Need a part of the schema to be generated only once as part of the response json whereas the rest fields are created based on count input

For example, I need to create data in this form:
“http_status_code”: 200,
“visible_sources”: 0,
“available_sources”: 101,
“persons_count”: 14,
“persons”: [
{
“first_name”:“xxxx”,
"last_name:“xxxxx”
},
{
“first_name”:“xxxx”,
"last_name:“xxxxx”
}
The first part should be treated as metadata of the response, the ones inside “persons” should be generated based on the count input.
Could it be possible through merging schemas or append dataset?