What problem does it solve? Hard bugs and performance regressions resist casual debugging because developers jump to fixes without a reliable reproduction signal. This Skill enforces a disciplined six-phase loop — build a feedback loop, reproduce, hypothesise, instrument, fix with regression test, and clean up — so root causes are found systematically instead of guessed. ## Core Features & Use Cases - Feedback loop construction: Ten ranked strategies for building a fast, deterministic pass/fail signal, from failing tests and CLI invocations to Playwright scripts, fuzz loops, and git bisect harnesses. - Ranked hypothesis testing: Generates 3-5 falsifiable hypotheses with explicit predictions before testing, then instruments one variable at a time using delve breakpoints or tagged slog logs. - Go debugging toolchain: Covers go test -run patterns, -count for flakiness, -race detection, delve, pprof benchmarks, and GODEBUG environment variables. - Use Case: A Go service intermittently returns wrong order data. The Skill guides you to build a 100-iteration test loop that reproduces the flake at a high rate, rank hypotheses about caching vs. concurrency, instrument with tagged logs, and land a regression test at the correct seam. ## Quick Start Ask the agent to diagnose the bug where the export endpoint intermittently returns stale data and have it build a reproduction loop first.
