🔨 In Development — This section is still being developed and may change.
Rules let you enforce organization-wide guardrails, tone, and compliance requirements across Freddy assistants and responses.
- Declarative policies that modify assistant behavior (style, prohibited topics, escalation paths).
- Stored centrally in Freddy Hub and referenced by ID in API requests.
- Evaluated alongside prompts to ensure consistent output quality.
- Create the rule in Freddy Hub with a clear name and description.
- Attach the rule ID when creating responses, assistants, or threads.
- Monitor rule analytics to confirm adoption and adjust messaging.
- Keep rules focused and composable; avoid mega-rules that are hard to maintain.
- Version rules when making breaking changes so clients can migrate gradually.
- Pair rules with tests in staging environments before promoting to production.
title: Rules badge: draft
🔨 In Development — This section is still being developed and may change.
Rules provide additional behavior modifications and can be attached to API responses to customize their behavior.
Rules allow you to define custom logic that modifies how responses are processed and returned. Each rule has a unique ID and can be referenced by responses to apply specific modifications.