Given that the "MELI" and "MikroTik" devices are visible as neighbors at two interfaces each, and as these interfaces are probably member ports of the same bridge, the loop may actually be there, but you've provided too little information. So a configuration export (see my automatic signature below for an anonymisation hint) of all three devices as well as a diagram of the network interconnections is necessary to find out what has happened.
I can imagine that there is an interconnection Yameli-GW <-> MELI <-> MikroTik <-> Yameli-GW again, so if no L2 loop prevention protocol is running there, there is no wonder that the frames sent from the bridge return back to it.
Instead of writing novels, post /export hide-sensitive. Use find&replace in your favourite text editor to systematically replace all occurrences of each public IP address potentially identifying you by a distinctive pattern such as my.public.ip.1.