Instead of going to the STRING website, performing a query, exporting and importing into Cytoscape, you can simply use the stringApp in Cytoscape to perform queries and construct networks with a ton more information and detail: https://apps.cytoscape.org/apps/stringapp (see tutorials if not clear)
MCODE and other topology based clustering (like MCL) is really only useful when you have a massive network with complex, varied connectivity. These algos will identify highly intra-connected subnetworks within your network and list them from largest to smallest for you to continue working with (e.g., data overlays).
In your example, you started with simple network of 11 nodes, evenly connected. So, MCODE returned the entire network as the only "subnetwork" it could find. There are no subnetworks in this tiny network :)