research_codebase_generic

To analyze a codebase and understand its architecture, this Skill spawns parallel sub-agents that investigate various components concurrently, identify patterns, and synthesize findings into structured documentation for actionable developer insights.

The best way to document an unfamiliar codebase automatically is by using parallel agents to conduct targeted research, which then generates detailed documents containing findings, code references, and architectural insights.

Yes, you can research specific implementation details like user authentication by breaking down complex queries into manageable research areas, allowing parallel agents to consolidate all relevant files and functions into a single report.

Parallel agent execution improves codebase research by investigating multiple components simultaneously, significantly reducing the time developers spend understanding existing code and synthesizing the results into structured output.

A key limitation of using automated agents for codebase analysis is that the depth of findings depends on how well queries are structured, requiring targeted research breakdowns to accurately identify relevant components.

You do not need to manually prepare research areas; the Skill breaks down complex queries into manageable research areas automatically, spawning sub-agents to investigate each domain concurrently.