CI `test_script.py` does nothing

[Tomaqa]

I believe this generally applies to multiple cases, but I will stick to one concrete submission of mine.

Following submission PR #176, I got suspicious that the CI test scripts do not run the solver on actual benchmarks.
When I tracked down the generated script, it turned out that it does not do anything at all.

The job related to the PR:
https://github.com/SMT-COMP/smt-comp.github.io/actions/runs/15822493191/job/44594752384?pr=176

The script generated by the job:
https://github.com/SMT-COMP/smt-comp.github.io/actions/runs/15822493191/artifacts/3382181727

test_script.py:

from test_solver import *
init_test()

print("Testing provers")
print('OpenSMT (min-ucore)')
end()

init_test just changes the working directory, and end exits.
Am I missing something?

I do not necessarily require this to be fixed in the CI, but I need to check if the commands of my submissions specified in the JSON config files do what I intended.

Thanks!

[bobot]

For incremental track we are not generating any benchmark for testing the solvers. Another improvement would be to use real unit-tests to check the solver instead of the current dumb example generated.

[Tomaqa]

The referenced solver participates only in the Unsat Core track. Are some tests generated only for the Non-incremental track, while for all others, only dummy tests are generated?