team-shinchan:deepsearch

You can search a large codebase by running a fast filename and keyword scan to locate API or configuration key references, followed by a deep content analysis to return file paths with surrounding code context.

Tracing dependencies and finding documentation uses a two-stage exploration process that matches file patterns first, then analyzes file contents and dependency relationships to surface related docs and traces.

Yes, you can locate symbols and files across a large repository without manual inspection by orchestrating a lightweight fast search that identifies matching paths and an optional deep agent for content analysis.

Deep codebase exploration answers find queries by first performing a rapid filename and keyword scan, then optionally deploying a deep agent to analyze surrounding code context, documentation, and dependency relationships.

Yes, deep codebase search works for large repositories by applying fast pattern matching to quickly locate files and a deeper content analysis to trace dependency relationships and surface associated documentation.

Use a two-stage code search when initial keyword scan results are insufficient, triggering the optional deep agent to perform deeper content analysis and trace dependency relationships for context-aware results.