nemo-automodel-model-onboarding

Pass

Audited by Gen Agent Trust Hub on Jun 30, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: The skill provides documentation and implementation patterns for model onboarding into the NeMo AutoModel framework.
  • [PROMPT_INJECTION]: No attempts to bypass safety filters or override system instructions were found.
  • [DATA_EXFILTRATION]: No credential exposure or unauthorized data transmission patterns detected.
  • [EXTERNAL_DOWNLOADS]: The skill mentions fetching model configuration files from the Hugging Face Hub, which is a standard operation for machine learning model development and architecture discovery.
  • [REMOTE_CODE_EXECUTION]: The skill contains code templates for model implementations but does not include any commands to download and execute arbitrary remote scripts.
Audit Metadata
Risk Level
SAFE
Analyzed
Jun 30, 2026, 10:58 AM
Security Audit — agent-trust-hub — nemo-automodel-model-onboarding