LLM Sandbox
Securely run LLM-generated code in isolated containers across 7 languages and 3 container backends.
Open the source (opens github.com) https://github.com/vndee/llm-sandbox
Before you use it
- Runs on your computer with your permissions. Read what it can do before connecting it to an agent.
What it is
MCP servers from io.github.vndee.
The problem it solves and who it suits haven't been reviewed by InnerVault. The description above is the source's own.
How to get it
pypi package (stdio)
llm-sandbox@0.3.43
From the server's entry in the official MCP Registry
Settings it asks for
BACKEND· Container backend to use. Must match the installed extra: mcp-docker, mcp-podman, or mcp-k8s.DOCKER_HOST· Docker or Podman socket URL, e.g. unix:///var/run/docker.sockKUBECONFIG· Path to kubeconfig file when BACKEND=kubernetes.NAMESPACE· Kubernetes namespace used for sandbox pods when BACKEND=kubernetes.COMMIT_CONTAINER· Commit the container after a run so installed libraries persist between sessions.KEEP_TEMPLATE· Keep the base image after the session ends to avoid re-pulling it on the next run.SANDBOX_NETWORK_MODE· Network mode for the sandbox container. Set to 'none' for hardened isolation. Docker and Podman backends only.SANDBOX_READ_ONLY· Mount the sandbox root filesystem read-only. Recommended: true. Docker and Podman backends only.SANDBOX_CAP_DROP· Comma-separated Linux capabilities to drop. Recommended: ALL. Docker and Podman backends only.SANDBOX_SECURITY_OPT· Comma-separated container security options, e.g. no-new-privileges. Docker and Podman backends only.SANDBOX_MEMORY· Memory limit for the sandbox container, e.g. 4g. Docker and Podman backends only.SANDBOX_CPUS· Fractional CPU allocation for the sandbox container, e.g. 1.5. Docker and Podman backends only.
Related
-
Containarium
MCP serversOpen-source agent runtime — SSH-native isolation, eBPF egress policy, Kubernetes + LXC backends, GPU passthrough, MCP-native CLI
git clone https://github.com/FootprintAI/Containarium.git -
LitterBox
MCP serversA self-hosted sandbox for red teams to test payloads against modern detection before deployment. MCP integration lets an LLM agent drive analysis end to end.
git clone https://github.com/BlackSnufkin/LitterBox.git -
wirken
MCP serversThe enterprise gateway for autonomous agents. Identity management, per-channel isolation, credential vault, per-session tamper-evident audit log.
git clone https://github.com/gebruder/wirken.git -
OpenRath
MCP serversAn open-source, PyTorch-like runtime for dynamic multi-agent and multi-session workflows.
git clone https://github.com/Rath-Team/OpenRath.git -
agentenv-framework
MCP serversCreating realistic RL environments requires collaboration between researchers, engineers, and domain experts across many dimensions: artifacts, environments tools, dynamism of the environment, reproducibility, and more. There is no open source framework for building these environments effectively. Until now.
git clone https://github.com/scaleapi/agentenv-framework.git -
console
MCP serversOpen-source, self-hosted AI gateway and control plane: OpenAI-compatible LLM routing, RAG, an MCP hub, guardrails and agent tracing for multi-tenant teams.
git clone https://github.com/Cognipeer/console.git