Author here. The post doesn't cover testing, and that's where I need some help. Right now it's all manual against a local set of dockerized SSH servers combined with some regression testing for basic stuff.

What I can't easily cover is the stuff that breaks in the wild: servers with odd algorithm settings, flaky networks, half-open connections, auth failures that shouldn't retry.

If anyone is interested I have some promo codes to exchange for some testing, hit me up via the support email from the support page and I'll send a promo code.