nemotron-lightning35
Pass
Audited by Gen Agent Trust Hub on Oct 28, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill serves as a reference desk for technical information regarding the Nemotron 3.5 Lightning model. It contains no executable code, shell commands, or network-active components. All external references target official NVIDIA repositories on Hugging Face (e.g.,
nvidia/NVIDIA-Nemotron-3.5-Lightning-30B-A3B-Base-BF16) or trusted container registries (nvcr.io/nvidia/nemo). - [SAFE]: The instructional content in SKILL.md focuses on model identity, architecture, and evaluation metrics. It includes clear boundaries for the AI agent, such as distinguishing between different model versions (Lightning vs. Super 3.5) and warning against using outdated documentation. No prompt injection or behavior-override attempts were found.
- [SAFE]: The skill documents known technical caveats, such as typos in the original technical report and conflicts between model card descriptions and paper findings, which improves the accuracy and safety of the information provided to the user.
- [SAFE]: No obfuscation (Base64, homoglyphs, or hidden characters), persistence mechanisms, or unauthorized privilege escalation patterns were detected across the analyzed files.
Audit Metadata