pytorch-patterns
Pass
Audited by Gen Agent Trust Hub on Apr 6, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill contains educational content on PyTorch best practices with no malicious patterns detected. All code snippets are pedagogical and represent idiomatic use of the library.
- [SAFE]: The documentation explicitly promotes secure coding practices, such as advising the use of
weights_only=Trueintorch.load()to mitigate risks associated with unsafe deserialization of pickle data.
Audit Metadata