What problem does it solve? Hard bugs and performance regressions often lead developers to guess at fixes without reproducing the failure, wasting time and shipping unverified patches. This Skill enforces a disciplined diagnosis loop so every fix is grounded in a reproducible failure and locked down by a regression test. ## Core Features & Use Cases - Feedback loop construction: Builds a fast, deterministic pass/fail signal first using ten strategies, from failing tests and curl scripts to replayed traces and bisection harnesses. - Ranked falsifiable hypotheses: Generates 3-5 ranked, falsifiable hypotheses before testing, with optional recall of past root causes from the memory plugin. - Instrumented fix and regression test: Changes one variable at a time with tagged debug logs, writes the regression test before the fix, and finishes with cleanup and a post-mortem. - Use Case: A user reports that an API endpoint intermittently returns stale data after a deploy. The Skill loops the trigger 100 times to raise the reproduction rate, ranks caching-layer hypotheses, instruments the boundary, and lands a fix with a failing-then-passing regression test. ## Quick Start Ask the agent to diagnose the bug where the checkout endpoint intermittently returns stale totals after the latest deploy.
