About explain-codebase
A repository skill for mapping and explaining a requested area of a codebase without changing it. It helps developers trace entry points, modules, callers, and data flow, while separating evidence from inference and unknowns.
Key features
- 01Read repository guidance and relevant documentation
- 02Identify a focused code area and entry points
- 03Trace modules, callers, tests, and data flow
- 04Distinguish evidence, inference, and unknowns
- 05Explain findings with focused file references
Use cases
- 01Onboarding to an unfamiliar code area
- 02Tracing how a request or data moves through a repository
- 03Understanding design choices from current code and relevant ADRs
Install
Copy the command and run it in a terminal.
$
npx skills add https:/After installing, type/explain-codebasein Claude Code to confirm it loaded.
Bundled filesagents/openai.yamlskill.toml
Details
From the project READMEExplain Codebase
Map the requested code area without changing it.
Process
- Read repo instructions and relevant current docs, Glossary, or ADRs.
- Infer the smallest useful scope when the user did not name one.
- Inspect entry points, key modules, callers, tests, and important data/control flow.
- Read relevant ADRs before using history. Inspect targeted history only when the user asks or current evidence cannot explain a consequential design choice.
- Explain ownership and flow at the user's level; distinguish verified evidence, inference, and unknowns.
Prefer a compact shape:
- Purpose
- Entry points and key modules
- Caller/data/control flow
- Relevant tests or checks
- Important terms
- Unknowns or assumptions
Use focused file references, not a search dump. Do not edit, diagnose, review, or refactor unless the user switches workflows. If evidence reveals a likely bug or architecture friction, note it briefly and recommend diagnose-issue or improve-architecture as a separate next step.
Comments
Markdown supported