gnd-doku.json

{
	"fields": {
		"P58": {
			"label": "Person – preferred name",
			"description": "GND data field for the preferred name of a person or a family",
			"aliases": [
				"PICA+ 028A",
				"PICA3 100",
				"MARC 21 100"
			],
			"viewLink": "https://doku.wikibase.wiki/wiki/datafield?property=P58",
			"editLink": "https://doku.wikibase.wiki/wiki/Property:P58",
			"subfields": {
				"P41": {
					"label": "Personal name",
					"occurrence": false,
					"description": "Hinweis zur Erfassung von Personen, die nur unter einem Nachnamen bekannt sind"
				},
				"P21": {
					"label": "Surname",
					"occurrence": false
				}
			},
			"codings": {
				"PICA+": {
					"fieldName": "028A"
				},
				"PICA3": {
					"fieldName": "100"
				},
				"MARC 21": {
					"fieldName": "100"
				}
			},
			"rulesOfUse": {
			},
			"examples": {
				"Q626": {
					"label": "Sarah Hartmann"
				}
			}
		},

		"P150": {
			"label": "Identification number"
		}
	}
}

* instead of "fields" I would suggest "gndfields"
* instead of "subfields" I would suggest "gndsubfields"

That is your call, but I'd not do this, since the whole JSON already is GND specific and is returned by a GND specific API. In other words, there will never be non-GND stuff in there so that the subfields are GND subfields should be clear from the context.

Okay makes sense, I agree with you.