Currently we need to manually configure the following. This is prone to error and would need to be documented. I think we can just do this automatically, i.e. scan for the corresponding endpoint/decorator.

supports_evaluation = truesupports_graded_feedback_requests = truesupports_non_graded_feedback_requests = false

Comment:
Can we automatically detect this? :)
Also for evaluation.

Originally posted by@FelixTJDietrich in#284 (comment)