shap-and-model-explainability v1.0.0A

Immutable. This exact content is served forever at /api/v1/blob/1110b6e2c0b2175e.

---
name: shap-and-model-explainability
description: >
  Use this skill for model explainability with SHAP, LIME, and permutation importance, including healthcare-safe interpretation patterns.
version: 1.0.0
authors:
  - Marie-Lynne Block
tags:
  - data-science
  - [TODO]
---

## What this skill does

[TODO] Define the specific workflow this skill standardizes, including default libraries,
quality checks, and expected deliverables.

## When to use it

[TODO] List concrete user intents and trigger phrases that should activate this skill.

## Instructions

1. Clarify the objective, data assumptions, and success metrics.
2. Execute a leakage-safe and reproducible workflow for this skill domain.
3. Validate outputs with diagnostics, edge-case checks, and documented caveats.

## Output format

- A concise plan of action
- Executable code or commands
- Validation summary with assumptions and risks

## Examples

### Example 1 - baseline workflow

**Input:** User asks for help in shap-and-model-explainability.
**Expected output:** A reproducible, validated workflow using the skill's core tools.

## Notes

- Prefer documented, stable APIs over experimental shortcuts.
- Record assumptions explicitly when data quality or labels are uncertain.