technology-selection
Pass
Audited by Gen Agent Trust Hub on May 8, 2026
Risk Level: SAFE
Full Analysis
- Security-First Architecture: The skill emphasizes secure development practices by explicitly instructing developers to avoid hardcoding API keys and instead use secure configuration providers like Azure Key Vault or environment variables.
- Orchestration Guardrails: For agentic workflows, the skill mandates the use of iteration limits and token budgets to prevent resource exhaustion or runaway processes, which is a standard safety measure for autonomous agents.
- Input Validation: It provides guidance on implementing schema validation and content assertions for LLM outputs, reducing the risk of processing malformed or malicious data from non-deterministic models.
- Trusted Dependency Management: The skill references official Microsoft and established community NuGet packages (e.g., Microsoft.Extensions.AI, OllamaSharp). These are standard libraries for the .NET ecosystem and are used according to their intended functional scope.
- Deterministic Fallbacks: There is a clear focus on using deterministic technologies like ML.NET for structured data tasks, which reduces the attack surface associated with the non-deterministic nature of large language models.
Audit Metadata