Skip to content

Views

Capture specific interactive states of a page: open modals, filled forms, active tabs, and more.

What are views?

A view is a child screenshot that captures a specific state of its parent page. While a screenshot captures the default state of a page, views capture what the page looks like after specific interactions—clicking a button, opening a dropdown, switching to a tab, or filling in a form.

Creating views

From a screenshot’s detail page, click the + button in the view tree sidebar to create a new view. You provide:

  • Name — A label describing this state (e.g., “Settings modal open”).
  • View instructions — Natural language instructions telling the agent what to do (e.g., “Click the Settings button in the top right corner”).

The capture agent navigates to the parent page, executes the instructions, then takes the screenshot.

Point & click creation

For precise interactions, use the Point & Click mode. In this mode, you click directly on the screenshot image to mark click points. Each click is recorded as a step in a click sequence. The agent will replay these exact clicks during capture.

View hierarchy

Views can nest. A view of a page can have its own child views, forming a tree:

Dashboard (page)
├── Settings Modal (view)
│   ├── Notifications Tab (view)
│   └── Billing Tab (view)
├── User Dropdown (view)
└── Search Results (view)

Each level inherits the parent’s URL and builds upon its interaction sequence.

View discovery

The AI can automatically discover interactive states on a page. Click Auto-Discover in the view tree on a screenshot’s detail page. The agent will explore the page, find clickable elements that reveal new states, and create views for them automatically.