Most of these things are pretty secure out of the box.
Even without fail2ban disabling root login and only allowing SSH key authentication makes those scripts just a waste of time for the attacker. That game is a low effort attempt to just get the low hanging fruit for botnets though.
Yes, My home network setup is a bit complicated but I am using Pfsense so I have things on separate vlans with internal firewall rules to reduce risks.
All traffic in on port 443 is routed from Cloudflare to an NginX reverse proxy which decides how to connect back into my network for things
Years ago I would just run a server on the network with 443, 80 and 22 exposed directly to the world and never had any major issues. (Other than the normal automated attacks trying to gain shell access over SSH)
I got 1gbps internet (symmetrical) and a raspberry pi cluster… running my own Wordpress never made more sense… AND that botch should scale!
Local models are actually pretty great! They are not great at everything… but for what most people are using LLMs for they do a fine job.
Thats from llama3.1:8b and the answer is decent, it took about 20seconds to generate my answer and used no more power than if I were to play a video game for the same amount of time.