Simple as dirt setup would be two soundcards in a computer. EchoLink uses one, Skype uses the other. Then cross-connect the two sound cards together. Voila! Skype + EchoLink. Usability of this would be so-so. Inbound Skype callers would be right onto the echolink node. Outbound callers can not specify who they are calling, but it's workable... If it's Linux, you can do some cli magic to tie the audio inputs together (No cables). That would be a little tricky, however.