# LLM Sandbox MCP server

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

## Links
- Registry page: https://www.getdrio.com/mcp/io-github-vndee-llm-sandbox
- Repository: https://github.com/vndee/llm-sandbox
- Website: https://vndee.github.io/llm-sandbox/

## Install
- Command: `uvx llm-sandbox`
- Auth: Not captured

## Setup notes
- Package: Pypi llm-sandbox v0.3.43
- Environment variable: BACKEND (default docker; choices docker, podman, kubernetes)
- Environment variable: DOCKER_HOST
- Environment variable: KUBECONFIG
- Environment variable: NAMESPACE (default default)
- Environment variable: COMMIT_CONTAINER (default true)
- Environment variable: KEEP_TEMPLATE (default true)
- Environment variable: SANDBOX_NETWORK_MODE
- Environment variable: SANDBOX_READ_ONLY
- Environment variable: SANDBOX_CAP_DROP
- Environment variable: SANDBOX_SECURITY_OPT
- Environment variable: SANDBOX_MEMORY
- Environment variable: SANDBOX_CPUS

## Tools
Not captured

## Resources
Not captured

## Prompts
Not captured

## Metadata
- Owner: io.github.vndee
- Version: 0.3.43
- Runtime: Pypi
- Transports: STDIO
- License: Not captured
- Language: Not captured
- Stars: Not captured
- Updated: Aug 3, 2026
- Source: https://registry.modelcontextprotocol.io