We have a timeseries_crud_jscore_passthrough that implicitly creates a timeseries collection. However, most of our aggregation tests exist in jstests/aggregation and all timeseries queries run as aggregations. We should investigate if we can create a similar passthrough for aggregation tests.