The Agent Skills Directory

[SAFE]: The skill outlines standard software engineering practices for MLOps, such as static typing, linting, and automated testing using well-known tools like ruff, pytest, and pydantic.
[SAFE]: Includes a dedicated security section that recommends using bandit to detect vulnerabilities like hardcoded secrets or unsafe function calls (e.g., eval).
[SAFE]: Promotes good data handling practices by suggesting pandera for schema validation and pydantic for runtime data modeling.
[SAFE]: Explicitly instructs users to never log secrets and to sanitize outputs, reducing the risk of accidental data exposure.

mlops-validation