Sorry about that! There's rules in place for catching bulk quote posts within a given period of time, to prevent folks from spamming mentions, but the thresholds were set too low, so it caught some false positives that I wasn't expecting. You should be good to go now!
That was someone else! I don't think it exists anymore, because I can't find it (my search was only cursory though)
oh i'm glad he got the account back, unlabeled!
Fixed a broken rule and we should catch the accounts reposting 100s of users in a go now.
Following 500 accounts in 5 minutes, twice in one hour, has been a pretty strong "this account is fake" signal in the past. May need to adjust
Thank you for building this! We need to build a sustainable ecosystem and I love that this is an atproto native way to surface this information!
Folks: Please stop reporting @attie.ai. This is out of scope for us.
Unfortunately no, and I wish there was. If you can send me the account, I can take a look at it later today.
PDS hosters: folks should not have PDS_INVITE_REQUIRED=false in their config unless they know what they are doing.
at a minimum you need CAPTCHAs on all signup routes, to block/disable the createAccount API flow, etc.
we know this should be smoother.
Oh you found an edge case. Gonna need to fix this.
I hmm. I have no idea. Let me go look what rule fired on that and fix it for you
Unfortunately often times we only catch them with the repost event, but when we do they end up on the Platform Abuse & Manipulation list, though they often get taken down shortly after that by Bluesky Moderation: bsky.app/profile/did:...
Sorry about that!
He's good, false positive, fixed the rule. thanks!
I've realized I've been telling PDS admins how they can tackle these spam accounts from being created on their PDS, but just realized I never explained that as the end user, YOU can also hide them from your views with @skywatch.blue. Wrote how below.
@asukafield.xyz or the A Very British Bigotry blocklist maybe?
Label Scanner interface on white background with barcode icon and title. Search field labelled "Handle or URL:" contains DID "did:plc:23zfas5o6aufneieinzh2cq6" with Search button. Results display "6 labels found:" with orange tag icon. Six labels listed with linked sources: "time-machine-accident-past" from The Cave of Trophoniüs; "joined-nov-c" from Joined Labeler; "domain-instagram" from Girl Bye with winged emoji; "domain-x" from Girl Bye with winged emoji; "made-over-fifty-replies-yesterday" from Account Activity Labeler; "monthly-replies-over-ten-per-day" from Account Activity Labeler. Each label name and source is displayed as clickable blue hyperlink.
You don't have any of our labels
I removed him and I'll watch to see if they get picked up again, as it might have been due to the volume of shopping links. We can allowList the account if we have to.
I removed the the label and he should automatically be removed from the list. Let me know if he's still blocked for you and I can look into why and thank you for flagging!
The variant in your profile is now allow listed and shouldn't get picked up again! I will look into other variants to make sure they're added too.
If he's a false positive, can you send his handle our way and we can remove him from the list!
Oh thanks for letting me know, I try to allow list pride flag combos but was unaware of this one. I will address it and remove your label today
Oh, and I forgot Inauthentic Fundraising!
In addition, we very much recommend subscribing to @labeler.hailey.at for overlapping coverage.
Been inundated by spam / scammer accounts lately? We've recently added some labels to help clarify why accounts get labeled: Bulk Following, Engagement Abuse / Hacking and Spam. We also encourage you to use our Platform Manipulation label!
FYI: the automod is currently lagging behind substantially. It’s slowingly catching up, but will probably be another few hours
lol there is but it might be too sensitive you should be good to go now
Yeah! I try to filter it out, but sometimes the tool used to determine a profile's language is not accurate and guesses wrong. Sorry about that
Desole. Automod tries its best to flag only English language accounts, but sometimes can't accurately determine the profile language. In this case it flagged "Un Train de Retard" as an english slur :(
Looks like "follow me on twitter" got flagged, but that was inadvertent and the rule should be tighter because that's not follow back behavior. Sorry about that!