Installation¶

container-manager-mcp is a standard Python package and a prebuilt container image. Pick the path that matches how you want to run it.

Requirements¶

Python 3.11 – 3.14.
A reachable Docker or Podman engine on the controller host (or on the remote hosts you target — see Multi-Host).

From PyPI (recommended)¶

pip install container-manager-mcp

Optional extras¶

The base install pulls in the Docker engine client. Install the extra for what you need:

Extra	Install	Pulls in
`docker`	`pip install "container-manager-mcp[docker]"`	Docker engine client (`docker`)
`podman`	`pip install "container-manager-mcp[podman]"`	Podman engine client (`podman`)
`mcp`	`pip install "container-manager-mcp[mcp]"`	FastMCP MCP-server runtime (`agent-utilities[mcp]`)
`agent`	`pip install "container-manager-mcp[agent]"`	Pydantic-AI agent + Logfire tracing
`all`	`pip install "container-manager-mcp[all]"`	Everything above

# Typical: run the MCP server with both engines and the agent
pip install "container-manager-mcp[all]"

From source¶

git clone https://github.com/Knuckles-Team/container-manager-mcp.git
cd container-manager-mcp
pip install -e ".[all]"          # editable install with every extra

With uv:

uv pip install -e ".[all]"
uv run container-manager-mcp

Prebuilt Docker image¶

A multi-stage, slim image is published on every release (installs container-manager-mcp[all], entrypoint container-manager-mcp):

docker pull knucklessg1/container-manager-mcp:latest

docker run --rm -i \
  -v /var/run/docker.sock:/var/run/docker.sock \
  knucklessg1/container-manager-mcp:latest        # stdio transport (default)

For an HTTP server with a published port and the agent server, see Deployment.

Verify the install¶

container-manager-mcp --help
python -c "import container_manager_mcp; print(container_manager_mcp.__version__)"

Next steps¶

Deployment — run it as a long-lived MCP / agent server behind Caddy + DNS.
Usage — call the tools, the DockerManager API, and the CLI.
Configuration — every environment variable.