This article talks about using squid. I wouldn't recommend Squid for this, as they're understaffed and took years to fix critical vulnerabilities I found[0].
Using ssh-over-https with ssh -R works wonders everywhere though. You could probably even make the ssh packets look like html so it's opaque to a mitm-proxy too.
The sort of firewalls this post is discussing are close to snake oil imo. Sure they help with automated script kiddie attacks and whatnot, but yes, if you control both ends, it's nearly always possible to connect back.
Using ssh-over-https with ssh -R works wonders everywhere though. You could probably even make the ssh packets look like html so it's opaque to a mitm-proxy too.
The sort of firewalls this post is discussing are close to snake oil imo. Sure they help with automated script kiddie attacks and whatnot, but yes, if you control both ends, it's nearly always possible to connect back.
0: https://megamansec.github.io/Squid-Security-Audit/