moe-training
Pass
Audited by Gen Agent Trust Hub on Feb 17, 2026
Risk Level: SAFE
Full Analysis
- EXTERNAL_DOWNLOADS (SAFE): The skill references standard machine learning libraries such as vLLM and DeepSpeed. It includes instructions to clone the Megatron-DeepSpeed repository from a trusted organization (microsoft), which is classified as a low-risk external reference.
- COMMAND_EXECUTION (SAFE): The provided shell commands are routine for setting up AI development environments, installing dependencies via pip, and launching distributed training sessions. No suspicious or unauthorized command patterns were detected.
- DATA_EXPOSURE (SAFE): The documentation does not access sensitive system paths, hardcode credentials, or perform any unauthorized data exfiltration. All URLs point to reputable research (arXiv) or official documentation and code repositories.
Audit Metadata