AnyQR

Use to open the QR Codes widget whenever the conversation involves viewing, decoding, creating, or editing a QR code. This tool accepts exactly one of three input modes: a ChatGPT file containing a QR image, a remote image URL pointing to a QR image, or structured QR information to encode. When a QR image is supplied, the tool downloads and decodes the QR payload on the server before rendering the widget, so the model can immediately reason over the decoded information and the widget can open with the same data loaded. When structured information is supplied, the tool normalizes that payload and opens the widget with a generated QR representation. When no input is supplied, the tool opens an empty editor so the user can start from scratch. The tool is best for situations where the user wants an interactive surface: checking what a QR code contains, iterating on a QR code's meaning, switching between structured fields and raw encoded text, or exporting the latest QR image back into the conversation. It supports common QR payload families including plain text, web URLs, mail actions, telephone actions, SMS actions, vCard contact cards, Wi-Fi credentials, calendar events, geolocation coordinates, and custom raw payloads. Because the widget persists edited model-visible state, it is also appropriate when the model needs to stay aware of the latest QR contents after the user changes them in the interface. Do not use this tool for non-QR barcode formats, generic OCR, or when the user has asked not to open a widget. If the user only wants a textual explanation of QR codes or a direct answer unrelated to QR manipulation, respond normally instead.

qr-code