codebase-walkthrough

Tracing an end-to-end codebase flow involves mapping the main path and side effects to build a journey diagram. It walks every leg including persistence and downstream consumers, providing file paths and line numbers before you start coding.

Planning changes across multiple services requires mapping the change impact to specific affected files and data shapes. This approach surfaces assumptions early and produces a prioritized files-to-read list, ensuring you only modify directly relevant code.

Mapping ticket impact links your requested changes to specific affected legs, files, and before/after data shapes. It explicitly traces downstream consumers and persistence points to ensure architectural understanding and prevent wasted implementation time.

Architecture tracing provides code-grounded explanations by showing relevant file paths, line ranges, and key function signatures for each leg of the flow. It extracts specific code blocks without dumping entire files to maintain context.

Implementation planning surfaces assumptions by explicitly calling out open questions and marking what must be verified rather than guessed. This prevents wasted effort by ensuring architectural understanding before proposing any implementation changes.