The Agent Skills Directory

Standard Development Workflow: The skill provides templates for Python scripts used in model verification. These scripts use well-known libraries like torch, diffusers, and safetensors to compare model outputs.
Safe Data Handling: Instructions explicitly recommend persistent storage over temporary directories and advise against modifying global caches, which aligns with best practices for development environments.
No External Risks Detected: No unauthorized network operations, credential harvesting, or obfuscation techniques were found. The skill operates within the expected scope of ML research and integration.

testing-parity