codebase-recon

To understand an unfamiliar codebase, use script-driven discovery to detect project type, map the skeleton, and identify entry points. This minimizes file reads by surfacing core abstractions and providing a context map for targeted exploration.

The best way to find entry points and major modules is running a recon workflow. It detects the project type, maps the structural skeleton, and surfaces core abstractions to guide your exploration with minimal file reads.

To locate definitions and usages across a project, use targeted deep-dives. This technique traces data flow and scopes changes by identifying exactly where specific functions or components are called without reading entire files.

Yes, you can trace data flow and dependencies using codebase analysis scripts. The recon workflow maps the skeleton, identifies core abstractions, and performs targeted deep-dives to locate definitions, usages, and connections between components.

Yes, this codebase analysis approach works for large projects by using script-driven discovery to map dependencies and identify hot files. It scopes changes and guides targeted exploration, making it efficient for complex projects with many interconnected modules.

You should use targeted scripts instead of brute-force file reading when exploring unfamiliar code, planning features, or debugging. This approach provides a concise context map and identifies hot files, preventing unnecessary reads and accelerating understanding.