tao-train-mask-auto-encoder
Pass
Audited by Gen Agent Trust Hub on Oct 29, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill uses official NVIDIA container images from the NVIDIA Container Registry (
nvcr.io) for training and deployment actions. This is consistent with the skill's stated purpose and originates from a trusted vendor source. - [SAFE]: Integration with Weights & Biases (
wandb) is provided for experiment tracking. W&B is a well-known and standard service for machine learning workflows. - [SAFE]: Command execution is restricted to the TAO Toolkit CLI tools (
mae) within defined Docker containers. No arbitrary or malicious shell commands were found. - [SAFE]: No hardcoded credentials, malicious obfuscation, or persistence mechanisms were detected in the skill's files or metadata.
Audit Metadata