> ## Documentation Index
> Fetch the complete documentation index at: https://mintlify.com/NVIDIA/NemoClaw/llms.txt
> Use this file to discover all available pages before exploring further.

# NVIDIA NemoClaw

> Run OpenClaw AI agents safely inside NVIDIA OpenShell — with sandboxed execution, cloud inference, and declarative network policy.

<CardGroup cols={2}>
  <Card title="Quickstart" icon="rocket" href="/get-started/quickstart">
    Install NemoClaw and run your first sandboxed OpenClaw agent in minutes.
  </Card>

  <Card title="How It Works" icon="gear" href="/about/how-it-works">
    Understand the plugin, blueprint, and sandbox lifecycle.
  </Card>

  <Card title="CLI Commands" icon="terminal" href="/reference/commands">
    Full reference for all `nemoclaw` and `openclaw nemoclaw` commands.
  </Card>

  <Card title="Architecture" icon="diagram-project" href="/reference/architecture">
    Plugin structure, blueprint lifecycle, and sandbox environment.
  </Card>
</CardGroup>

## What is NemoClaw?

NemoClaw is the [OpenClaw](https://openclaw.ai) plugin for [NVIDIA OpenShell](https://github.com/NVIDIA/OpenShell). It moves OpenClaw into a sandboxed environment where every network request, file access, and inference call is governed by declarative policy.

<Note>
  NemoClaw is alpha software. Interfaces, APIs, and behavior may change without notice. It is shared to gather feedback and enable early experimentation — it is not yet production-ready.
</Note>

## Key capabilities

<CardGroup cols={2}>
  <Card title="Sandboxed execution" icon="shield-halved">
    Every agent runs inside an OpenShell sandbox with Landlock, seccomp, and network namespace isolation. No access is granted by default.
  </Card>

  <Card title="NVIDIA cloud inference" icon="microchip">
    Agent traffic routes through cloud-hosted Nemotron 3 Super 120B via [build.nvidia.com](https://build.nvidia.com), transparent to the agent.
  </Card>

  <Card title="Declarative network policy" icon="network-wired">
    Egress rules are defined in YAML. Unknown hosts are blocked and surfaced to the operator for approval.
  </Card>

  <Card title="Blueprint lifecycle" icon="box-archive">
    Versioned blueprints handle sandbox creation, digest verification, and reproducible setup.
  </Card>

  <Card title="Single CLI" icon="square-terminal">
    The `nemoclaw` command orchestrates the full stack: gateway, sandbox, inference provider, and network policy.
  </Card>

  <Card title="Remote GPU deployment" icon="server">
    Deploy a sandboxed agent to a remote GPU instance via Brev for persistent operation.
  </Card>
</CardGroup>

## How it fits together

| Component     | Role                                                                                       |
| ------------- | ------------------------------------------------------------------------------------------ |
| **Plugin**    | TypeScript CLI commands for launch, connect, status, and logs.                             |
| **Blueprint** | Versioned Python artifact that orchestrates sandbox creation, policy, and inference setup. |
| **Sandbox**   | Isolated OpenShell container running OpenClaw with policy-enforced egress and filesystem.  |
| **Inference** | NVIDIA cloud model calls, routed through the OpenShell gateway, transparent to the agent.  |

## Get started

```bash theme={null}
curl -fsSL https://nvidia.com/nemoclaw.sh | bash
```

The installer sets up Node.js if needed, then runs the guided onboard wizard to create a sandbox, configure inference, and apply security policies.

After setup, connect to your sandbox:

```bash theme={null}
nemoclaw my-assistant connect
```

See the [Quickstart](/get-started/quickstart) for the full walkthrough.

## License

NemoClaw is released under the [Apache License 2.0](https://github.com/NVIDIA/NemoClaw/blob/main/LICENSE).