agentskills.codes
CO

code-review

by jason-kerneySource

>

Install

mkdir -p .claude/skills/code-review-jason-kerney && curl -L -o skill.zip "https://agentskills.codes/api/skills/download/15082" && unzip -o skill.zip -d .claude/skills/code-review-jason-kerney && rm skill.zip

Installs to .claude/skills/code-review-jason-kerney

Activation

This is the description your AI agent reads to decide when to run this skill — the better it matches your request, the more reliably it fires.

Technical code review guidance for WorkMood and similar .NET/MAUI projects. Covers correctness, testability, performance, idiomatic C#/.NET, and API design. Provides checklists, review heuristics, and verification steps for safe, incremental improvements.
255 charsno explicit “when” triggerlonger than Claude Code's old 250-char listing cap (fine on current versions)

About this skill

Code Review Skill for WorkMood

Purpose

This skill guides technical code review for C#/.NET (including MAUI) projects, with a focus on:

  • Correctness and safety (does it work, is it robust?)
  • Testability (can it be covered by unit/integration tests?)
  • Performance (allocations, CPU, UI responsiveness)
  • Idiomatic .NET/C# and project conventions (MVVM, DI, SOLID)
  • API and architecture clarity

Review Checklist

  • Correctness: Does the code do what it claims? Are edge cases handled?
  • Tests: Are there tests? Do they cover happy/sad paths? Are they clear and minimal?
  • Performance: Any obvious allocation, CPU, or UI responsiveness issues? (Use BenchmarkDotNet, PerfView, VS Diagnostics)
  • Readability: Is the code clear, small, and idiomatic? Any long methods, magic values, or code smells?
  • API Design: Are interfaces clear, minimal, and DI-friendly? Any leaky abstractions?
  • MVVM/Architecture: Does it follow project patterns (MVVM, DI, shims, factories)?
  • Documentation: Are comments, XML docs, and codexes up to date?

Review Heuristics

  • Prefer small, focused diffs and single-responsibility changes.
  • Write or update tests before/after changes to guarantee behavior.
  • Use Arrange-Act-Assert in tests; prefer xUnit patterns.
  • For performance: measure before/after, optimize only when needed.
  • For refactoring: preserve behavior, commit in small increments, update codexes.

Example Review Prompts

  • "/code-review [file.cs] Review for testability and performance."
  • "/code-review [method] Is this idiomatic C#? Any code smells?"
  • "/code-review [ViewModel] Does this follow MVVM and DI patterns?"
  • "/code-review [diff] Are there any risks or missing tests?"

Verification Steps

  • Build: dotnet build --framework net9.0-windows10.0.19041.0
  • Test: dotnet test (or filter for specific tests)
  • Benchmark: Add BenchmarkDotNet, run microbenchmarks for hot paths
  • Profile: Use PerfView or VS Diagnostic Tools for UI/CPU/memory issues

Resources

Maintenance

  • Update this skill as project patterns or review priorities evolve.
  • Ensure checklists and heuristics match current codexes and architecture docs.

Search skills

Search the agent skills registry