The Agent Skills Directory

[SAFE]: The skill's primary function is to guide the agent through statistical workflows including hypothesis testing, regression, and Bayesian modeling. All operations are local and consistent with the stated purpose.
[COMMAND_EXECUTION]: Technical review of scripts/assumption_checks.py shows it is limited to performing mathematical operations and generating data visualizations using standard libraries. There are no calls to subprocess, os.system, or other dangerous execution functions.
[EXTERNAL_DOWNLOADS]: The skill mentions common scientific packages (e.g., scipy, statsmodels, pingouin, pymc) but does not include any automated installation scripts or remote code fetching mechanisms.
[DATA_EXFILTRATION]: There are no network operations, credential identifiers, or access patterns targeting sensitive system files. All data processing is confined to the provided datasets in memory.
[INDIRECT_PROMPT_INJECTION]: The skill processes structured data (DataFrames) for analysis. While it lacks explicit sanitization for column names, its capabilities are restricted to statistical testing and plotting, which does not present a significant path for privilege escalation or malicious instruction execution. Ingestion points are localized in scripts/assumption_checks.py, and the capability inventory is limited to standard data science tools.

statistical-analysis