Model Explainability Skill for Single-Row SHAP Values
Explain one synthetic tree-model prediction and verify that the feature contributions match the model output scale.
中文Install in Claude Code
Use Python, a tree model trained on synthetic data, and a compatible SHAP release. The example predicts packing time.
- Get the complete `plugins/ai-ml/model-explainability-tool/skills/explaining-machine-learning-models` directory from the pinned GitHub commit.
- Copy it into the target project's `.claude/skills/`. Check `.claude/skills/explaining-machine-learning-models/SKILL.md`, start Claude Code at the project root, and call `/explaining-machine-learning-models`.
Example task
Explain one held-out order with item_count and parcel_size. Generate a contribution table and check the base value and contributions on the configured model scale.What to check
The report records the prediction, both contributions, and the output scale. It can be reproduced from the same model and row.
Original SKILL.md on GitHub · Updated
Frequently asked question
Does a large SHAP value prove a causal effect of changing that feature?
It describes the model's prediction attribution. A causal claim needs separate evidence.