Overview
What is Containerlab MCP Server?
Containerlab MCP Server is an experimental Model Context Protocol server that enables AI assistants (such as Claude) to interact with containerlab network topologies. It is tailored for use with Claude Desktop and requires a running containerlab API server.
How to use Containerlab MCP Server?
Build the MCP server binary for your platform (Windows, Mac, or Linux) using go build, then configure Claude Desktop by placing the executable and adding a claude_desktop_config.json that sets the API_SERVER_URL environment variable (default http://localhost:8080). Ensure the containerlab API server is running.
Key features of Containerlab MCP Server
- List available containerlab labs
- Deploy new network topologies
- Inspect lab details
- Execute commands on lab nodes
- Destroy and clean up labs
Use cases of Containerlab MCP Server
- Ask Claude to deploy a lab topology using natural language
- Inspect the status of running labs via Claude
- Execute commands on lab nodes through conversational AI
- Destroy and clean up labs with simple prompts
FAQ from Containerlab MCP Server
What dependencies are required?
You need a running containerlab API server and the latest build of containerlab from the main branch. The MCP server is a compiled Go binary.
What transport does the server use?
The server communicates with the containerlab API server via HTTP, using the API_SERVER_URL environment variable (default http://localhost:8080).
Where do lab configurations and data live?
Lab topologies and state are managed by containerlab itself. The MCP server acts only as a proxy to the API server; no lab data is stored by the MCP server.
Is authentication supported?
The README does not mention any authentication mechanism. The server is intended for local use.
Is this ready for production?
No. The README describes this as a "quick and dirty trial" for exploring MCP with containerlab. It is experimental.
