The Agent Skills Directory

[EXTERNAL_DOWNLOADS]: The skill requires standard, well-known Python packages for data science including xgboost, scikit-learn, shap, pandas, numpy, matplotlib, marimo, and ibis-framework. These dependencies are resolved through standard package registries (PyPI).
[COMMAND_EXECUTION]: The documentation suggests the use of the marimo edit --sandbox demo.py command to run the interactive demo. This is a standard and safe method for executing the provided local script within a sandboxed environment.
[INDIRECT_PROMPT_INJECTION]: The skill is designed to process external tabular data from files such as Parquet or CSV. While this represents a data ingestion surface, it is a primary and necessary function for a binary classification tool. The instructions follow standard data science workflows for materializing and cleaning data before model training.
[SAFE]: No malicious patterns, obfuscation, persistence mechanisms, or unauthorized privilege escalation attempts were detected. The skill uses modern frameworks like Ibis and Marimo appropriately and safely.

binary-classification