Hey everyone,
I’ve been testing Cursor Agent Auto and wanted to use it to analyze a screenshot of a website or form in order to automatically recreate its structure (HTML, fields, layout, etc.).
However, it seems the agent can’t interpret screenshots at all — it only works with text input or existing code.
Has anyone figured out a way to extract a website or form layout from a screenshot, maybe using another tool or workflow that could be combined with Cursor?
Any tips or ideas would be greatly appreciated! 
Cheers