debugging

Diagnose failures systematically before changing code.

23|2|Updated May 27, 2026

Search Tags:#testing #debugging #reproduce #root-cause-analysis #hypotheses #bug-triage

Ready to run

npx skills add https://github.com/zhouguoqing/QianYuan.AIAgenticFramework --skill debugging-zhouguoqing

💬 Prompt for ChatGPT / Claude Chat▼

Please help me install this Agent Skill.
Skill: debugging
Source: https://github.com/zhouguoqing/QianYuan.AIAgenticFramework/tree/main/samples/skills/debugging
Command: npx skills add https://github.com/zhouguoqing/QianYuan.AIAgenticFramework --skill debugging-zhouguoqing

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

Diagnose failures systematically before changing code.

Core Features & Use Cases

Reproduce or inspect the failure before proposing a fix.
Separate symptoms, confirmed facts, hypotheses, and unknowns.
Prefer the smallest experiment that can disprove the leading hypothesis.
Fix the root cause when practical, not just the visible symptom.
Add or update focused tests when the bug has a clear behavioral contract.

Quick Start

Reproduce the failure in your environment and document observed symptoms, suspected causes, and attempted experiments to guide a fix.

Frequently Asked Questions about debugging

High-intent search queries and answers about installing and using this skill.

FAQPage Schema

How do I systematically debug a failing test before changing code?▼

To debug failing tests systematically, first reproduce the failure and capture observed symptoms. Then separate confirmed facts from hypotheses, run minimal experiments to disprove leading causes, and apply a root-cause fix before focused retesting.

What is the best way to isolate root causes during bug triage?▼

Root-cause analysis during bug triage involves separating symptoms from confirmed facts and unknowns. You generate hypotheses, prefer the smallest experiment that can disprove each one, and fix the underlying cause rather than the visible symptom.

How do I diagnose production incidents without applying premature fixes?▼

Diagnose production incidents by forcing symptom capture and structured failure analysis. Document confirmed facts and hypotheses, run minimal experiments to test them, and only apply root-cause fixes once the leading hypothesis is disproven.

Why should I reproduce a bug before attempting a code fix?▼

Reproducing a bug before fixing code prevents premature or incorrect patches. It allows you to inspect the actual failure, separate symptoms from hypotheses, and design the smallest experiment to disprove your suspected cause.

Can I use this structured failure analysis for exceptions across different codebases?▼

Yes, structured failure analysis is applicable to bugs, failing tests, exceptions, and production incidents across various codebases and environments. It systematically forces symptom capture, hypothesis generation, minimal experiments, and root-cause fixes.

When should I add or update tests during debugging?▼

Add or update focused tests during debugging when the bug has a clear behavioral contract. This follows applying a root-cause fix, ensuring the minimal experiment results are verified and the original failure is resolved.