v1.0.0 to v1.0.0
18 added, 64 removed. Audit A to A.
---
name: evaluating-machine-learning-models
description: |
- Build this skill allows AI assistant to evaluate machine learning models using a comprehensive suite of metrics. it should be used when the user requests model performance analysis, validation, or testing. AI assistant can use this skill to assess model accuracy, p... Use when appropriate context detected. Trigger with relevant phrases based on skill purpose.
+ Evaluate trained machine learning models with the right metrics and comparison logic.
+ Use for benchmark review, threshold selection, calibration, validation, and model comparison; not for feature engineering or leakage auditing.
allowed-tools: Read, Write, Edit, Grep, Glob, Bash(cmd:*)
version: 1.0.0
author: Jeremy Longshore <jeremy@intentsolutions.io>
license: MIT
---
# Model Evaluation Suite
- This skill provides automated assistance for model evaluation suite tasks.
+ Use this skill when the model exists and the question is whether it is good enough.
## Overview
- This skill empowers Claude to perform thorough evaluations of machine learning models, providing detailed performance insights. It leverages the `model-evaluation-suite` plugin to generate a range of metrics, enabling informed decisions about model selection and optimization.
-
- ## How It Works
-
- 1. **Analyzing Context**: Claude analyzes the user's request to identify the model to be evaluated and any specific metrics of interest.
- 2. **Executing Evaluation**: Claude uses the `/eval-model` command to initiate the model evaluation process within the `model-evaluation-suite` plugin.
- 3. **Presenting Results**: Claude presents the generated metrics and insights to the user, highlighting key performance indicators and potential areas for improvement.
+ This skill focuses on choosing and interpreting the right evaluation metrics for the problem, then comparing candidate models or thresholds.
## When to Use This Skill
- This skill activates when you need to:
- - Assess the performance of a machine learning model.
- - Compare the performance of multiple models.
- - Identify areas where a model can be improved.
- - Validate a model's performance before deployment.
-
- ## Examples
-
- ### Example 1: Evaluating Model Accuracy
-
- User request: "Evaluate the accuracy of my image classification model."
-
- The skill will:
- 1. Invoke the `/eval-model` command.
- 2. Analyze the model's performance on a held-out dataset.
- 3. Report the accuracy score and other relevant metrics.
-
- ### Example 2: Comparing Model Performance
-
- User request: "Compare the F1-score of model A and model B."
-
- The skill will:
- 1. Invoke the `/eval-model` command for both models.
- 2. Extract the F1-score from the evaluation results.
- 3. Present a comparison of the F1-scores for model A and model B.
-
- ## Best Practices
-
- - **Specify Metrics**: Clearly define the specific metrics of interest for the evaluation.
- - **Data Validation**: Ensure the data used for evaluation is representative of the real-world data the model will encounter.
- - **Interpret Results**: Provide context and interpretation of the evaluation results to facilitate informed decision-making.
-
- ## Integration
-
- This skill integrates seamlessly with the `model-evaluation-suite` plugin, providing a comprehensive solution for model evaluation within the Claude Code environment. It can be combined with other skills to build automated machine learning workflows.
-
- ## Prerequisites
-
- - Appropriate file access permissions
- - Required dependencies installed
-
- ## Instructions
+ - Comparing candidate models with consistent metrics
+ - Reviewing precision/recall/F1/AUC, regression error, calibration, or ranking quality
+ - Stress-testing validation strategy before deployment or publication
- 1. Invoke this skill when the trigger conditions are met
- 2. Provide necessary context and parameters
- 3. Review the generated output
- 4. Apply modifications as needed
+ ## Not For / Boundaries
- ## Output
+ - Building the training pipeline itself: use `training-machine-learning-models`
+ - Engineering features: use `engineering-features-for-machine-learning`
+ - Checking train/test contamination: use `ml-data-leakage-guard`
- The skill produces structured output relevant to the task.
+ ## Typical Outputs
- ## Error Handling
+ - Metric suite recommendations
+ - Model comparison tables
+ - Notes on threshold tradeoffs, calibration, and validation weaknesses
- - Invalid input: Prompts for correction
- - Missing dependencies: Lists required components
- - Permission errors: Suggests remediation steps
+ ## Related Skills
- ## Resources
+ - `confusion-matrix-generator` for class-level error breakdowns
+ - `scientific-reporting` when the evaluation must become a deliverable
- - Project documentation
- - Related skills and commands