Use ML.NET to train, evaluate, or integrate machine-learning models into .NET applications with realistic data preparation, inference, and deployment expectations. USE FOR: ML.NET integration; local model training or retraining; inference pipelines, model loading, evaluation, and deployment review. DO NOT USE FOR: unrelated stacks; generic tasks that do not need this specific guidance. INVOKES: inspect the repository context, edit targeted files, and run relevant build, test, lint, or validation
SkillScore91Maintained
Automatic
◆AI Engineer pack
Last verified 8/19/2026
What it changes
Passed 20 of 22 test cases with the skill, and 19 of 22 without it.
Without skill86%
With skill91%
Checks it passed that the model alone missed
The recommendation requires inspecting and validating the resulting pipeline architecture prior to production deployment.
1 fixed19 passed either way2 still failing0 broke
Pass-rate lift
+5 pp
Tokens
-4%
Agent turns
0%
Response time
-16%
Model gemini-3.6-flashMeasured 8/19/2026v120 of 22 cases passed with skill
Tokens, turns and time describe resource use, not a dollar price.
What's inside
The instructions your model reads when the skill loads.
SKILL.md33 lines3 files
ML.NET
Trigger On
integrating machine learning into a .NET application
training or retraining ML.NET models from local data
reviewing inference pipelines, model loading, or AutoML-generated code
Workflow
Start from the prediction task and data quality, not the algorithm or package list.
Separate training code from inference code so the production path stays lean and predictable.
Review feature engineering, normalization, label quality, and evaluation metrics before trusting model output.
Use Model Builder or the ML.NET CLI when they speed up exploration, but inspect the generated C# before treating it as production architecture.
Plan how the model is loaded, versioned, and refreshed in the application lifecycle.
Validate with representative datasets and explicit evaluation, not only with a sample that happens to run.
Deliver
ML.NET pipelines that fit the prediction task
production-usable inference integration
evaluation evidence tied to the business scenario
Validate
model quality is measured, not assumed
training and inference responsibilities are separated
deployment and versioning expectations are explicit
References
patterns.md - Data loading, training pipelines, evaluation metrics, deployment strategies, and feature engineering patterns
examples.md - Complete examples for sentiment analysis, price prediction, image classification, anomaly detection, recommendations, clustering, fraud detection, text classification, object detection, and AutoML