investigate

Root-cause debugging identifies the actual source of software bugs rather than treating surface symptoms. A four-phase workflow of investigate, analyze, hypothesize, and implement enforces systematic evidence gathering to prevent recurring issues.

To debug a software crash deterministically, you should reproduce the error, gather evidence using file reading and searching, form a hypothesis, and only implement a fix after verifying the underlying cause.

Yes, this structured investigation workflow works across general software bugs, crashes, and unexpected behavior. It proactively surfaces questions and tests to guide the debugging process regardless of the specific error type.

The best way to find root-causes is to enforce a strict rule: no fixes without investigation. By following a structured process of analyzing evidence and hypothesizing, you avoid patching symptoms and reduce error reoccurrence.

To start a systematic root-cause investigation, simply describe the failing behavior or unexpected error. The workflow will then guide you through evidence gathering and analysis to identify the source before implementing any code changes.

Yes, this debugging approach supports web searches to gather external evidence. It leverages tools like Bash, Grep, and WebSearch during the investigation phase to collect the necessary information for forming an accurate hypothesis.