What problem does it solve? Understanding a large or unfamiliar codebase requires hours of manual file reading, and ad-hoc exploration often loses context or produces undocumented findings. This Skill automates structured codebase research by spawning parallel read-only sub-agents and producing a durable, permalinked research document. ## Core Features & Use Cases - Parallel Sub-Agent Research: Spawns locator, analyzer, and pattern-finder agents concurrently to map where components live and how they work. - Structured Research Documents: Writes findings to flow/research/ with YAML frontmatter, git metadata, and clickable per-line repository permalinks. - Tracker Integration: Fetches work items from GitHub Issues or Azure DevOps and transitions their state as research progresses. - Use Case: A developer joining a project asks how authentication flows through the system; the Skill documents the current implementation with exact file and line references, without proposing changes. ## Quick Start Invoke /research-codebase and provide a research question or tracker issue URL to generate a documented map of the relevant codebase areas.
