📌Overview

Purpose: The MCP server is designed to manage Docker environments using natural language, making container management more intuitive.

Overview: The MCP server allows users to create and manage Docker containers through simple, conversational commands. It is particularly useful for server administrators and hobbyists who want to efficiently handle containers without needing extensive command-line knowledge.

Key Features:

Natural Language Composition: Users can define container requirements in plain language, and the server generates and executes Docker configurations based on these instructions.
Container Management Tools: The MCP server provides various utilities to list, create, run, and debug containers, images, networks, and volumes, ensuring comprehensive management capabilities.
Interactive Planning: Through a plan+apply mechanism, the server enables users to interactively refine and execute their container configurations, enhancing the user experience and flexibility.

🐋 Docker MCP Server

An MCP server for managing Docker with natural language!

🪩 Features

🚀 Compose containers with natural language
🔍 Introspect & debug running containers
📀 Manage persistent data with Docker volumes

❓ Target Users

Server administrators managing remote Docker engines, e.g., for public-facing websites
Tinkerers running containers locally and experimenting with Docker apps
AI enthusiasts exploring LLM capabilities

Demo

A quick demo showing a WordPress deployment using natural language.

🏎️ Quickstart

Install

Claude Desktop Config Paths

MacOS: ~/Library/Application\ Support/Claude/claude_desktop_config.json
Windows: %APPDATA%/Claude/claude_desktop_config.json

Install from PyPi with uv

If you don't have uv installed, follow the installation instructions on its official website.

Add the following to your MCP servers file:

"mcpServers": {
  "mcp-server-docker": {
    "command": "uvx",
    "args": [
      "mcp-server-docker"
    ]
  }
}

Install with Docker

This server can run inside a Docker container.

Build the Docker image:

docker build -t mcp-server-docker .

Add the following to your MCP servers file:

"mcpServers": {
  "mcp-server-docker": {
    "command": "docker",
    "args": [
      "run",
      "-i",
      "--rm",
      "-v",
      "/var/run/docker.sock:/var/run/docker.sock",
      "mcp-server-docker:latest"
    ]
  }
}

Note: Mounting the Docker socket allows the MCP server to control the local Docker daemon.

📝 Prompts

🎻 `docker_compose`

Compose containers using natural language. Provide a project name and a description of desired containers, and the LLM will handle the rest.

The prompt instructs the LLM to execute a plan+apply loop involving:

You specify containers to bring up.
The LLM presents a concise natural language plan.
You either apply the plan or provide feedback to recalculate it.

Examples

Name: nginx
Containers: deploy an nginx container exposing it on port 9000
Name: wordpress
Containers: deploy a WordPress container and a supporting MySQL container, exposing WordPress on port 9000

Resuming a Project

When starting a new chat, the LLM receives the current status of any containers, volumes, and networks associated with the project name, useful for cleanup.

📔 Resources

The server supports these container resources:

Stats: CPU, memory, etc.
Logs: tail container logs

🔨 Tools

Containers

list_containers
create_container
run_container
recreate_container
start_container
fetch_container_logs
stop_container
remove_container

Images

list_images
pull_image
push_image
build_image
remove_image

Networks

list_networks
create_network
remove_network

Volumes

list_volumes
create_volume
remove_volume

🚧 Disclaimers

Sensitive Data

DO NOT CONFIGURE CONTAINERS WITH SENSITIVE DATA, such as API keys or database passwords. Exchanging sensitive data with the LLM is insecure unless the LLM runs locally.

If you need secure secrets management, please file an issue with your use case.

Reviewing Created Containers

Docker is not a secure sandbox. The MCP server may impact the host machine. Review all containers created by the LLM carefully.

Sensitive Docker options like --privileged or --cap-add/--cap-drop are disabled for safety. Request these features via issues if needed.

🛠️ Configuration

This server uses the Python Docker SDK's from_env method. For details, see the official Docker SDK documentation.

💻 Development

Using Devbox is recommended to configure your development environment.

Example MCP config for Claude MCP with Devbox:

"docker": {
  "command": "/path/to/repo/.devbox/nix/profile/default/bin/uv",
  "args": [
    "--directory",
    "/path/to/repo/",
    "run",
    "mcp-server-docker"
  ]
}

This document provides a clear overview and usage guide for the Docker MCP server with unnecessary details and local links removed.

mcp-server-docker