Write with AI. Run locally. Deploy to Xano.
1
Write with AI
Scaffold a project, and give your coding agent the skill that teaches it the SDK.
init writes a TypeScript backend in xano/, a React frontend, and an AGENTS.md brief
every coding agent reads.2
Run locally
Deploy the backend to the Xano Engine on your machine, then start the frontend against it.No account and no sign-in. The deploy prints the backend URL and a link that opens the
Xano editor on the Engine.
3
Deploy to Xano
Deploy the same code to an ephemeral in Xano’s cloud and run its tests, then cut a release
from it and promote it to your workspace.Where your code goes lists every target and what a deploy to each
one changes.
Governed from the first release
Code reaches your workspace through a release: a stored, named record of a backend that came up and answered. A promote lands it on a new Xano branch and leaves live alone until you say so. The same rules apply to every person and every AI agent on the team, and the Xano editor shows everyone what got built. Read more about governing AI-built code.Start where you are
A new app
From an empty folder to a backend on your machine and a frontend that reads from it.
With your coding agent
Install the agent skill and let Claude Code, Codex or Cursor build on the Xano Engine.
An existing Xano workspace
Read a workspace back as TypeScript and take it into Git.
An existing app
Add a Xano backend beside a frontend you already have.
Shipping to production
Cut a release, promote it to your workspace or deploy it to a tenant.
How it works
Def objects, bundles, and where the SDK sits next to XanoScript and the Xano editor.
Which command goes where
No account to start
The Xano Engine runs on macOS (Apple silicon) and Linux, and needs no Xano account. You need one only when you deploy to Xano’s cloud, and the Free plan covers it (Connect to Xano). The SDK is MIT licensed: source on GitHub and@xano/sdk on npm.