I'll try to implement this if nobody is on it.

What info do we want to show, exactly? Which one(s) of the following is preferred?

• Hostname or IP address (and port) of the server, as given by the user as argument when starting the client
• Hostname of the local machine (where the mosh client is running)

By the way, mosh does not support the server to roam, does it? So the initial IP address of the server when the session is established will always remain the same. I'm asking because the README says:

Mosh allows the client and server to "roam" and change IP addresses, while keeping the connection alive.

But I always thought that only the client could roam.

Why not print both local hostname and remote IP?

in what senario would you using mosh in and out of a jump box? Wouldn't at least one of those connections be low latency, and could just use ssh? Unless you need to connect to a central server and then to another remote host, blegh. What about instead of double mosh, running a UDP proxy on the jump box? A wrapper script that ssh'ed to the final target to run mosh-server, parse the returned port, start UDP proxy and return new port. You could even add the proxy functionality to mosh-server, and then execute: mosh --server='ssh -tt finalhost | mosh-server proxy' jumpbox.