Browser & Computer Control
Open URLs, inspect DOM snapshots, click, type, and move the mouse - natively.
Last updated
Open URLs, inspect DOM snapshots, click, type, and move the mouse - natively.
When the agent needs to use your machine the way a person would - open a page, inspect its DOM snapshot, click a button, type a phrase - these tools are how it does it.
Open a URL in an embedded webview the agent can read back from.
Snapshot the current page's accessibility/DOM structure, including stable element references for later actions.
The browser surface runs through CEF (Chromium Embedded Framework) and includes a security layer that scopes what pages can do. See Chromium Embedded Framework for the platform details.
Mouse - move, click, drag.
Keyboard - type text, send key chords.
Human path - moves and clicks follow human-like trajectories rather than teleporting, so they don't trip naive bot detection.
Driving sites that don't have an API or a native integration.
Multi-step UI flows where an interactive DOM snapshot provides the next actionable element.
Automating local apps from inside a chat.
Web Scraper - when you only need the article, not the whole page.
Chromium Embedded Framework - the runtime browser layer.
Last updated