# NVIDIA NemoClaw

An open-source reference stack for running AI agents such as OpenClaw, Hermes and LangChain Deep Agents more safely inside NVIDIA OpenShell sandboxes, with managed inference and lifecycle operations.

NVIDIA NemoClaw is an open source reference stack for running supported AI agents more safely inside NVIDIA OpenShell sandboxes. It provides guided onboarding, managed inference, network policy, managed integrations, snapshots and lifecycle operations through the NemoClaw CLI and agent-specific aliases.

It currently supports OpenClaw as its default agent, along with Hermes and LangChain Deep Agents Code. Installation runs through an interactive installer, with an express mode on supported DGX or WSL hosts that installs OpenClaw with recommended settings, or an interactive mode for choosing an agent, sandbox name, inference provider and model. A starter prompt lets an existing coding agent such as Cursor, Claude Code, Codex or Copilot install NemoClaw on the user's behalf, asking one question at a time and running commands only with explicit approval.

It is written in TypeScript and licensed Apache-2.0.

Website: https://docs.nvidia.com/nemoclaw/latest/