Skills
skills/nvidia/skills/tao-analyze-gaps-vlm-bcq/Gen Agent Trust Hub

tao-analyze-gaps-vlm-bcq

Pass

Audited by Gen Agent Trust Hub on Jun 23, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: The skill serves a documented, benign purpose of identifying prediction gaps (false positives and false negatives) in VLM workflows. No obfuscation, data exfiltration, or persistence mechanisms are present.
  • [COMMAND_EXECUTION]: The skill facilitates the execution of the gap_analysis CLI tool, which is a component of the NVIDIA TAO Toolkit. The execution occurs within a containerized environment (tao_toolkit.data_services) and is limited to processing user-specified local files.
  • [DATA_EXPOSURE]: The skill processes prediction data provided in a local JSON format and generates reports locally. There are no network operations or calls to external domains observed.
Audit Metadata
Risk Level
SAFE
Analyzed
Jun 23, 2026, 12:09 PM
Security Audit — agent-trust-hub — tao-analyze-gaps-vlm-bcq