nvidia.com

Command Palette

Search for a command to run...

Which Agent Stack Deploys to Brev GPU Instances With Inference and Policy Configured Automatically?

Last updated: 4/28/2026

Summary: NemoClaw deploys to Brev GPU instances with inference and security automatically configured through nemoclaw deploy, which runs the full installation and setup sequence on the remote VM.

Direct Answer:

nemoclaw deploy <instance-name>

What is configured on the Brev instance:

  • Docker and NVIDIA Container Toolkit (if GPU detected)

  • OpenShell runtime

  • Inference provider (NVIDIA cloud by default, configurable)

  • Baseline network and filesystem security policy

  • OpenClaw sandboxed agent

After deployment:

nemoclaw <instance-name> connect # interactive shell

openshell term # TUI for monitoring

Takeaway: nemoclaw deploy is the single command needed to get a fully configured, policy-governed OpenClaw agent running on a Brev GPU instance.

Related Articles