nemoclaw-user-monitor-sandbox

Inspect NemoClaw sandbox health and trace agent behavior via status probes and log streaming.

22.1k|3.0k|Updated Mar 15, 2026
One-click install
npx skills add https://github.com/NVIDIA/NemoClaw --skill nemoclaw-user-monitor-sandbox
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: nemoclaw-user-monitor-sandbox
Source: https://github.com/NVIDIA/NemoClaw/tree/main/.agents/skills/nemoclaw-user-monitor-sandbox
Command: npx skills add https://github.com/NVIDIA/NemoClaw --skill nemoclaw-user-monitor-sandbox

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

Provides operators a concise, repeatable way to inspect NemoClaw sandbox health, trace agent behavior, and diagnose failures so issues can be identified and resolved quickly.

Core Features & Use Cases

  • Sandbox Health Checks: Run host-side status probes to verify sandbox state, gateway connectivity, and inference backend reachability.
  • Log Streaming & Tracing: Stream blueprint runner and sandbox logs to surface errors and trace agent execution for debugging.
  • TUI Network Monitoring: Use the OpenShell terminal UI to view active connections, blocked egress requests, and routed inference activity for operator approval workflows.
  • Inference Testing: Execute simple inference requests and validation steps to confirm provider responsiveness and routing configuration.
  • Use Case: When an assistant returns errors or times out, use this Skill to check status, follow logs, inspect network activity, and run a test inference to isolate the failure.

Quick Start

Run the host command nemoclaw <name> status to check sandbox state, probe inference health, and confirm gateway connectivity.

Frequently Asked Questions about nemoclaw-user-monitor-sandbox

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I monitor sandbox health and trace agent behavior in NemoClaw?

You can monitor sandbox health and trace agent behavior by running the host command nemoclaw <name> status to check sandbox state, probe inference health, and confirm gateway connectivity.

How do I stream logs to debug an OpenShell-managed agent that returns errors?

To debug an OpenShell-managed agent returning errors, stream blueprint runner and sandbox logs to surface errors and trace agent execution for troubleshooting.

Can I view blocked egress requests and routed inference activity in the terminal UI?

Yes, you can use the OpenShell terminal UI to view active connections, blocked egress requests, and routed inference activity for operator approval workflows.

How do I test inference routing to confirm provider responsiveness in my sandbox?

You can test inference routing by executing simple inference requests and validation steps to confirm provider responsiveness and verify routing configuration.

Do I need a running NemoClaw sandbox to use the OpenShell CLI for diagnostics?

Yes, you need a running NemoClaw sandbox with the OpenShell CLI on PATH to execute status probes, stream logs, inspect TUI network activity, and test inference routing.

Why does my assistant time out and how can I isolate the failure in NemoClaw?

When an assistant times out, check sandbox status, follow logs, inspect network activity in the TUI, and run a test inference to isolate the failure using OpenShell diagnostics.