- cross-posted to:
- [email protected]
- fediverse
- cross-posted to:
- [email protected]
- fediverse
cross-posted from: https://lemmy.basedcount.com/post/225787
I have built an AutoMod bot for my instance, lemmy.basedcount.com. The bot covers the following features:
- Automated removal
- of posts, based on their title, content or link
- of comments, based on their content
- configurable with either regular expressions or substrings
- User whitelisting and exceptions for moderators to selectively lift some or all of the aforementioned rules for certain users.
- Mention based pinning and locking of a post, through commands exclusively available to the mod team
- Discord notifications for new registration applications through a webhook. [only for admins]
Naturally, the bot is completely open source. I have also written a rather comprehensive (albeit long-winded) documentation and some examples.
This project is mainly targeted towards admins of small instances, however anyone can spin up their own AutoMod instance for their favourite community (provided they are a moderator there).
The automoderator is also available as a Docker image, for ease of installation.Feel free to suggest any additional features that you might want to see added to this bot.
Interesting. Thanks for this! Can it automatically post a reply or send a PM when something is removed explaining why?
Yes! When setting a removal rule for a post or a comment, you can fill an optional
message
field. After having removed the content, the bot will reply to the user with whatever you have written there.Relevant pages from the docs:
Awesome! Thanks for all your work on this :).
What, treat the user like a human being? Are you mad?
It’s a big problem in how Lemmy handles removals. I currently have to send a PM to each person explaining why and where they can post instead (e.g., asking them to post tech support questions to [email protected] instead of [email protected]).
Thankfully Photon makes it easier to do this in one step, but it doesn’t support managing multiple types of responses.