Vite and Nitro

Why no index.html

A plain Vite project exposes index.html because Vite uses it as the application entry point. The developer owns it, edits it, and controls the HTML structure from it.

BionicJS is a framework, not a Vite template. The difference matters: when BionicJS owns the application model, the developer should not need to edit an HTML file. They write app/layout.tsx and app/page.tsx. That is the application root.

Internally, BionicJS generates an HTML entry point that Vite uses. The developer never sees it or edits it. Changes to the <html> element, <head> tags, or favicon are handled through app/layout.tsx and bionicjs.config.ts.

Vite inside BionicJS

BionicJS uses Vite as its frontend development and build tool. Vite is not exposed to the developer as a configuration surface. There is no vite.config.ts in a BionicJS project.

Internally, BionicJS creates the Vite configuration programmatically when the dev server starts. The configuration includes:

1. @vitejs/plugin-react for JSX and Fast Refresh
2. The BionicJS Vite plugin for route discovery and manifest generation
3. The proxy rule forwarding /api requests to Nitro
4. The historyApiFallback configuration for SPA routing
5. The root and entry pointing to internal BionicJS files

A developer who needs to extend Vite's configuration does so through bionicjs.config.ts, not by writing a Vite config directly.

Nitro inside BionicJS

Nitro provides the server runtime. It handles HTTP requests, compiles server code, and supports deployment to multiple targets - Node.js, Cloudflare Workers, Deno, Bun, and more.

There is no nitro.config.ts in a BionicJS project. BionicJS creates the Nitro configuration internally; API requests run through h3 handlers. Server code lives in server/ and is auto-discovered by Nitro's filesystem convention for server routes.