Installation¶

systems-manager 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 Linux host (apt, dnf, zypper, or pacman) or Windows for the package-management surface; the MCP and agent servers run anywhere Python does.

From PyPI (recommended)¶

pip install systems-manager

Optional extras¶

The base install ships the CLI and the cross-platform package managers. Install the extra for the interface you need:

Extra	Install	Pulls in
`mcp`	`pip install "systems-manager[mcp]"`	FastMCP MCP-server runtime (`agent-utilities[mcp]`)
`agent`	`pip install "systems-manager[agent]"`	Pydantic-AI agent + Logfire tracing
`all`	`pip install "systems-manager[all]"`	The MCP server, the agent, and tracing
`test`	`pip install "systems-manager[test]"`	`pytest`, `pytest-xdist`, `pytest-asyncio`, `pytest-cov`

# Typical: run the MCP server and the agent
pip install "systems-manager[all]"

From source¶

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

With uv:

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

Prebuilt Docker image¶

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

docker pull knucklessg1/systems-manager:latest

docker run --rm -i \
  knucklessg1/systems-manager:latest systems-manager-mcp   # stdio transport (default)

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

Verify the install¶

systems-manager --help
systems-manager-mcp --help
python -c "import systems_manager; print(systems_manager.__version__)"

Next steps¶

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