VeylarkStop hate raids in Twitch chat
In a hate raid, many hateful messages from new accounts arrive within a short time. Veylark checks every single one and deletes whatever breaks your rules.
What the protection package covers
The Managed Ruleset in the protection package comes with ready-made categories: racism, discrimination, hate speech, harassment, threats, sexual content, doxxing, scams and attempts to manipulate the bot. It also recognises deliberately misspelled or paraphrased terms. You switch it on with one button and choose per category whether to delete or to time out repeat offenders.
Alongside Twitch's own tools
Veylark does not replace Twitch's protections such as Shield Mode or follower-only chat. It works next to them on single messages: whatever gets through is checked and removed. Copypasta sent by many accounts at once is scored only once.
Careful, even under load
When in doubt nothing happens, and nobody is banned unless you set it up. Every action is in the log and can be undone. If the AI is busy, the word list keeps working and the dashboard shows "Moderation disrupted".
Example rule: Protection package: racism and hate speech
go back to your country
: deletedthey should all be gone
: deletedracism is a real problem in games
: staysgreetings from turkey!
: stays
Free: 1 rule, delete with reply, 10,000 messages a month. New rules start in test mode.