The Agent Skills Directory

[SAFE]: The skill provides comprehensive instructions for implementing the Shiny Mediator library in .NET applications. It correctly identifies and explains critical library requirements, such as the partial class requirement for middleware-decorated handlers to support source generation.
[SAFE]: The skill instructions emphasize security and best practices for modern .NET development, including AOT (Ahead-of-Time) and trimming compatibility, use of IMediatorContext for scoped operation chaining, and proper CancellationToken propagation.
[SAFE]: Data processing instructions focus on standard serialization practices using System.Text.Json and provide clear guidance on maintaining AOT safety via JsonSerializerContext.
[SAFE]: The AI Tools integration instructions properly utilize the Microsoft.Extensions.AI library for exposing mediator contracts as AI-callable tools with structured JSON schemas, following standard vendor integration patterns.
[SAFE]: Network-related features (HTTP clients, OpenAPI scaffolding, and Server-Sent Events) follow established .NET patterns using IHttpClientFactory and standard ASP.NET Core routing, with no evidence of unexpected network behavior or data exfiltration.
[SAFE]: The deterministic static detector flagged a potential prompt injection/concealment pattern, but analysis confirms this is a false positive triggered by documentation regarding how logging and configuration can be optionally skipped or suppressed at the application level during library setup, which is standard framework behavior.

shiny-mediator