training-machine-learning-models
This skill trains machine learning models using automated workflows. It analyzes datasets, selects appropriate model types (classification, regression, etc.), configures training parameters, trains the model with cross-validation, generates performance metrics, and saves the trained model artifact. Use this skill when the user requests to "train" a model, needs to evaluate a dataset for machine learning purposes, or wants to optimize model performance. The skill supports common frameworks like scikit-learn.
90
49%
Does it follow best practices?
Impact
98%
1.07xAverage score across 9 eval scenarios
Passed
No known issues
Optimize this skill with Tessl
npx tessl skill review --optimize ./backups/skills-migration-20251108-070147/plugins/ai-ml/ml-model-trainer/skills/ml-model-trainer/SKILL.mdQuality
Discovery
92%Based on the skill's description, can an agent find and select it at the right time? Clear, specific descriptions lead to better discovery.
This is a well-crafted skill description that clearly articulates specific capabilities, includes natural trigger terms, and provides explicit guidance on when to use it. Its main weakness is potential overlap with adjacent data science skills (e.g., data analysis, model evaluation, or model deployment), though the focus on training workflows and scikit-learn helps differentiate it somewhat.
| Dimension | Reasoning | Score |
|---|---|---|
Specificity | Lists multiple specific concrete actions: analyzes datasets, selects model types (with examples like classification, regression), configures training parameters, trains with cross-validation, generates performance metrics, and saves the trained model artifact. | 3 / 3 |
Completeness | Clearly answers both 'what' (analyzes datasets, selects models, configures parameters, trains, evaluates, saves artifacts) and 'when' with an explicit 'Use this skill when...' clause covering training requests, dataset evaluation, and model optimization. | 3 / 3 |
Trigger Term Quality | Includes strong natural keywords users would say: 'train a model', 'dataset', 'machine learning', 'classification', 'regression', 'cross-validation', 'performance metrics', 'scikit-learn', 'optimize model performance'. Good coverage of terms a user would naturally use. | 3 / 3 |
Distinctiveness Conflict Risk | While it specifies ML model training, it could overlap with skills for data analysis, data preprocessing, or model deployment. The mention of scikit-learn helps narrow scope, but terms like 'analyzes datasets' and 'evaluate a dataset' could conflict with general data analysis skills. | 2 / 3 |
Total | 11 / 12 Passed |
Implementation
7%Reviews the quality of instructions and guidance provided to agents. Good implementation is clear, handles edge cases, and produces reliable results.
This skill content reads like a marketing description rather than actionable instructions for Claude. It contains no executable code, no specific commands, no concrete implementation details, and explains concepts Claude already understands. The entire file could be replaced with a concise code template showing the actual scikit-learn pipeline with validation steps.
Suggestions
Replace the abstract examples with actual executable Python code using scikit-learn (e.g., a complete pipeline with `train_test_split`, `cross_val_score`, model fitting, `classification_report`, and `joblib.dump`).
Remove the 'Overview', 'How It Works', and 'When to Use' sections entirely — they restate information Claude already knows and add no actionable value.
Add explicit validation checkpoints: check for missing values, verify target variable type, validate train/test split ratios, check for class imbalance, and include a feedback loop for poor metrics.
Add concrete code for model persistence (e.g., `joblib.dump(model, 'model.pkl')`) and specify output format expectations (what metrics to print, what files to save).
| Dimension | Reasoning | Score |
|---|---|---|
Conciseness | The content is verbose and explains concepts Claude already knows (what classification/regression are, what cross-validation is, what accuracy/precision/recall are). The 'Overview' section restates the description, 'How It Works' is generic, and 'When to Use' repeats obvious triggers. Nearly every section could be eliminated or drastically shortened. | 1 / 3 |
Actionability | There is no executable code, no concrete commands, no specific library imports, no actual training pipeline, and no copy-paste ready examples. The examples describe what the skill 'will do' in abstract terms rather than providing actual Python code using scikit-learn or any other framework. This is entirely descriptive rather than instructive. | 1 / 3 |
Workflow Clarity | The workflow is described at a very high level ('analyze data, select model, train, evaluate') with no concrete steps, no validation checkpoints, no error handling, and no feedback loops. There's no guidance on what to do if training fails, data is malformed, or metrics are poor. | 1 / 3 |
Progressive Disclosure | The content is organized into logical sections with clear headers, which provides some structure. However, there are no references to external files, no bundle files to support deeper content, and inline content that is mostly filler rather than substantive material that would benefit from being split out. | 2 / 3 |
Total | 5 / 12 Passed |
Validation
100%Checks the skill against the spec for correct structure and formatting. All validation checks must pass before discovery and implementation can be scored.
Validation — 11 / 11 Passed
Validation for skill structure
No warnings or errors.
- Commit
13d35b8
- Reviewed
Table of Contents
Is this your skill?
If you maintain this skill, you can claim it as your own. Once claimed, you can manage eval scenarios, bundle related skills, attach documentation or rules, and ensure cross-agent compatibility.