Idea: Cause poly test to fail when no tests are run (optionally using extra arg to the command)

[svdo]

Hi Team!

I have an idea for a feature, I'm curious how you feel about this.

Is your feature request related to a problem? Please describe.
Throughout my career I have very often experienced a situation where something, either internal or external to the project, caused that tests were not actually being run in CI, which then could go unnoticed for quite a while. Typically this would be because of some dependency issue or accidental misconfiguration. Recently I experienced this with the kaocha wrapper, which needed a newer version of clj-poly but that was not immediately clear. So it turned out in CI it didn't find any tests to run and pretended that everything was fine. But in my experience that can happen for any number of reasons.

Describe the solution you'd like
I would like to be able to start the poly test runner in such a way that it fails (non-zero exit code) when it runs no tests. Either using an additional arg, or using config.

Describe alternatives you've considered
An alternative method might be to, in a script, capture the output of poly test and analyse the output to make sure that non-zero amount of tests was run. This is a more complex method that requires using a non-standard script that devs have to know about.

Additional context
N/A

[svdo]

Just as an additional thought: I'd argue that the whole point of a test runner is... 🥁 ... to run tests. So if there are no tests, shouldn't that be considered an error for any test runner? Ok, maybe over-generalising a bit, but still...