Helm charts tend to become quite complex over the time, leading to potential rendering issues and bugs.

To avoid that, we should have automated tests in place, to make sure the templates are rendered the way we expect them to be when different values are provided.

The Kubewarden team found the helm unittest to be particularly useful.

You can see that in action inside of this repo.

Acceptance criteria

• We have a set of tests assessing the helm charts being created
• Test can be run locally with make
• Test are run on each PR. They must be green in order for the PR to be merged