skills/ascend-ai-coding/awesome-ascend-skills/external-cannbot-model-model-infer-graph-mode/Gen Agent Trust Hub
external-cannbot-model-model-infer-graph-mode
Pass
Audited by Gen Agent Trust Hub on Jul 7, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill provides legitimate technical guidance for optimizing machine learning models on Ascend NPUs using standard industry practices and official vendor libraries like torch_npu and torchair.- [SAFE]: External links point to the official documentation of TorchAir on the GitCode platform, which is consistent with the skill's purpose and the Ascend hardware ecosystem.- [SAFE]: No suspicious patterns such as prompt injection, data exfiltration, credential exposure, or obfuscation were detected. The skill instructions focus entirely on model performance and graph mode adaptation.
Audit Metadata