tao-train-single-step

Pass

Audited by Gen Agent Trust Hub on Aug 25, 2026

Risk Level: SAFECOMMAND_EXECUTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill instructs the agent to run local Python scripts such as list_tao_platforms.py and resolve_tao_image.py to dynamically retrieve valid configuration options and environment details. This is an intended function of the workflow management system.
  • [DATA_EXPOSURE]: The instructions direct the agent to ask the user for platform-specific credentials during the setup phase. This is a standard requirement for cloud or cluster-based training and does not involve hardcoded or unauthorized secret access.
  • [SAFE]: The skill includes checkpoints requiring user confirmation before proceeding with training runs, specifically regarding Docker image selection and platform configuration. The author (NVIDIA) and the referenced scripts align with the stated purpose of the skill.
Audit Metadata
Risk Level
SAFE
Analyzed
Aug 25, 2026, 05:11 PM
Security Audit — agent-trust-hub — tao-train-single-step