Vision input in Cursor SDK

Ideas Feature Requests
1

Feature request for product/service

Cursor SDK

Describe the request

Could you pretty please add support for visual input via the SDK? :backhand_index_pointing_right::backhand_index_pointing_left::flushed_face:

5

@Artemonim What’s missing from the SDK now? You can send images along with text.. Also see SDKImage.

const run = await agent.send({
  text: "What's in this screenshot?",
  images: [{ data: base64Png, mimeType: "image/png" }],
});

More details please. :slight_smile:

7

AI ​​implemented the SDK integration incorrectly, and I trusted the AI too much :roll_eyes:

I’ve now fixed it. Thank you.