In a thread in the ginkgo/onsi repo I asked about extracting the test names - onsi replied 1 suggesting I take a look at the dry-run json output which features quite a lot of data while it also has the full test names.
So the task is to improve the property extraction in the way he suggested. In that way we can also extract labels and add them to the data.
This will benefit on what @fossedihelm is wanting to work on (topic is labelling PRs if tests with a certain decorator are changed). Also we can use this to improve on the check-tests-for-flakes, where @jean-edouard suggested to find an approach for determining whether we really need three nodes 2