scan-source-index

Incremental scanning progressively scans source code to build a cacheable source-index.json, reducing token usage. It performs a three-level scan—tree, header, and optional detail—to balance structural analysis depth with token budgets for large codebases.

Yes, the source indexing tool is framework-aware and supports common Java, Spring, and eGovFrame project structures. It detects these frameworks during the scan to produce structured metadata including packages, controllers, services, entities, mappers, and configs.

The best way to cache project structure analysis is generating a source-index.json file. This cache-based approach stores recent scan results, allowing you to speed up repeated analyses and keep project context fresh without re-reading the entire source tree.

To protect sensitive information when scanning source code, the incremental scan automatically masks sensitive data during processing. This ensures that secrets remain secure while building the project structure overview and caching the source index results.

Yes, you can reuse a cached source-index.json to speed up subsequent codebase analysis. The cache-based mechanism preserves recent scan results, enabling quick reuse for repeated reviews without performing a full, token-heavy scan of the source tree again.