Reply Rules engine

Rules let you steer the AI per comment — skip, flag, switch tone, or append a custom instruction — based on conditions you control. Evaluated top to bottom, first match wins.

Reply Rules engine

Settings → ReplyMind Pro → Reply Rules.

Build rules that decide what happens for each comment before the AI generates a reply. Rules evaluate top-to-bottom; first match wins.

8 condition types

  • Always — matches every comment (use last as a catch-all).
  • Keyword — comma-separated keywords, case-insensitive substring match against the comment body.
  • Category — comma-separated category IDs.
  • Word count min — comment must have at least N words.
  • Word count max — comment must have at most N words.
  • Author email — comma-separated emails, or wildcard patterns like *@example.com.
  • Post type — post, page, product, etc.

4 actions

  • Skip — do not generate a reply for this comment.
  • Flag — generate normally, but mark the comment with a "Flagged" badge for manual review.
  • Tone — override the global tone for this comment (e.g. switch to "Empathetic" for complaint keywords).
  • Prompt — append custom instructions to the system prompt for this comment ("Always offer a refund link" for keywords matching billing terms).

Drag rules to reorder. Use the per-rule toggle to disable a rule without deleting it.

Next step

→ Sentiment analysis