What problem does it solve?
Four-phase hypothesis-driven debugging methodology with root cause analysis. Activates when investigating bugs, fixing test failures, diagnosing runtime errors, troubleshooting unexpected behavior, analyzing performance issues, or resolving race conditions. Enforces NO FIXES WITHOUT ROOT CAUSE FIRST.
Core Features & Use Cases
- Phase 1: Root Cause Investigation — read error messages thoroughly, reproduce the issue consistently, examine recent changes, gather diagnostic evidence, and trace data flow to the origin.
- Phase 2: Pattern Analysis — locate working examples, compare implementations completely, list all differences, understand dependencies, and check underlying assumptions.
- Phase 3: Hypothesis and Testing — formulate ONE clear hypothesis, design a minimal test, predict the outcome, run the test, compare actual vs expected, and iterate as needed.
- Phase 4: Implementation — create a fix only after root cause confirmation, verify with tests, and perform a brief post-mortem to prevent regression.
- Use cases include debugging runtime errors, failing tests, performance issues, race conditions, and unexpected behavior in complex systems.
Quick Start
Reproduce the issue, document the error messages, formulate one hypothesis, and validate it with a minimal, verifiable change.