shap

Pass

Audited by Gen Agent Trust Hub on Mar 31, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: The skill provides legitimate technical documentation and workflows for the SHAP (SHapley Additive exPlanations) library.
  • Evidence: Found throughout SKILL.md and the references/ directory, detailing model explanation, debugging, and visualization techniques.
  • [EXTERNAL_DOWNLOADS]: Includes standard installation instructions for the shap package and its dependencies from official package registries.
  • Evidence: uv pip install shap in SKILL.md.
  • [DYNAMIC_EXECUTION]: Describes standard practices for loading serialized machine learning models and explainer objects using industry-standard libraries.
  • Evidence: References to joblib.load(), tf.keras.models.load_model(), and torch.load() are included in references/workflows.md as part of production deployment examples.
Audit Metadata
Risk Level
SAFE
Analyzed
Mar 31, 2026, 08:21 AM
Security Audit — agent-trust-hub — shap