What problem does it solve?
Understanding how an unfamiliar or large codebase works requires hours of manual searching, reading, and tracing. This Skill automates that investigation by dispatching parallel specialized agents to locate code, trace data flows, and document findings with precise file:line references.
Core Features & Use Cases
- Parallel Agent Research: Decomposes a question into 2-5 research areas and assigns them to codebase-explorer, codebase-analyst, or web-researcher agents running in parallel.
- Evidence-Based Documentation: Produces a structured research document where every claim is backed by a file:line reference, with optional GitHub permalinks.
- Follow-Up and Web Research: Supports --follow-up to append new findings to existing research files and --web to include external documentation.
- Use Case: Ask "how does authentication work" and receive a dated research document in your PRP store describing the actual implementation, architecture, and code references—without opinions or refactoring suggestions.
Quick Start
Ask the agent to research how authentication works in this codebase using /prp-codebase-question and save the findings as a research document.
