Open-source AI coworkers that each get a computer of their own: a browser, files and tools, with every action decided before it happens and recorded after. Bring any AG-UI agent.
/**
* Screenshot-to-page coordinate conversion.
*
* Geometry only. Talking to a computer moved to `lib/computers/control.ts`; this stayed because it is
* about the image on screen rather than about the machine behind it.
*/
/**
* Convert display coordinates on a scaled screenshot into browser viewport coordinates.
*/
export function pageCoordinates(
image: { naturalWidth: number; naturalHeight: number },
rect: { left: number; top: number; width: number; height: number },
event: { clientX: number; clientY: number },
): { x: number; y: number } | null {
if (!rect.width || !rect.height) return null;
if (!image.naturalWidth || !image.naturalHeight) return null;
const withinX = event.clientX - rect.left;
const withinY = event.clientY - rect.top;
return {
x: Math.round((withinX / rect.width) * image.naturalWidth),
y: Math.round((withinY / rect.height) * image.naturalHeight),
};
}