Install OrbStack MCP Server

Pick your client, copy the command, done.

claude mcp add orbstack-mcp -- orbstack-mcp

Available Tools (6)

Once configured, OrbStack MCP Server gives your AI agent access to:

orbstack_statusCheck the current running status of OrbStack.

orbstack_machine_listList all available Linux machines.

orbstack_docker_psList all running Docker containers.

orbstack_docker_runRun a new Docker container.

orbstack_compose_upStart a Docker Compose project.

orbstack_k8s_statusCheck the status of the Kubernetes cluster.

After setup, try these prompts with your AI agent:

→List all my currently running Docker containers and tell me which ones are using the most resources.

→Create a new Linux machine named 'dev-server' and start it.

→Stop the Docker Compose project in the current directory.

→Check the status of my Kubernetes cluster and restart it if it is not running.

→Pull the latest nginx image and run it as a container.

Prerequisites & system requirements

An MCP-compatible client (Claude Code, Cursor, Windsurf, Claude Desktop, or Codex)
Python 3.8+ with pip installed

Alternative installation methods

claude mcp add orbstack -- python -m orbstack_mcp.server

Save the docs, env vars, and workflow around OrbStack MCP Server in Conare so Claude Code, Codex, and Cursor remember it next time.