Rendering modes
How the assistant shows results — built-in components, your own components, or sending the user to the right page.
The assistant doesn’t only reply with text. When a result is clearer as a table, a form, or a card, it renders one — themed to your brand. There are three ways it can present what it does.
Built-in components
Out of the box, the assistant can render a set of polished components — tables, lists, cards, forms, charts, code, and more — styled automatically from your design system so they look like part of your product, not a bolt-on.
Your own components
When you want the assistant to show something specific to your app, register your own React components. The assistant can then render them with validated inputs, so a result appears exactly as you’ve designed it.
Deep-link navigation
Sometimes the best answer isn’t to render anything — it’s to take the user to the right place. The assistant can navigate them straight to the relevant page in your app, with the right context already loaded.
How it chooses
The assistant picks the clearest way to present each result. You stay in control: you decide which components are available, and can keep the experience as simple or as rich as you like.