vastai-reference-architecture
Pass
Audited by Gen Agent Trust Hub on Sep 13, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill provides structural guidance and configuration templates for Vast.ai GPU orchestration without any malicious instructions.
- [SAFE]: Includes a Python implementation for a checkpoint manager using the
boto3library to interact with S3. The implementation follows best practices by using variable placeholders for bucket names and prefixes rather than hardcoded credentials. - [SAFE]: Referenced external links target official documentation for Vast.ai and PyTorch.
Audit Metadata