nemoclaw-user-guide

Pass

Audited by Gen Agent Trust Hub on Jun 24, 2026

Risk Level: SAFEEXTERNAL_DOWNLOADSREMOTE_CODE_EXECUTIONCOMMAND_EXECUTION
Full Analysis
  • [EXTERNAL_DOWNLOADS]: The skill instructs the agent to fetch documentation index files and Markdown pages from NVIDIA's official domain (docs.nvidia.com). This is a standard operation for documentation retrieval and relies on the official infrastructure of the skill's author.
  • [REMOTE_CODE_EXECUTION]: The skill guides the user to configure a Model Context Protocol (MCP) server located at https://docs.nvidia.com/nemoclaw/_mcp/server. This is a legitimate mechanism for extending the agent's capabilities using tools provided by the vendor.
  • [COMMAND_EXECUTION]: The instructions permit the agent to run commands for non-technical users, but this is explicitly conditioned on the agent first explaining the command's purpose and obtaining the user's informed consent.
  • [SAFE]: The skill contains specific safety protocols to protect user data, including directives for the agent to stop before requesting credentials (API keys, tokens) and instructions to use redacted placeholders in examples rather than asking for actual secrets.
Audit Metadata
Risk Level
SAFE
Analyzed
Jun 24, 2026, 09:06 PM
Security Audit — agent-trust-hub — nemoclaw-user-guide