Auto-moderating comments
Rule-based filtering, plus optional AI for what keywords miss.
How it works
Create rules that automatically Hide or Delete matching comments. Rule types: Word/phrase (profanity list), Keyword (e.g. a competitor name or promo code), Regex pattern (for advanced matching like buy.*followers), Any link/URL (catches spam links with no pattern needed), and AI: hate/toxic, a Premium-only option that uses AI to catch hate speech and harassment a keyword list would miss (needs an Anthropic API key configured).
Every rule can be toggled on/off without deleting it, and each shows a running count of how many comments it has actioned.
Set it up
- 1
- 2
Add a rule
Pick a rule type, enter a word/phrase or pattern if needed, choose Hide or Delete, then Add rule.
- 3
Review hit counts
Check back periodically, a rule with a high hit count is working; one with zero hits after a while may need a broader pattern.
Tips to get the most out of it
- Start with Hide rather than Delete for anything you’re not 100% sure about, hidden comments are invisible to other visitors but recoverable, deleted ones aren’t.
- An "Any link/URL" rule set to Hide is a fast first line of defense against the most common spam pattern, promo links dropped under popular posts.