ReplyMint can automatically reply to buyer comments and DMs using MintSense. This guide covers how to enable auto-reply, set confidence thresholds, and configure moderation settings.

1. Go to MintSense AI → Automation
2. Toggle "Enable Auto-Reply" on
3. Set your confidence threshold (default: 0.85)
4. Save changes

Once enabled, MintSense will automatically reply to comments and DMs that meet your confidence threshold and show buyer intent.

The confidence threshold determines which messages get auto-reply. Only messages above this threshold (and meeting other safety checks) will trigger auto-reply.

• 0.85 (recommended): Balanced — catches most clear buyer intent without false positives
• 0.75: Aggressive — more auto-replies, higher risk of misclassification
• 0.95: Conservative — only the clearest buyer signals

Start at 0.85. If you see false positives (auto-replies to non-buyers), raise it. If you're missing too many buyers, lower it.

MintSense follows these rules when deciding whether to auto-reply to a message:

1. Intent check: Message must show buyer intent (pricing, availability, product questions)
2. Confidence check: Confidence score must be above your threshold
3. Sentiment check: Sentiment must be positive or neutral (never negative)
4. Safety check: No spam, hate speech, threats, or personal information
5. Context check: Reply must be relevant to the message and post context

If any check fails, the message surfaces to your inbox for manual review.

Example: "How much is this?"
Auto-reply: "Hi! This is $24. Free shipping on orders over $50. Let me know if you'd like to order!"

Example: "Do you have this in blue?"
Auto-reply: "Yes! We have this in blue. DM me to place an order."

Example: "Do you ship to Canada?"
Auto-reply: "Yes, we ship worldwide via USPS. Shipping to Canada takes 7-10 business days."

• Negative sentiment: Complaints, refund requests, negative feedback
• Spam: Links, self-promotion, bot comments
• Unsafe content: Hate speech, threats, harassment
• Generic praise: "Love this!" without a question
• Low confidence: Ambiguous messages below your threshold

These messages appear in your inbox for manual handling.

In addition to auto-reply, you can enable auto-moderation actions in MintSense AI → Automation. The Moderation Actions card includes four toggles:

• Like positive comments — Automatically likes positive comments on Facebook (Instagram doesn't support comment likes via API)
• Hide spam — Automatically hides comments classified as spam
• Hide negative comments — Automatically hides negative feedback
• Hide hate speech — Automatically hides unsafe content (threats, hate speech, harassment)

When enabled, MintSense will automatically perform these actions on Instagram and Facebook comments. You can review hidden comments in the Moderation section of your inbox. All hiding actions are reversible — you can unhide any comment manually.

Before going live with auto-reply, test it:

1. Enable auto-reply with a high threshold (0.9+)
2. Post a test comment from a friend's account asking "How much?"
3. Check your inbox to see if MintSense auto-replied
4. Review the reply — does it sound like you?

If the reply looks good, lower the threshold to your target (0.85). If it needs work, go to MintSense AI → Knowledge or MintSense AI → Reply style and refine your business profile and brand voice.

You can see all auto-replies in your inbox — they're marked with an "Auto-reply" badge. Click any message to see:

• The original comment
• MintSense's classification and confidence score
• The reply that was sent

If you see an auto-reply that shouldn't have been sent, you can manually follow up with a correction. Then adjust your threshold or business profile to prevent similar issues.

To turn off auto-reply, go to MintSense AI → Automation and toggle "Enable Auto-Reply" off. This stops all new auto-replies immediately.

Messages will still be classified by MintSense and suggested replies will appear in your inbox, but nothing will be sent automatically.