Import the call in a piece of software called postman.
From there you can see any headers you need to set as well as what body and query parameters need to be set. It’s super handy and it’s what I use most of the time when I have to figure out a curl call
this is the final block that works - it’s calling the correct endpoint /recognition/v2/classify (the above was mistakenly pointing to /detection/v2/detect)
this is based on a classify run to classify only one image. (see object inside list in records parameter). if more photos are needed, simply them as new objects to the list - but then you have to modify the response processing to use a for loop so that the property string “records[1].best_label.name” is adjusted so the index to records reflects the current value.)