ uv run tests/demo.py "Read and summarize the file ./LICENSE"
Secure MCP Filesystem Server running on stdio
Allowed directories: [ '/users/aw/projects/rectalogic/pydantic-mcp' ]
The file ./LICENSE is a MIT License agreement. It states that the software is provided "as is" without warranty and that the authors and copyright holders are not liable for any claims, damages, or other liability arising from the software or its use.
