90% of your leveling should be done using volume automation - Not compression.
That all said -- Yes, you generally want to *correctively* EQ before compression (as you don't want something that might not be there anymore to trigger the compressor - Usually). Shaping EQ can certainly be applied post compression for the same reason - well, the opposing reason actually.
I'm not a Cubaser, but somewhere in there you should be able to find/change the routing of whatever inserts you're using.
* Both EQ and compression would generally be done using inserts - not sends. When you want to affect *THE SIGNAL* you're going to use an insert. When you want to *ADD SOMETHING TO* the signal (verb, echo, modulative effects, etc.), you're going to use an auxiliary send.