deepstream-inspect-rtsp

Pass

Audited by Gen Agent Trust Hub on Oct 31, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: The skill is designed for diagnostic purposes, providing triage workflows and command templates to isolate network, camera, and decoder issues in DeepStream environments.
  • [SAFE]: The skill manifest and README include clear instructions for users to redact PII and credentials, such as RTSP URLs and tokens, before providing input to the agent.
  • [SAFE]: Execution boundaries are well-defined. The skill explicitly forbids automatic installation of software, system modifications, or starting packet captures without user intervention.
  • [SAFE]: The allowed-tools configuration is restricted to standard Linux and NVIDIA diagnostic utilities (e.g., nvidia-smi, ip, ethtool, gst-inspect-1.0) used only for read-only system inspection.
  • [SAFE]: No evidence of prompt injection, data exfiltration, obfuscation, or malicious persistence mechanisms was found. The technical content is consistent with official NVIDIA DeepStream documentation and community best practices.
Audit Metadata
Risk Level
SAFE
Analyzed
Oct 31, 2026, 09:41 PM