▶
ExecutionTrigger
Automation/LLM/Healing
Uses vision LLM to find a visually similar element when template matching fails
Scores range from 0 to 10. Higher values mean more impact, exposure, or operational weight.
Trigger
Vision-capable LLM model
Current screenshot as base64 PNG, JPEG, WebP or GIF (a data URL is fine). Ignored when Image is connected
Current screenshot as an image, e.g. from the Screenshot node. Takes precedence over the base64 Screenshot
Screen frame of the screenshot, from the capture node. Coordinates are desktop input coordinates (ready for the mouse nodes) when Frame is connected, otherwise pixels of the original screenshot
Base64-encoded template image (PNG, JPEG, WebP or GIF) that failed to match
Description of what the template represents
Where the element was previously found (x,y in desktop input coordinates (ready for the mouse nodes) when Frame is connected, otherwise pixels of the original screenshot)
Continue
Could not heal, or the model gave no point on the screenshot
Healed template result; points and regions are in desktop input coordinates (ready for the mouse nodes) when Frame is connected, otherwise pixels of the original screenshot
X coordinate of the found element's center, in desktop input coordinates (ready for the mouse nodes) when Frame is connected, otherwise pixels of the original screenshot
Y coordinate of the found element's center, in desktop input coordinates (ready for the mouse nodes) when Frame is connected, otherwise pixels of the original screenshot