tao-train-mask2former
Pass
Audited by Gen Agent Trust Hub on Oct 29, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill instructions and associated files were analyzed for security risks including prompt injection, data exfiltration, and obfuscation. No malicious patterns or security vulnerabilities were identified.
- [EXTERNAL_DOWNLOADS]: The skill references pretrained model weights hosted on GitHub (SwinTransformer repository), which is a standard practice for machine learning workflows and originates from a well-known service.
- [COMMAND_EXECUTION]: Provides templates and instructions for executing the
mask2formerCLI tool within Docker containers. This is standard functionality for the intended purpose of the skill.
Audit Metadata