← ClaudeAtlas

diagnosing-bugslisted

Diagnosis loop for hard bugs and performance regressions in .NET codebases. Use when the user says "diagnose"/"debug this", or reports something broken/throwing/failing/slow.
lamuniercosta/dotnet-agent-harness · ★ 0 · AI & Automation · score 72
Install: claude install-skill lamuniercosta/dotnet-agent-harness
# Diagnosing Bugs (.NET) A discipline for hard bugs. Skip phases only when explicitly justified. Assumes only .NET 8+, xUnit, and Docker. Concrete examples below name specific technologies (MongoDB, SQL Server, Redis, Hot Chocolate, gRPC) because a vague technique is useless — treat them as *instances of a pattern*, not as the stack. Read `harness.yml` and the repo's own `CLAUDE.md`/`AGENTS.md` for what is actually in use, and substitute the equivalent tool. When exploring the codebase, read `CONTEXT.md` (if it exists) to get a clear mental model of the relevant modules, and check ADRs in the area you're touching. ## Phase 1 — Build a feedback loop **This is the skill.** Everything else is mechanical. If you have a **tight** pass/fail signal for the bug — one that goes red on _this_ bug — you will find the cause; bisection, hypothesis-testing, and instrumentation all just consume it. If you don't have one, no amount of staring at code will save you. Spend disproportionate effort here. **Be aggressive. Be creative. Refuse to give up.** ### Ways to construct one — try them in roughly this order 1. **Failing xUnit test** at whatever seam reaches the bug — unit, integration, or end-to-end. Run it narrowly and fast: ```bash dotnet test --filter "FullyQualifiedName~OrderCancellationTests" --nologo -v q ``` For integration seams, prefer `WebApplicationFactory<Program>` so the whole ASP.NET pipeline (middleware, DI, Hot Chocolate) is in the loop without a running