shap

Pass

Audited by Gen Agent Trust Hub on Oct 23, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: The skill provides comprehensive documentation and a reference script for machine learning model explanation using the SHAP library.
  • [SAFE]: All identified dependencies, including shap, scikit-learn, numpy, pandas, and matplotlib, are well-known, industry-standard libraries for data science.
  • [SAFE]: The skill explicitly includes security best practices, such as warning users against loading untrusted pickle or model artifacts which could lead to arbitrary code execution.
  • [SAFE]: The bundled script scripts/tabular_report.py follows a deterministic workflow for training a model on a built-in dataset and generating reports, with no evidence of malicious behavior or unsafe data handling.
  • [SAFE]: Installation instructions use uv, a standard tool for Python package management, to install the specific version 0.52.0 of the SHAP library.
Audit Metadata
Risk Level
SAFE
Analyzed
Oct 23, 2026, 03:12 AM