What problem does it solve? Hard bugs and performance regressions stall when developers jump straight to guessing fixes without a reproducible 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 by luck. ## Core Features & Use Cases - Feedback loop construction: Ten ranked strategies for building a deterministic pass/fail signal, from failing tests and curl scripts to replayed traces, fuzz loops, and git bisect harnesses. - Ranked falsifiable hypotheses: Generates 3-5 hypotheses with explicit predictions before testing, avoiding single-hypothesis anchoring. - Tagged instrumentation and cleanup: Every debug log gets a unique [DEBUG-xxxx] prefix so cleanup is a single grep, and the post-mortem captures the confirmed root cause in the commit message. - Use Case: A user reports "the export button throws intermittently." The Skill builds a loop (possibly using the bundled human-in-the-loop bash template), reproduces the failure, tests ranked hypotheses one variable at a time, and lands a fix guarded by a regression test. ## Quick Start Ask the agent to diagnose the bug where the export endpoint intermittently returns a 500 error and have it build a reproduction loop first.
