Shifted about some Agent+DM code to Agent+SmartBehaviour that deals with auto-completing a "current" SmartBehaviour if it's action chain is completed and a new behaviour is being stacked on top of it.