Very nice job @AlexandreBelling! I do believe that proof aggregation and EC cycles is one of the most challenging problems in the zk space, mostly because of the difficulty of finding the “perfect” cycle sort to speak.
At the ZKProof workshop there was a breakout session lead by Izaak Mekler on SNARK composition and it was very interesting. I have not mapped the conversation here to the notes, but here is the link, in case you find anything useful!