codebase-onboardinglisted

# Codebase Onboarding ## What this skill is for A new contributor cloning an unfamiliar repo usually gets one of two things: a sparse README, or an AI-generated wall of bullet points that lists facts without ever explaining how they fit together. Neither actually helps someone _understand_ the project. This skill produces something different: a single `ONBOARDING.md`, placed at the project root, that reads like a thoughtful senior engineer sat down with a new hire and walked them through the codebase — what it is, why it's built the way it is, where things live, and what you need to know before you touch the code. It is long. It is prose-heavy. It uses diagrams only where a diagram earns its place, and when it does, the diagram is precise and grounded in the actual code, not a generic box-and-arrow sketch. The single biggest failure mode to avoid: **don't just enumerate facts**. Any tool can run `ls -R` and dependency-parse a manifest file. What makes this valuable is the connective tissue — explaining _why_ a dependency was chosen, _why_ the architecture is shaped this way, _how_ a request actually flows through the system end to end. If you find yourself writing "This project uses the following libraries:" followed by a bare bulleted list with one-line descriptions, stop — you've slipped into the failure mode this skill exists to avoid. ## Before you start: read the reference files - `references/writing-style.md` — the conversational voice this document needs, with co