llm-training
Pass
Audited by Gen Agent Trust Hub on Sep 23, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill functions as a documentation and reference guide for large language model (LLM) training frameworks and memory optimization techniques. It contains no executable code or scripts.
- [SAFE]: All external resource links target official documentation or public repositories of trusted organizations and well-known services within the machine learning community, such as HuggingFace and Microsoft DeepSpeed.
- [SAFE]: No patterns of prompt injection, data exfiltration, or obfuscation were identified within the skill body or metadata.
Audit Metadata