Docker MCP Server

Docker container and Compose stack management for AI-driven infrastructure workflows.

Data last scanned today · Reviewed today

284 tools501PythonMIT

Overview

This server provides direct control over local Docker environments, enabling AI assistants to list containers, monitor statuses, and retrieve system logs. It bridges the gap between natural language commands and container infrastructure, facilitating hands-on management of isolated service instances. Beyond basic container inspection, it supports lifecycle actions such as creating new containers and deploying complete Docker Compose stacks. By utilizing UV for dependency management, the server provides a streamlined interface for developers aiming to automate their containerized testing or deployment tasks within AI environments.

Our verdict

With a security score of 28/100, this tool is categorized as high risk. The repository has not been updated in over 633 days and remains a community-maintained project rather than an official vendor release. Given that critical security aspects such as read-only mode and authentication methods have not been editorially reviewed, this server should be used with extreme caution, ideally in sandboxed or local-only environments where container access control is managed externally.

  • Supports deployment of full Docker Compose stacks
  • Requires zero direct dependencies
  • Licensed under the permissive MIT license
  • High security risk due to lack of editorial review
  • Abandoned codebase with no commits in 633 days
  • Lacks documented verification for read-only modes

Tools

ToolDescriptionRisk
create-containerCreates a standalone Docker containermedium
deploy-composeDeploys a Docker Compose stackmedium
get-logsRetrieves logs from a specific containerlow
list-containersLists all Docker containerslow

Compatibility

ClientLocalDockerRemoteRead-only
Claude Desktop
Cursor
VS Code
Windsurf
ChatGPT

Frequently asked questions

What software prerequisites are required to use the Docker MCP server?

You must have Python 3.12+, the UV package manager, a running Docker environment (Desktop or Engine), and the Claude Desktop application installed.

Are there any specific resource management or networking features currently missing?

The server does not support container resource limits, network management, volume management, or container health checks.

How can I debug or troubleshoot the server?

You can launch the MCP Inspector by running the command 'npx @modelcontextprotocol/inspector uv --directory <path-to-docker-mcp> run docker-mcp' to access a web-based debugging interface.

Can I configure container restart policies through this MCP server?

No, container restart policies are currently listed as a known limitation of the server.

How do I install the server for production use in Claude Desktop?

You can add the server to your Claude Desktop configuration file using the 'uvx' command, or by using the Smithery CLI to install it automatically.

Badge

Docker MCP Server security score, rated on RepoAI

Maintain this server? Add the live badge to your README.