The Agent Skills Directory

[SAFE]: The skill provides instructional content and security best practices for Machine Learning (ML) engineering, including guidance on avoiding unsafe model deserialization and managing data privacy.
[SAFE]: Audit checklists utilize standard, benign search utilities (grep, find, ls) intended for identifying configuration patterns in local development environments.
[SAFE]: The skill explicitly warns against security risks such as arbitrary code execution via pickle and provides recommendations for using safer formats like safetensors or ONNX.
[SAFE]: External references point to authoritative and trusted sources, including Google Research, NIST, and MITRE ATLAS documentation.

sota-ml-engineering