cuequivariance-torch

Pass

Audited by Gen Agent Trust Hub on Jun 22, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: The skill provides equivariant mathematical operations and PyTorch layers (e.g., BatchNorm, Linear, Tensor Products). All code appears functional and aligned with its stated purpose of scientific and machine learning computation.
  • [COMMAND_EXECUTION]: The skill utilizes PyTorch's internal compilation and optimization features such as torch.compile, torch.jit.script, and torch.jit.trace within its test utilities and core primitives. These are standard deep learning library practices for model optimization and do not constitute arbitrary or malicious command execution.
  • [EXTERNAL_DOWNLOADS]: The skill references and depends on cuequivariance and cuequivariance-ops-torch. These are related vendor packages from the NVIDIA ecosystem required for full functionality and hardware acceleration, particularly on CUDA-enabled GPUs.
Audit Metadata
Risk Level
SAFE
Analyzed
Jun 22, 2026, 01:49 PM
Security Audit — agent-trust-hub — cuequivariance-torch