Happy slice of mcp pizzaMCP.pizza
Fire in da houseTop Tip:Paying $100+ per month for Perplexity, MidJourney, Runway, ChatGPT and other tools is crazy - get all your AI tools in one site starting at $15 per month with Galaxy AI Fire in da houseCheck it out free

arrakis-mcp-server

MCP.Pizza Chef: abshkbh

The Arrakis MCP Server is a Model Context Protocol server that integrates the Arrakis VM sandbox environment with large language models. It enables LLMs to interact with and control virtual machine sandboxes securely, providing resources to list and manage VMs and tools to start sandboxes or restore snapshots. This server facilitates safe, isolated execution environments accessible via MCP, enhancing LLM capabilities with sandboxed VM operations.

Use This MCP server To

Expose Arrakis VM sandbox controls to LLMs for secure code execution List and retrieve detailed information about available virtual machines Start new VM sandboxes dynamically from LLM commands Restore VM snapshots to previous states via MCP tool calls Integrate sandboxed VM environments into AI-enhanced workflows Enable isolated testing and execution environments for AI-generated code

README

Arrakis MCP Server

A Model Context Protocol (MCP) server that exposes Arrakis VM sandbox functionality to Large Language Models.

Setup

Setup

  1. Clone this repository:

    git clone https://github.com/abshkbh/arrakis-mcp-server.git
cd arrakis-mcp-server

  2. Configure your LLM tool to use the Arrakis MCP server by adding the following to your MCP configuration:

    "mcpServers": {
    "arrakis": {
        "command": "<path-to-uv>",
        "args": [
            "--directory",
            "<path-to-repo>",
            "run",
            "arrakis_mcp_server.py"
        ]
    }
}

API

The Arrakis MCP Server exposes the following MCP resources and tools:

Resources

  • arrakis://vms - List all available VMs
  • arrakis://vm/{vm_name} - Get information about a specific VM

Tools

  • start_sandbox - Start a new VM sandbox
  • restore_snapshot - Restore a VM from a snapshot
  • snapshot - Create a snapshot of a VM
  • run_command - Run a command in a VM
  • upload_file - Upload a file to a VM
  • download_file - Download a file from a VM
  • destroy_vm - Destroy a specific VM
  • destroy_all_vms - Destroy all VMs
  • update_vm_state - Update the state of a VM (pause/stop)

License

This project is licensed under the GNU Affero General Public License v3.0 - see the LICENSE file for details.

arrakis-mcp-server FAQ

How do I configure the Arrakis MCP Server with my LLM tool?
Clone the repository and add the server command and arguments to your MCP configuration under 'mcpServers' as shown in the setup instructions.
What resources does the Arrakis MCP Server expose?
It exposes 'arrakis://vms' to list all VMs and 'arrakis://vm/{vm_name}' to get specific VM information.
What tools are available through the Arrakis MCP Server?
Tools include 'start_sandbox' to start a new VM sandbox and 'restore_snapshot' to restore a VM from a snapshot.
Can the Arrakis MCP Server be used with different LLM providers?
Yes, it is provider-agnostic and can work with OpenAI, Anthropic Claude, and Google Gemini models.
Is the Arrakis MCP Server suitable for running untrusted code?
Yes, it uses VM sandboxing to isolate execution, enhancing security for running untrusted or experimental code.
How do I start a new VM sandbox using the Arrakis MCP Server?
Use the 'start_sandbox' tool exposed by the server via MCP commands to initiate a new sandbox environment.
Can I restore a VM to a previous state?
Yes, the 'restore_snapshot' tool allows restoring a VM from a saved snapshot for state rollback.
What programming languages or environments does the Arrakis VM support?
The server exposes the Arrakis VM sandbox, which supports multiple environments depending on the VM configuration.