To block off-topic responses, you can use the classifier rule. This rule helps ensure that the AI’s responses stay relevant to the intended topic or context. For more details, see our Rules Catalog.

Rule structure:

  • type: classifier
  • classes: List of possible topics or categories
  • expected: The expected topic or category
  • threshold: Confidence level for classification (e.g., 0.8 for 80% confidence)