Basically two routes, they could coexist. 1. Write some tests and create optimized resource requests for modules. These would just get dropped in the

nextflow.config

and then we'd run some "Mega"tests for modules and pass those resources. The added benefit is when some full sized test data breaks something we can actually run those tests. Not sure what the cost/testing strategy would be on these, it would be a "something broke, let's go back in the git timemachine and figure out where" https://github.com/nf-core/modules/pull/4899 2. Copy the tower optimized configs as some sane defaults (But might be more likely to fail) https://github.com/nf-core/rnaseq/pull/1209 https://github.com/nf-core/fetchngs/pull/269