Docker MCP

🚀 Docker MCP Service

The Docker MCP service is a powerful model context protocol server designed to execute code within Docker containers and return the results to large language models (LLMs), such as Claude. This service supports a wide range of features, including isolated code execution, multi - language support, and integration with various package management tools.

🚀 Quick Start

The Docker MCP service is a powerful model context protocol server. It can execute code in Docker containers and return the results to LLMs like Claude. It supports features such as isolated code execution, multi - language support, and integration with package management tools.

✨ Features

• Isolated Code Execution: Ensures code runs in an independent environment through Docker containers.
• Multi - Language Support: Supports multiple programming languages, including Python and Node.js.
• Complex Script Support: Capable of handling complex scripts and tasks.
• Package Management: Built - in support for pip, npm, apt - get, and apk, automatically detects and uses the appropriate package management tool.
• Container Management: Provides functions for creating, monitoring, and cleaning up Docker containers.
• Robust Error Handling: Equipped with timeout detection and recovery mechanisms to ensure reliable task execution.
• Color Output: Enhances the user experience through terminal color coding.

📦 Installation

Installation Requirements

• Python Version: Python 3.6 or higher must be installed.
• Docker Installation: Ensure Docker is correctly installed and running (verify with the docker --version command).
• fastmcp Library: Install the fastmcp library to support integration with LLMs.

Installation Steps

1. Clone the code repository:

   git clone https://github.com/your-repository.git
   

2. Create and activate a virtual environment:

   python -m venv venv
   source venv/bin/activate  # On Linux/Mac
   venv\Scripts\activate     # On Windows
   

3. Install dependencies:

   pip install -r requirements.txt
   

💻 Usage Examples

Running the Server

Start the MCP service:

python src/docker_mcp.py

Access http://localhost:8090 to open the console bound to the MCP.

Toolset

List Containers

View all running Docker containers:

list_containers [--all]

• --all: Show all containers, including stopped ones.

Create a New Container

Create and start a new Docker container:

create_container [-i IMAGE] [-- privileged|nonPrivileged] [COMMAND]

• IMAGE: Specify the image, default is python:3.8 - slim.
• privileged or nonPrivileged: Set the container's permission mode.
• COMMAND: Specify the command to execute.

Execute a Command

Execute a command in an existing container:

execute_command [-c CONTAINER_ID] [-t COMMAND]

• CONTAINER_ID: Specify the ID of the target container.
• COMMAND: The specific command to execute.

Remove a Container

Remove a specified Docker container:

remove_container [-c CONTAINER_ID] [--volumes]

• CONTAINER_ID: Specify the ID of the container to remove.
• --volumes: Also remove the volumes associated with the container.

Basic Workflow Example

1. List all containers:

   list_containers
   

2. Create a Python container and execute a command:

   create_container -i python:3.8-slim -c "echo 'Hello, Docker!'" --nonPrivileged
   

3. After execution, remove the container:

   remove_container [-c CONTAINER_ID]
   

Python Data Analysis Example

1. Create a Node.js container and install dependencies:

   create_container -i node:16 --privileged npm install express axios
   

2. Execute a script:

   execute_command [-c CONTAINER_ID] node script.js
   

📚 Documentation

Package Manager Support

The service automatically detects and uses the appropriate package management tool:

• Python: Uses pip.
• Node.js: Uses npm.
• Debian/Ubuntu: Uses apt - get.
• Alpine: Uses apk.

In cases where the package manager cannot be inferred from the image name, the system will attempt to detect available package management tools.

Integration with LLMs

Supports integration with LLMs like Claude through the fastmcp library for more advanced functionality.

⚠️ Important Note

• Permission Issues: Running Docker commands may require administrative privileges.
• Resource Limitations: Be aware of the CPU and memory usage of containers to avoid over - using system resources.
• Error Handling: Ensure all possible exceptions, such as network interruptions or missing images, are handled.

🤝 Contribution Guidelines

Welcome to contribute to the project! Please refer to the project's contribution guidelines for more information.