Debugging Workflow Skill

A research-first debugging workflow formalizes problem discovery and evidence collection to find root causes before implementing fixes. It enforces a structured discipline of reproducing, diagnosing, fixing, and validating issues across software systems.

To find the root cause of failing tests and integration errors, follow a structured diagnostic sequence: research, understand, implement, validate, and document. This approach ensures you gather evidence and review documentation before applying fixes.

You can debug an unfamiliar codebase by enforcing a four-step discipline: reproduce, diagnose, fix, and validate. This process integrates with built-in debug tools and agents to guide your investigation and prevent regression.

No specific dependencies are required to start. The workflow integrates with your existing built-in debug tools and agents, guiding the process through problem discovery, evidence collection, and documented learning.

The best way to document debugging findings is using spike-based knowledge capture. You create a spike with a clear investigation title, assign it to an agent, and record your diagnostic evidence in HtmlGraph spikes for reuse.

You might skip this structured debugging workflow for trivial syntax errors requiring immediate fixes. It is designed for complex issues like unfamiliar codebases and system integrations demanding thorough root-cause analysis and documented learning.