In some cases, its possible to mess up your policies and end up with none available for conftest to evaluate.
Currently atlantis will bark if your policy path doesn't exist, but it doesn't have a mechanism to determine if 0 tests where ran
This can cause problems if you are expecting your policies to surface problems in the proposed changes, but instead there are no tests being ran.
I'd propose that we have an option added to require 0 tests not be present in the output, or even have a minimal number of tests expected per policy_set
In some cases, its possible to mess up your policies and end up with none available for conftest to evaluate.
Currently atlantis will bark if your policy path doesn't exist, but it doesn't have a mechanism to determine if 0 tests where ran
This can cause problems if you are expecting your policies to surface problems in the proposed changes, but instead there are no tests being ran.
I'd propose that we have an option added to require
0 tests
not be present in the output, or even have a minimal number of tests expected per policy_set