The Agent Skills Directory

[SAFE]: The skill outlines a structured approach to technology selection using official and well-known .NET libraries from Microsoft and the community.- [SAFE]: It explicitly warns against anti-patterns like hardcoding API keys in source control and recommends secure alternatives such as Azure Key Vault and environment variables.- [SAFE]: The guidance includes critical safety controls for agentic workflows, such as mandatory iteration limits and token budget ceilings to prevent runaway costs or execution.- [SAFE]: It promotes deterministic and reproducible machine learning practices by recommending fixed random seeds and proper data splitting techniques.- [SAFE]: External references and suggested NuGet packages originate from trusted organizations (Microsoft, GitHub) or well-known open-source projects (OllamaSharp, MathNet).

technology-selection