I'm evaluating what is the better technique for SSHing into my machines securely, without trusting third parties.
I can connect to my machines through an SSH tunnel, but I need a server in between.
As far as I understood, the target local machine connects to the server through SSH tunneling, that is, it forwards everything from server:port to local:22. This is in the TCP level, so it's equivalent as having an open SSH port in the wild, which is not a problem.
However, can the server do something malicious through this reverse connection? Besides something that anyone with TCP connection could do, is there something tricky that the server can do?
I mean, something publicly known that can be done. Of course there could be exploits on the openssh server/client