codebase-scanner

You can scan a codebase of any size to generate hierarchical documentation from L0 to L3, providing a high-level overview down to specific modules. This allows developers and AI agents to instantly understand project architecture and navigate complex code efficiently.

Incremental scanning works by analyzing only changed files using git-diff, rather than rescanning the entire project. This enables faster documentation updates and allows AI agents to quickly process recent modifications without repeating a full codebase scan.

Yes, semantic analysis automatically extracts business rules, security risks, and technical debt from your codebase. This helps identify critical code sections and navigate complex project structures without manually reviewing every file.

Yes, the codebase scanner supports projects of any size, including codebases with millions of lines. A new developer can use it to get an L0 overview, understand the L1 architecture, and dive into L2 domains and L3 modules to quickly become productive.

The best way is to perform a full codebase scan to generate hierarchical documentation and semantic analysis. This enables AI agents to understand project architecture, identify critical sections, and navigate complex codebases efficiently without manual context building.

Use incremental scanning when you need faster documentation updates for changed files only, rather than regenerating all documentation levels. A full scan is necessary initially to establish the L0-L3 baseline, after which git-diff scanning maintains it efficiently.