What problem does it solve? Debugging stubborn bugs and performance regressions often devolves into random log-grepping and guesswork. This Skill enforces a disciplined six-phase loop — build a feedback loop, reproduce, hypothesise, instrument, fix with regression test, and clean up — so bugs get root-caused instead of patched over. ## Core Features & Use Cases - Feedback Loop Construction: Guides you through ten concrete strategies for building a fast, deterministic pass/fail signal, from failing tests and curl scripts to headless browser replays and bisection harnesses. - Ranked Hypothesis Testing: Forces generation of 3-5 falsifiable hypotheses with explicit predictions before any probing, preventing anchoring on the first plausible idea. - Regression Test Discipline: Requires writing the regression test before the fix at a correct seam, and flags architectural problems when no valid seam exists. - Use Case: A user reports that an export endpoint intermittently returns corrupted data. The Skill walks you through building a replay harness from a captured request, ranking hypotheses about concurrency vs serialization, instrumenting with tagged debug logs, and landing a fix with a regression test. ## Quick Start Ask the agent to diagnose the bug where the export endpoint intermittently returns corrupted data and follow the structured diagnosis loop.
