***

title: NVIDIA OpenShell Developer Guide
sidebar-title: Home
slug: get-started
description: OpenShell is the safe, private runtime for autonomous AI agents. Run agents in sandboxed environments that protect your data, credentials, and infrastructure.
keywords: Generative AI, Cybersecurity, AI Agents, Sandboxing, Security, Privacy, Inference Routing
position: 1
---------------------

For clean Markdown of any page, append .md to the page URL. For a complete documentation index, see https://docs.nvidia.com/openshell/llms.txt. For full documentation content, see https://docs.nvidia.com/openshell/llms-full.txt.

<BadgeLinks
  badges={[
    {
      href: "https://github.com/NVIDIA/OpenShell",
      src: "https://img.shields.io/badge/github-repo-green?logo=github",
      alt: "GitHub",
    },
    {
      href: "https://github.com/NVIDIA/OpenShell/blob/main/LICENSE",
      src: "https://img.shields.io/badge/License-Apache_2.0-blue",
      alt: "License",
    },
    {
      href: "https://pypi.org/project/openshell/",
      src: "https://img.shields.io/badge/PyPI-openshell-orange?logo=pypi",
      alt: "PyPI",
    },
  ]}
/>

NVIDIA OpenShell is the safe, private runtime for autonomous AI agents. It provides sandboxed execution environments
that protect your data, credentials, and infrastructure. Agents run with exactly the permissions they need and
nothing more, governed by declarative policies that prevent unauthorized file access, data exfiltration, and
uncontrolled network activity.

## Get Started

Install the CLI and create your first sandbox in two commands.

```shell
uv tool install -U openshell
openshell sandbox create --from openclaw
```

Refer to the [Quickstart](/get-started/quickstart) for more details.

***

## Explore

<div>
  <Cards>
    <Card title="About OpenShell" href="/about/overview">
      Learn about OpenShell and its capabilities.

      <Badge intent="tip" minimal outlined>
        Concept
      </Badge>
    </Card>

    <Card title="Quickstart" href="/get-started/quickstart">
      Install the CLI and create your first sandbox in two commands.

      <Badge intent="tip" minimal outlined>
        Tutorial
      </Badge>
    </Card>

    <Card title="Tutorials" href="/tutorials">
      Hands-on walkthroughs from first sandbox to custom policies.

      <Badge intent="tip" minimal outlined>
        Concept
      </Badge>
    </Card>

    <Card title="Gateways and Sandboxes" href="/sandboxes/manage-gateways">
      Deploy gateways, create sandboxes, configure policies, providers, and community images for your AI agents.

      <Badge intent="tip" minimal outlined>
        Concept
      </Badge>
    </Card>

    <Card title="Inference Routing" href="/inference/about">
      Keep inference traffic private by routing API calls to local or self-hosted backends.

      <Badge intent="tip" minimal outlined>
        Concept
      </Badge>
    </Card>

    <Card title="Observability" href="/observability">
      Understand sandbox logs, access them via CLI and TUI, and export OCSF JSON records.

      <Badge intent="tip" minimal outlined>
        How-To
      </Badge>
    </Card>

    <Card title="Reference" href="/reference/default-policy">
      Policy schema, environment variables, and system architecture.

      <Badge intent="tip" minimal outlined>
        Reference
      </Badge>
    </Card>

    <Card title="Security Best Practices" href="/security/best-practices">
      Every configurable security control, its default, and the risk of changing it.

      <Badge intent="tip" minimal outlined>
        Concept
      </Badge>
    </Card>
  </Cards>
</div>

***

<Warning title="Notice and Disclaimer">
  This software automatically retrieves, accesses or interacts with external materials. Those retrieved materials are not distributed with this software and are governed solely by separate terms, conditions and licenses. You are solely responsible for finding, reviewing and complying with all applicable terms, conditions, and licenses, and for verifying the security, integrity and suitability of any retrieved materials for your specific use case. This software is provided "AS IS", without warranty of any kind. The author makes no representations or warranties regarding any retrieved materials, and assumes no liability for any losses, damages, liabilities or legal consequences from your use or inability to use this software or any retrieved materials. Use this software and the retrieved materials at your own risk.
</Warning>