tao-train-bevfusion

Pass

Audited by Gen Agent Trust Hub on Jun 30, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: The skill is a technical resource for managing BEVFusion models within the NVIDIA TAO Toolkit ecosystem. Analysis did not find any indicators of malicious intent, prompt injection, or unauthorized data access.
  • [COMMAND_EXECUTION]: The skill facilitates model operations through defined CLI subtasks such as bevfusion train, bevfusion evaluate, and torchrun. These commands are executed within a specific, vendor-provided TAO container image (nvcr.io/nvidia/tao/tao-toolkit:5.5.0-pyt).
  • [EXTERNAL_DOWNLOADS]: References the nvcr.io registry for containerized execution environments. This is a well-known and trusted vendor service, and the use of these resources is standard operational behavior for this technology stack.
Audit Metadata
Risk Level
SAFE
Analyzed
Jun 30, 2026, 05:48 PM
Security Audit — agent-trust-hub — tao-train-bevfusion