How does this skill prevent 'guess-and-check' debugging?

It enforces an 'Iron Law' where Phase 1 (Root Cause Investigation) must be completed and evidence must be gathered before any code fixes can be proposed.

What happens if I've tried three different fixes and none worked?

The skill triggers a specific architectural review phase, mandating that the developer stop patching and discuss whether the underlying pattern or system design is fundamentally flawed.

When should I use the Systematic Debugging skill?

Use it for any technical issue, especially when under time pressure, after multiple failed fix attempts, or when dealing with complex multi-component systems where the failure point is unclear.

How does it handle multi-component system errors?

It provides specific patterns for adding diagnostic instrumentation at every component boundary to log data entry/exit, allowing you to see exactly where a process breaks.

Does this skill work with test-driven development (TDD)?

Yes, it explicitly integrates with TDD by requiring the creation of a minimal failing test case in Phase 4 before any fix is implemented.

Systematic Debugging

Name: Systematic Debugging
Author: dkmaker

bydkmaker

•

安全与测试

Enforces a rigorous, four-phase root cause analysis process to eliminate guesswork and prevent recurring bugs during development.

The Systematic Debugging skill transforms the troubleshooting process by enforcing an 'Iron Law' that prohibits code fixes until the root cause is fully identified and documented. It guides Claude through a structured workflow—Root Cause Investigation, Pattern Analysis, Hypothesis Testing, and Implementation—to ensure that developers stop patching symptoms and start solving underlying issues. By utilizing diagnostic instrumentation and data flow tracing, this skill reduces development thrashing, prevents the introduction of new bugs, and identifies when an issue requires fundamental architectural changes rather than simple code patches.

主要功能

01Rigorous pattern analysis comparing broken code against known working reference implementations.

02Scientific hypothesis testing with minimal, single-variable code modifications for verification.

03Multi-component diagnostic instrumentation for identifying failures in complex CI/CD or API systems.

04Architectural red-flag detection that triggers fundamental reviews after three failed fix attempts.

051 GitHub stars

06Mandatory root cause investigation phase before any code modifications are proposed.

使用场景

01Transitioning from 'guess-and-check' patching to high-confidence, evidence-based architectural fixes.

02Resolving intermittent test failures or production bugs in complex distributed systems.

03Debugging deeply nested data flow issues through systematic backward call-stack tracing.

What are Skills?·How to Install

Install with 🐟 Skill.Fish

npx skillfish add dkmaker/my-claude-plugins systematic-debugging

For use in Claude.ai and ChatGPT

Download Skill