Python MCP server

Super simple implementation of reference python MCP for GPT-OSS built-in tools.

More information can be found in GPT-OSS repository

The model was trained to use a python tool to perform calculations and other actions as part of its chain-of-thought. During the training the model used a stateful tool which makes running tools between CoT loops easier. This reference implementation, however, uses a stateless mode.

Warning

This implementation runs in a permissive Docker container, which could be problematic in cases like prompt injections. It's serving as an example and you should consider implementing your own container restrictions in production.

Requirements:

• Docker must be installed, and the current user must have permission to run it.
• The image python:3.11 will be pulled automatically if missing.

Installing and running MCP

To generate binary run ./prepare_binary.sh, it should work on Linux.

To build binary, you have to install box-project/box to generate PHAR.

Thanks to amazing projects like Static PHP and FrankenPHP we are able to run PHP applications as a single binary now.

The easiest way is to just download binary from releases for your platform.

MCP config:

STDIO is only supported transport for now, just add entry to mcp.json with a path to binary

{
    "command": "./dist/python-mcp",
    "args": [],
    "env": {
        "APP_LOG_DIR": "/tmp/.symfony/python-mcp/log"
    }
}

You can also use python-mcp.phar PHAR file.

If you want to use other transports use some wrapper for now, for example, MCPO

uvx mcpo --port 8000 -- ~/mcp/dist/python-mcp

Env variables

Underneath it's just Symfony CLI application, so typical ENV variables can be used, but not recommended for running as binary.

### Set log level, default INFO, with log action level ERROR
LOG_LEVEL=info
# Where to store logs
APP_LOG_DIR="/tmp/mcp/python-mcp/log"

Development

If you need to modify or want to run/debug a server locally, you should:

• git clone repository
• run composer install
• ./bin/python-mcp contains server, while ./bin/console holds Symfony console

To debug server you should use npx @modelcontextprotocol/inspector

npx @modelcontextprotocol/inspector ./bin/python-mcp
# for compiled version with log dir
npx @modelcontextprotocol/inspector -e APP_LOG_DIR="${HOME}/.symfony/python-mcp/log" ./dist/python-mcp