LLM Sandbox

Level CMCP ecosystemNot an MCP ServerLocal

Securely run LLM-generated code in isolated containers across 7 languages and 3 container backends.

Overview

LLM Sandbox securely runs LLM-generated code in isolated environments through a unified API, supporting Docker, Kubernetes, Podman, Bubblewrap, NsJail, and Windows MXC backends. It executes Python, JavaScript, Java, C++, Go, and R code with resource limits, network isolation, security policies, dependency installation, file transfer, and artifact extraction, and also offers MCP integration.

⭐ 0Development ToolsRepo ↗Updated 2026-08-03

Installation

MCP client config (paste into mcp.json)

{
  "mcpServers": {
    "llm-sandbox": {
      "command": "python3",
      "args": [
        "-m",
        "llm_sandbox.mcp_server.server"
      ]
    }
  }
}

Agent install prompt

Copy and paste into Cursor / Claude / other agents to install this MCP for you

Please read this MCP repository and install it into my current environment:
https://github.com/vndee/llm-sandbox

Known working client config (can write into mcp.json):
{
  "mcpServers": {
    "llm-sandbox": {
      "command": "python3",
      "args": [
        "-m",
        "llm_sandbox.mcp_server.server"
      ]
    }
  }
}

Please: 1) check README for deps and env vars; 2) write config for my client; 3) if secrets are missing, tell me what to fill — do not invent values.

Related servers