When should I use the systematic debugging skill?

Use it for any technical issue including test failures, production bugs, build errors, or performance problems, especially when you feel tempted to apply a 'quick fix' without full understanding.

What happens if I've tried multiple fixes and the bug still persists?

After three failed attempts, the protocol mandates an architectural review to determine if the problem is a fundamental design flaw rather than a simple logic bug.

What is the 'Iron Law' of this debugging skill?

The Iron Law dictates that no code fixes should be attempted or proposed until a definitive root cause has been identified and verified through evidence.

Does this skill require writing tests?

Yes, Phase 4 of the process requires creating a failing test case that reproduces the bug before implementing the final fix to ensure the solution is verified.

How does this skill handle complex multi-component systems?

It guides the user to add diagnostic instrumentation at every component boundary to log data entry/exit, allowing you to isolate exactly which layer is failing before investigating the code.

Systematic Debugging

Name: Systematic Debugging
Author: claudiodearaujo

byclaudiodearaujo

•

Developer Tools

Enforces a rigorous four-phase root cause analysis protocol to eliminate guess-and-check coding and superficial bug fixes.

Systematic Debugging is a specialized Claude Code skill designed to stop the cycle of 'bug thrashing' and regressions. By mandating 'The Iron Law'—no fixes without root cause investigation—it guides developers through a disciplined workflow of evidence gathering, pattern analysis, and minimal hypothesis testing. This skill is essential for resolving complex issues in multi-component systems, intermittent test failures, and performance bottlenecks. It prevents the introduction of new bugs by requiring a verified understanding of the problem and the creation of failing test cases before any implementation begins, ensuring that every fix is both precise and permanent.

Key Features

01Integrated verification steps requiring failing test cases (TDD) before final fix deployment

021 GitHub stars

03Architectural 'STOP' signals that trigger after three failed fix attempts to re-evaluate system design

04Strict 4-phase debugging workflow: Investigation, Pattern, Hypothesis, and Implementation

05Mandatory root cause verification protocol to prevent 'guess-and-check' patches

06Diagnostic instrumentation patterns for tracing data across multi-component system boundaries

Use Cases

01Tracing deep-seated data flow issues across microservices or multi-layered architectures

02Resolving intermittent or 'flaky' test failures in complex CI/CD pipelines

03Breaking out of unproductive debugging loops where previous 'quick fixes' have failed

What are Skills?·How to Install

Install with 🐟 Skill.Fish

npx skillfish add claudiodearaujo/sistema-de-narra-o-de-livro systematic-debugging

For use in Claude.ai and ChatGPT

Download Skill