Abuse Guard

Open Abuse Guard

After purchasing Abuse Guard, open Configurations → Abuse Guard in the dashboard sidebar. Settings apply to the selected chatbot.

Intelligent Guard

Intelligent Guard detects flooding, off-topic replies, and abusive behavior. Turn it on for semantic detection. Static keywords remain available as fallback protection.

Choose Enforcement Mode

Select one active strategy:

• Session Rate Limit applies a temporary cooldown after the selected number of strikes.
• Identity Ban permanently blocks an identity after the selected number of strikes.

Use the sliders to set strike thresholds. Start with a measured threshold, review real conversations, then tighten it only when false positives stay low.

Blacklist and Identity Duration

Active Blacklist shows blocked targets. Under Unban Target, enter a user ID or IP address and choose Remove to clear that target.

Identity Duration controls how long identity enforcement lasts. Available choices include 1 hour, 24 hours, 7 days, and permanent.

Static Keywords

Add a word or phrase under Static Keywords. Matching uses substring filtering when Intelligent Guard is off or semantic detection fails. Remove entries that are too broad, because substrings can also match legitimate messages.

Do not use keyword filtering as the only review process for high-impact identity bans. Test terms against normal customer language first.

Save and Review

Choose Save after changing guard settings. Test spam, repeated messages, benign messages containing similar terms, and expected cooldown or ban behavior. Review Usage Log for Abuse guard events and adjust thresholds when needed.