I have to say this was perplexing me for the longest time as well. I was wondering if I didn’t understand interfaces as well as I did.
So, in summary we create the ItemLayerInterface and populate it with any functions that we want to share across the Items (Like a normal interface blueprint). We then define the base functionality of these functions in our ItemLayerBase anim blueprint. We then can make child blueprints from the base and customize them for the different items.
The MannequinBase anim blueprint inherits the ItemLayerInterface so that we have placeholders to the LinkedLayers to use to further define the logic. Unlike the normal interfaces we didn’t define anything here, just used as a placeholder for the linked layer that we will set during runtime.
Then on the character we can use the node LinkAnimClassLayers to use one of the child ItemLayers ie (Unarmed) and set it on the mesh.
(Keeping It Simple) This could happen at BeginPlay for the character for instance . Could also then be applied when equipping a weapon etc.
It would have been nice if in the Lyra writeup they would have added that little bit of additional explanation. I guess there are a lot of new things going on with Lyra and where do you stop.
Cheers!