This could be done either by parsing the downloaded json, or, by making API calls. For now, the code is using API calls. A very preliminary version that parses the json is also in this repository but is not being maintained.
See issue #1 for report content checklist and discussion.