Skip to main content

Custom Intents

Define your own detection categories with examples.

Overview​

Custom intents extend the guardrails system with your own detection logic. Provide a name, description, and example prompts (both positive matches and negative non-matches) to train a lightweight classifier. Once created, your intent runs on every request alongside the built-in categories.

Key Features​

  • Create custom detection intents with positive and negative examples
  • Intents are evaluated alongside built-in guardrails
  • Assign block, monitor, or redact actions per intent

How It Works​

  1. Name your intent (e.g., competitor-mentions)
  2. Describe what the intent should detect
  3. Add positive examples - prompts that should trigger the intent
  4. Add negative examples - prompts that should not trigger the intent
  5. Assign an action - block, monitor, or redact

The classifier learns from your examples and applies the configured action when a match is detected.