AI coding agents: Run
vtz create vertz my-app to scaffold a working project, then customize
from there. See the LLM Quick Reference for condensed patterns.Define your schema
d.table() and d.model() — define columns, constraints, and relations. This is the single
source of truth for your data shape. Types flow from here to every other layer.Create an entity
entity() — turn your schema into API endpoints. Define which operations to expose, who can
access them, and add custom actions. One entity = a full REST API for that resource.Serve the API
createServer() — wire up entities and serve your API. Routes and OpenAPI docs are
auto-generated from your entities.Build the UI
A compiled UI framework with fine-grained reactivity, scoped CSS, type-safe routing, SSR, and
built-in data fetching — replacing React, Next.js, and Tailwind in one stack. Use
query() and
form() to connect to your API with full type safety end-to-end.See it in action
- Schema
- API
- Server
- UI
Key ideas
Types flow everywhere
One schema definition drives your database, API, and UI types. Change a column — the type error
shows up in your component, not at runtime.
Compiler-driven reactivity
Write plain
let and const in your UI. The compiler handles signals, computed values, and
reactive props — you never call signal() directly.Production-ready by default
OpenAPI generation, access control, environment validation, SSR — built in from day one, not
bolted on later.
One way to do things
Strong conventions mean less guessing for your team and your LLM. If there’s a Vertz way, that’s
the only way.
The stack
Install the granular packages you need:
@vertz/testing, @vertz/fetch, @vertz/agents, etc. as your app grows.
Next steps
Quickstart
Build your first Vertz app in 5 minutes.
Installation
Add Vertz to an existing project.
UI Components
Write components with JSX and the compiler.
Reactivity
Understand signals, computed values, and effects.