Interface: ImageInput
@kortexya/reasoninglayer / ImageInput
Interface: ImageInput
Section titled “Interface: ImageInput”Defined in: src/types/query.ts:412
An image attached to a new TRIZ session.
Remarks
Section titled “Remarks”The vision model reads the image, and the engine adds the reading to the
problem text before it structures the problem. The engine refuses with
400 a data that is not standard base64 (RFC 4648, with padding; a
data: URL is refused) and a mime that is not an image/* type.
Properties
Section titled “Properties”data:
string
Defined in: src/types/query.ts:414
Standard base64 of the raw image bytes. Not a data: URL.
description?
Section titled “description?”
optionaldescription?:string
Defined in: src/types/query.ts:418
Optional description, given to the vision model with the image.
mime:
string
Defined in: src/types/query.ts:416
MIME type of the image; must be image/* (e.g. "image/png").