tao-train-ocrnet
Pass
Audited by Gen Agent Trust Hub on Jun 23, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill provides a structured set of schemas and instructions for managing the lifecycle of OCRNet models within the NVIDIA TAO Toolkit ecosystem.
- [SAFE]: All remote references, such as container images from
nvcr.io, target well-known and trusted infrastructure associated with the skill's author. - [SAFE]: The skill contains no embedded executable scripts, relying on the agent's use of standard CLI tools in a containerized environment to perform documented tasks.
- [SAFE]: Analysis of the 28 files, including JSON schemas and specification templates, reveals no evidence of prompt injection, data exfiltration, or obfuscated malicious content.
Audit Metadata