I would argue, I mean, if it's happening in the user's own network and not being exposed to the internet, there's minimal risk. If a user is exposing ADBTuner to the internet, they're a freaking moron. And if there's risk to ADBTuner on their home network, they have bigger problems than ADBTuner. For example, I have my home network accessible over Tailscale, but I'm the only one with access to every asset. The only other two people are my wife and my mother, and they have both very strict ACLs on their accounts to prevent them from accessing anything but a few assets that they need just to protect everything else on my network in case they were compromised. I mean, granted, I also control the authentication endpoint because I run my own OIDC provider, but that's beside the point.
More to the point, Channels DVR itself is accessible on one's home network and allows direct file system access if you were to import a media library. So one could access Channels DVR on one's home network via IP address and start deleting things out of the library.
I feel like this is something that can be hidden behind a configuration flag with a warning and allow the user to choose if they want to execute scripts make them aware of the risks and that's it. I personally think user choice is kind of important. Allowing for some sort of scripting beyond static JSON would be really game changing.
This limitation is the reason why I've been focusing my contributions on adding new features to ah4c, to be honest. Besides the open source question, which I totally understand, because I work on FOSS and have gotten absurd user requests and abusive user requests to the point where I don't want to work on it anymore so I can totally empathize with not wanting to open source, but in terms of running scripts, I don't see why you can't hide that behind a flag/switch with a stern warning.
I mean, I think you've been amazing at adding features and adding new ways of doing things, but as someone highly technical, I want to be able to do that myself. And allowing the ability to script, whether it be Bash or Python, you're just letting me write tuning scripts in BASIC, jk is kind of a big deal for someone like me. I think quite a few of us feel that way.
Edit: Also, users should not be loading scripts from remote Git repositories that they're unaware of. That's just stupid. This should function like ah4c, where you just have a shell script you run to do something. I think you also kind of have to trust your users. Like, if people do stupid things, people are going to do stupid things. Problem exists between keyboard and chair. You can't prevent it. You can warn and mitigate, but you can't prevent. And you also don't want to limit functionality.