Currently we only return:
response = {
'name': cmd.name,
'description': cmd.description,
'required_parameters': cmd.required_parameters,
'optional_parameters': cmd.optional_parameters,
'default_parameter_sets': {
p.name: p.values for p in cmd.default_parameter_sets}}
which is useless when we want to test that the names/outputs of a command are fine from the plugin itself.
From: qiita-spots/qp-qiime2#61