I am trying to decrease the API calls as much as possible, so that is not an option.
I might just have the owner decide with which mod they want to share it with. That seems to be the easiest option tbh. It also shifts the responsibility from the developer (me) to the consumer (the owner).
I am thinking of having a simple login form with a community ID field (which will be provided by the bot along with the password) and a password field which will just be a UUID. There is no checking done whether the person logging in is really a moderator of community X or not so yeah.
LLMs are too expensive to run and slow as well. As for the lockdown system, DoS as in sending API calls to send messages? Yeah well that's considered spam and will indeed lock the community down. But it won't happen instantly and I still need to plan how it would detect spam and whether it will attempt to automatically ban the spammers or not.
The bot is still in the preparation phase, I haven't done any work yet.
As discussed in some other comment, the bot will not delete posts that are triggered by regex or word blacklist but report them.
The bot indeed won't understand context so it will do more harm than good. I am also thinking of transliterating content to catch people using symbols that look like certain characters like € for E and $ for S. That will be done at the webhook level though.
What exactly am I to do when mods use my bot maliciously? I just try to program features into the bot that might be used by the moderators. Everything is optional, if the features are used maliciously, it's not my fault but the moderators'. What fault does the creator of knives have when they are being used to murder people instead of cutting vegetables?
I do appreciate your comment though, some misc tasks don't necessarily need to exist such as welcome messages and auto comments on posts. Actually I'll remove welcome messages, they are a waste of API calls. Maybe auto comment on posts as well but scheduled posts are (most likely) staying. It is a moderation bot after all. I'll consider your compliant. Thanks.
I am thinking of making a website for it to act as a dashboard. The community owner will get a UUID as a password to access the dashboard of the community. For security reasons, the password will be reset when the owner of a community changes. I actually just thought of hosting the bot myself on lemmings.world.
As for the mods accessing the dashboard to configure the bot, I don't know yet. Any suggestions?
I am merely trying to give community mods options. This feature and the other features are optional. Direct your complaints to the community owners if they use some regex that matches usernames that you think shouldn't be banned.
Locking down an entire community because of a small group of people spamming.
The bot just locks it down to stop the spam, otherwise everyone's feed will just be filled with spam. I haven't seen such a spam yet, but that does not mean there won't be any in the future. Just trying to be prepared for it.
Deleting posts because an account isn’t old enough?
Again, I am just giving the mods options. If they enable the feature and use it, direct your complaints to them.
Why not throw in the system to have to approve posts before they get published? Really make the community welcoming.
That is possible with post locking and with a dashboard. I'll look into it.
It was said in another comment above that this tool is easily abused by “overzealous mods”, but I believe the real problem are overzealous programmers.
Again, I'm only giving them options.
Every tool can be used both in good and bad purposes. Why is it that it is the fault of the tool or its creator?
Why though? We don't need a symbol. Is it that hard to type "fediverse"? The fediverse logo is good enough.