Development

Kaja is one Bun monorepo. These pages cover how it fits together, how to run it, and how it ships.

🏛️Architecture The repository map and the ideas behind it. 💻Local setup Run Kaja from source: install, commands, hooks, URLs and tests. 🌀Agent brain Inside the agent loop, stores and model client. 📡API The Hono API: routes, auth, rate limits and the cloud agent. 🗄️Database Both databases, the tables and how the schema is managed. 🕸️Web The TanStack Start app: public pages and the signed-in portal. 📐Schema Where every Zod schema lives. 🏪Marketplace internals How the marketplace reaches local and cloud users. 🚀Deployment How kaja.io is deployed and what production needs. 🐘Postgres on a volume Keep Postgres data on a volume that outlives the server. 🐞Debug flow Dump a terminal conversation to markdown and see what happened. 🎧Vibe coding What the repo gives a coding agent, and what runs on commit and push.