The Agent Skills Directory

[SAFE]: The skill implements a rigorous data science workflow using bundled Python scripts and markdown-based guidance. It emphasizes data integrity and validation as core principles.
[COMMAND_EXECUTION]: The skill includes tests/test_scripts.py which uses subprocess.run to execute its own bundled scripts (profile_data.py, baseline_model.py) for validation. This is a standard and safe use of subprocess for internal testing.
[EXTERNAL_DOWNLOADS]: The skill README provides installation instructions using npx from the author's repository and suggests installing common, well-known Python packages such as pandas, scikit-learn, and pyarrow via standard package managers.
[INDIRECT_PROMPT_INJECTION]: The skill is designed to process external datasets (CSV, Parquet, Excel). While this is an ingestion surface, the skill utilizes specific boundary rules and mandates the use of analytical scripts that process data statistically, minimizing the risk of prompt injection from cell content.

data-scientist