code-research

To generate teaching documentation for an unfamiliar codebase, the Skill reads your code, traces entry points and data flows, and produces structured Markdown explanations starting from first principles. It analyzes local or public repositories to explain subsystems step-by-step without assuming prior knowledge.

Code research traces data flow and dependencies by reading the repository code, identifying entry points, and tracking how data moves through the system. It focuses strictly on code comprehension to produce structured Markdown documentation explaining the entire process step-by-step.

The best way to explain how existing code works is to produce teaching-oriented documentation that assumes no prior knowledge. This Skill reads your codebase and generates comprehensive Markdown docs with step-by-step rationale and examples, explaining data flows from first principles.

Yes, you can use this to understand a public repository provided by URL. The Skill applies code research to local codebases or public repositories, reading the provided code to trace dependencies and generate teaching documentation for the target subsystem.

No, code research does not work for planning new features in a codebase. The tool is designed strictly for code comprehension rather than feature planning, focusing on identifying and explaining how existing code functionality works end-to-end with step-by-step rationale.

You should use a teaching-oriented approach for codebase documentation because it explains code-paths and data flows starting from first principles, assuming no prior knowledge. This ensures comprehensive understanding of how specific subsystems operate within larger repositories through structured Markdown.