tao-train-visual-changenet
Pass
Audited by Gen Agent Trust Hub on Sep 22, 2026
Risk Level: SAFEEXTERNAL_DOWNLOADS
Full Analysis
- [EXTERNAL_DOWNLOADS]: The skill downloads a pretrained model backbone from HuggingFace (nvidia/C-RADIOv2-B) using a provided staging script (
scripts/stage_backbone.py). This download targets a public repository by the same vendor as the skill author. - [SAFE]: All included scripts perform legitimate model management and data validation tasks. The skill uses standard vendor infrastructure for model checkpoints and Docker registries. No indicators of prompt injection, data exfiltration, or malicious command execution were found.
Audit Metadata