skills/ascend-ai-coding/awesome-ascend-skills/migration-ascend-torchnpu-skills-torch-npu-reference/Gen Agent Trust Hub
migration-ascend-torchnpu-skills-torch-npu-reference
Pass
Audited by Gen Agent Trust Hub on Oct 22, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill provides instructions for installing software and referencing documentation. All external resources, including wheel downloads from
pytorch.organd source code fromgitcode.com/Ascend, are official or vendor-controlled. The use of established Chinese package mirrors (Alibaba, Tsinghua) forpipis standard practice and does not pose a security risk. No patterns of obfuscation, exfiltration, or unauthorized command execution were found.
Audit Metadata