Spam deleted before your customers ever see it
Every incoming comment is checked against your moderation rules first. Scams, shills and abuse are removed on sight — and AI writes the term list for you.
What it does
One “DM me to earn 💰” comment under your best post costs trust and sales. catais treats moderation as the first gate: before the agent considers replying to anything, the comment is screened against your auto-delete list, and matches are removed instantly via the platform API — then logged so you can audit every removal.
Building the list is effortless. Hit “Suggest with AI” and the agent generates a high-precision set of terms tailored to your business — crypto shills, fake giveaways, “check my profile” bait, competitor spam, slurs — phrases that are almost always abuse and almost never a real customer. You can add, trim or override anything.
Moderation is part of a wider safety net: excluded topics keep the agent from ever discussing what you've ruled out; stay-silent lists make it skip (and log) comments that need a human; and sensitive data rules keep order details and personal info out of public threads entirely.
How Moderation works
Set your rules
Auto-delete terms, excluded topics, stay-silent triggers and DM-only data — in Agent Context, with AI suggestions one click away.
Every comment screened
Real-time webhook comments and sweep-collected comments both pass the moderation gate before any reply logic.
Matches removed & logged
Matching comments are deleted via the platform API and recorded in your activity feed with the matched reason.
Edge cases escalate
Stay-silent topics are skipped with a logged reason; complaints raise alerts to your email or Telegram so a human steps in.
Who it's for
Pages targeted by scam rings
Crypto and “earn from home” spam disappears the second it lands.
Brands protecting community
Customers see a clean, on-topic thread — which keeps them commenting.
Regulated or sensitive niches
Health and finance pages can hard-exclude claims and topics the agent must never touch.
Everything included
- Auto-delete on sight, before any reply
- AI-generated term lists tailored to your business
- Case-insensitive matching tuned for precision
- Full audit log of every removal
- Excluded topics & stay-silent escalation
- Sensitive data restricted to private DMs
- Complaint alerts to email/Telegram
Moderation — questions
Could it delete a real customer's comment?+
The AI-suggested terms are deliberately high-precision (phrases that are almost always spam), and you control the final list. Every deletion is logged so you can review and refine.
Does moderation run before or after replies?+
Before — always. A comment that matches your list is removed without ever reaching the reply engine.
Can I see what was removed?+
Yes. Every moderation action appears in your agent activity feed with the comment text and the reason.
What about comments that need a human?+
Stay-silent rules make the agent skip them (logged, with reason), and complaint detection can alert you instantly by email or Telegram.
Works together with
Try Moderation on your own store
Free plan, one page, no card — see the agent work before paying anything.
Get started free