The Agent Skills Directory

[SAFE]: The skill is entirely instructional, providing a framework for machine learning engineering workflows. It does not perform any file operations, network requests, or command executions.
[SAFE]: Included Python code snippets in the 'Core Workflow' section are benign examples for generating unique hashes using standard libraries (hashlib) and checking metric thresholds. These serve as illustrative templates rather than active functional code.
[SAFE]: The skill actively encourages security-positive practices, such as explicitly instructing the agent to check for PII (Personally Identifiable Information) in logs and prompts, and recommending security scans for model artifacts and dependencies.

mle-workflow