zekroTJA
commented
2 years ago First of all, thank you very much for your many great and thought through ideas.
Generally, It would be awesome to extend the moderation features of shinpuru so that you could use it like a "swiss army knife" for general moderation.
I would generate separate issues for each of your proposals as soon as I will start working on them to track progress and also create concepts on how to implement these features.
Generally, I would prefer to finish the new web interface [#370] first before implementing these features, because It would require a lot of new bindings in the web interface and I really don't want to implement everything twice there, especially when keeping in mind that the old interface will become obsolete as soon as the new interface is ready. Or maybe it would be possible to make some sort of hybrid where I would port all guild settings first to the new interface and then relocate from the old to the new one.
I'm glad that these suggestions are coming now because I am almost finished with my last project Yuri69 and so I actually wanted to get back to this project. Really hurt a little bit that over 3 months or so, nothing has been updated over here. :D
voxain
Type
A new feature
Description
Here are some suggestions of new features that I'd like to propose for future shinpuru updates:
[ ] Anti-MassMention
Autoban (temporary or permanent) for users who mention more than x users in the same message OR more than y roles in the same message
Both x and y should be configurable, aswell as the action that is taken when the system is triggered (would be an idea for a global action system which allows you to select whether to mute, kick, ban, choose the duration etc.)
[ ] Anti-Spam
In addition to the anti-raid system, this should behave in a very similar way, but with messages instead of join events.
If user x has used up y “tokens” within z seconds before they are renewed, similar to the suggestion above, action would be taken. Now I would make that a temporary or permanent ban at first which can then be reviewed once admins get online, but in an ideal world it would use the same kind if action selector that the above suggested anti massmention system would.
An additional feature would be a global or channel-wide implementation of that system, in the cases that some tryhard would be setting up enough bots with just perfectly timed messages that dont trigger the system (even though I don't exactly see the use in a global implementation but whatever)
It would count all messages from all users inside a channel or even globally, and when the amount of messages x surpasses the amount of tokens y, before they are renewed in z seconds, all users that caused the “token overflow” would be temporarily muted/kicked/banned/…
[ ] Deduplicator
This one is still a kind of unfinished idea within my head, but it could also be used to detect messages that have the same content and were sent within x seconds from each other, which would identify anything that could be classified as “annoying as fuck”.
[ ] Global Banlist implementation
This one will definitely be controversial, but I would support the implementation of a trustworthy global banlist, such as https://ksoft.si. It would allow the bot to ban users from the guild that have been manually verified to offend a specific set of rules, before they even join the guild.
It would also allow any moderator or even user (while the latter should be regulated in a way) to report a user to the banlist, with enough clear evidence, the ban request could be manually approved and protect other server owners who chose that same implementation aswell.
[ ] Username filter
Allows setting either a global list or custom keywords that are forbidden in usernames, such as offensive names, slurs, or known scam bots such as "Discord Moderator academy" or whatever.
[ ] Antiantiping/Anti-Zalgo
I'm amost certain that i have suggested this at some point, but I can't remember which conclusion we've come to and if we ever implemented it. This feature would automaatically check new usernames for the eyecancerous ascii trickery they do because they want to be either different or edgy, and would clean up that username and set it as local nickname.
A handy little library capable of doing just that, would be ravy's sanitizer module, as it's also being used in their aero bot and has never failed to do what it's supposed to.
Additionally, it could be watching for members who change their username manually and do the same thing to these new nicknames.
That’s all for today, let me know what your thoughts are about these, if needed, i can split them up into separate issues as soon as im around a computer somewhere. I think they would only improve shinpurus solidity in the “dont go on my nerves” department, which is already very strong, but can always be better :)
Attachments
No response