I get loads of them when I'm on a Netsweeper filtered network... pretty much any time any asset a page loads is from a blocked site (social media pixels normally).
I'd hazard a guess that it's based on what the LLM can "design" without actually being able to see it or have taste and it still reliably look fine to humans.
When you have a simple tool you have written for yourself, that you need to be reliable and accessible but also that you don't use frequently enough that it's worth the bother of running on your own server with all of that setup and ongoing maintenance.
Not sure if this is the right place to ask, but I've got a semi-obscure DNS question.
I'd like to use Cloudflare's Zero Trust DNS filtering with DoH by running a DNS proxy on my network.
I can get this to work great with github.com/adguardTeam/dnsproxy (running on a Pi 4B) but what I would really like is to have different devices (based on their IP on the network) get their queries forwarded onto a different DoH upstream.