The Agent Skills Directory

[SAFE]: The skill serves as a comprehensive guide for ML development, promoting robust engineering practices such as temporal validation splits, reproducibility via seeding, and detailed logging.
[EXTERNAL_DOWNLOADS]: The skill references standard, well-known libraries in the machine learning ecosystem including torch, catboost, polars, numpy, coloredlogs, fire, and pyyaml. These are established tools typically sourced from official package registries.
[COMMAND_EXECUTION]: Code snippets demonstrate the use of the fire library to wrap Python functions into a command-line interface. This is a standard and safe practice for orchestrating machine learning workflows.
[DATA_EXPOSURE]: The guidelines explicitly recommend using yaml.safe_load() for configuration parsing, which is a security best practice to prevent unsafe deserialization attacks.

ml-project