What problem does it solve? Choosing the wrong AI or ML technology in a .NET project leads to wasted cost, non-deterministic behavior, and unmaintainable code. This Skill provides a decision tree and guardrails for selecting between ML.NET, Microsoft.Extensions.AI, Microsoft Agent Framework, ONNX Runtime, OllamaSharp, and the GitHub Copilot SDK. ## Core Features & Use Cases - Decision-tree technology selection: Classifies tasks (classification, regression, RAG, agentic workflows, local inference) and maps them to the correct .NET AI/ML library. - Layered library guidance: Enforces the MEAI abstraction → provider SDK → Agent Framework orchestration stack so layers are never skipped or mixed. - Production guardrails: Covers reproducibility seeds, retry logic, token budgets, iteration limits, RAG chunking thresholds, and anti-patterns to reject. - Use Case: When asked to add support-ticket classification to an ASP.NET Core app, the Skill directs you to ML.NET with a seeded MLContext and train/test split instead of an expensive, non-deterministic LLM call. ## Quick Start Ask the agent to recommend and implement the right .NET AI or ML technology for your task, such as adding a RAG pipeline or a classification model to your project.
