Interceptors & Guardrails

Interceptors and Guardrails

Interceptors are custom logic that can run before or after a request is routed, allowing you to enrich, validate, or transform requests and responses. Guardrails are a common type of interceptor used to enforce policies.

Type	Purpose	Business Value	Example Use Case/Config
Pre-request	Analyze or enrich request	Classify topic, check for risk, personalize	semantic_guardrail, toxicity_guardrail
Post-request	Analyze or modify response	Moderate output, add fallback, redact sensitive info	fallback_response

When an interceptor runs, it can inject its results into the routing context, making them available for your conditional logic.

Result Variable	Description	Example Value	Business Use
semantic_guardrail.result.topic	Detected topic from a guardrail	"billing"	Route to topic-specialized models
toxicity_guardrail.result	Toxicity score from a guardrail	0.8	Block or reroute harmful content
rate_limiter.result	Result of a rate limit check	true	Enforce usage quotas and prevent abuse

Note on Guardrails: Guardrails like semantic_guardrail and toxicity_guardrail are powerful examples of custom guardrails. Checkout Guardrails section for more details.