ai-for-science-ankh

Pass

Audited by Gen Agent Trust Hub on Apr 18, 2026

Risk Level: SAFE
Full Analysis
  • [COMMAND_EXECUTION]: The skill provides instructions for setting up the Ascend NPU environment by sourcing shell scripts (set_env.sh) and running system utilities (npu-smi). This is standard operational procedure for Huawei's CANN toolkit.
  • [EXTERNAL_DOWNLOADS]: The skill mentions package installation via pip and model weight loading via HuggingFace's from_pretrained method. These utilize well-known services and are focused on retrieving legitimate AI research dependencies and model architectures.
  • [DYNAMIC_EXECUTION]: The migration process utilizes torch_npu.contrib.transfer_to_npu, which is a standard library feature for dynamically redirecting CUDA-based PyTorch operations to NPU-compatible operations. This is the intended and documented way to perform device migration without manual code rewrites.
  • [DATA_EXPOSURE]: The skill references local paths for model weights and uses environment variables (e.g., ANKH_BASE_PATH) to locate them. It does not access sensitive user files or credentials outside the scope of the model migration task.
Audit Metadata
Risk Level
SAFE
Analyzed
Apr 18, 2026, 03:03 AM
Security Audit — agent-trust-hub — ai-for-science-ankh