Aaronontheweb/dotnet-skills

What it solves

This project provides a comprehensive library of professional .NET development patterns and expertise to AI coding assistants. It prevents AI from relying on outdated or generic pretraining data by providing a curated set of "skills" and specialized agents that enforce production-grade C# and .NET ecosystem standards.

How it works

The project is structured as a plugin for AI assistants like Claude Code, Codex, GitHub Copilot, and OpenCode. It consists of two primary components:

• Skills Library: A collection of 30 detailed reference documents (SKILL.md files) covering specific technical domains such as Akka.NET, EF Core, .NET Aspire, and modern C# coding standards.
• Specialized Agents: Five AI personas with deep domain expertise in areas like concurrency, performance analysis, and actor systems, which are automatically invoked by the assistant when relevant tasks are detected.

Users install the plugin via the assistant's marketplace or by copying the skills into their project's configuration directories.

Who it’s for

Professional .NET developers who use AI coding assistants and want to ensure the AI generates code that follows modern, high-performance, and battle-tested production patterns rather than generic LLM output.

Highlights

• Broad .NET Coverage: Includes specialized knowledge for Akka.NET, ASP.NET Core, .NET Aspire, and EF Core.
• Production-Grade Principles: Emphasizes immutability, type safety, and performance-aware coding (e.g., using Span<T>).
• Multi-Assistant Support: Compatible with Claude Code, Codex, GitHub Copilot, and OpenCode.
• Quality Gates: Includes tools like dotnet-slopwatch to detect LLM-generated anti-patterns in codebases.