The Agent Skills Directory

[SAFE]: The primary implementation in scripts/main.py uses localized decision tree logic and mathematical formulas to provide statistical recommendations. It does not perform any sensitive file system operations, network requests, or use dangerous functions like eval() or exec().
[INDIRECT_PROMPT_INJECTION]: The skill ingests user-provided data characteristics (e.g., number of groups, data distribution types) to generate recommendations. The attack surface is minimal as the script's output is limited to predefined statistical test names and rationales, and the skill does not grant the agent high-privilege capabilities that could be abused via malicious data input.
[EXTERNAL_DOWNLOADS]: The documentation in references/ mentions and provides snippets for standard scientific libraries such as NumPy, SciPy, Statsmodels, and Matplotlib. These are well-known, trusted industry-standard packages for statistical computing.
[CREDENTIALS_SAFE]: A review of all files, including the provided audit log statistical-analysis-advisor_audit_result_v2.json, confirmed that no hardcoded credentials, API keys, or secrets are present. The skill adheres to safe secret management practices.

statistical-analysis-advisor