debugging

Enforce root-cause investigation before fixing bugs and test failures.

Updated Aug 27, 2026

Search Tags:#debugging #systematic #triage #phases #root-cause #error-messages

Ready to run

npx skills add https://github.com/ArcadeAI/dexa --skill debugging-arcadeai

💬 Prompt for ChatGPT / Claude Chat▼

Please help me install this Agent Skill.
Skill: debugging
Source: https://github.com/ArcadeAI/dexa/tree/main/.claude/skills/safeword-debugging
Command: npx skills add https://github.com/ArcadeAI/dexa --skill debugging-arcadeai

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

Four-phase debugging framework that ensures root cause identification before fixes. Use when encountering bugs, test failures, unexpected behavior, or when previous fix attempts failed. It enforces investigative discipline (e.g., "debug this", "fix this error", "test is failing", "not working").

Core Features & Use Cases

Systematic Debugger: Find root cause before fixing; solving symptoms first is avoided.
Four-Phase Process: Root Cause Investigation, Pattern Analysis, Hypothesis Testing, and Implementation verification.
When to Use: In urgent bug scenarios, failed fixes, or flaky tests to ensure durable resolutions.

Quick Start

Start with Phase 1: Root Cause Investigation and proceed through Phases 2–4 as described. Document findings and confirm root cause before applying any fix.

Frequently Asked Questions about debugging

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

FAQPage Schema

How do I debug a problem systematically to find the root cause?▼

Root-cause debugging follows a four-phase process: investigate the error completely, analyze patterns across failures, form and test hypotheses against data flow, then verify the fix. This ensures you solve the underlying issue, not just symptoms.

What should I do when my fix didn't work and the error keeps happening?▼

Repeated failed fixes signal you're treating symptoms. Return to phase one: read error messages completely, reproduce the failure consistently, trace data flow through affected components, and compare differences between working and broken states before attempting another fix.

Can I use this debugging approach for test failures and unexpected behavior?▼

Yes. The four-phase framework applies to bugs, test failures, unexpected behavior, and emergency scenarios across multi-component systems. It enforces investigative discipline regardless of failure type, ensuring durable resolutions under time pressure.

Why does enforcing root-cause investigation before fixes matter?▼

Fixing symptoms first masks underlying problems and leads to repeated failures. Systematic root-cause investigation—reading error messages, reproducing consistently, tracing data flow, and testing hypotheses—identifies what actually broke and prevents the bug from recurring.

What's the best way to organize debugging work across multiple system components?▼

Trace data flow through all affected components while comparing differences between working and broken states. This phase-based approach isolates which component introduced the failure and reveals how errors propagate, preventing incomplete fixes in distributed systems.