For pipelines submitted with primitives, where to put some extra metadata

So the idea is that TA1s would submit example pipelines together with their primitive annotations in the same repo. But I realized that just pipelines are not enough. There should also be information which datasets and problems from known datasets and problems a pipeline is meant for. We could maybe store this into pipeline description itself, under something like "suggested datasets" top-level field. Or, we could just define that each pipeline description file should also have a .meta file with same base name which would contain this information.

So the question here is: do you think this is something which would be reasonable to add to pipeline description itself? Is this something which would be in general useful for pipeline descriptions, to have suggested datasets and problems listed there. So that it is a bit more useful stand-alone.