AI NVIDIA NeMo Guardrail Policy

Check requests with a stored guardrail configuration

Configure the AI NVIDIA NeMo Guardrail policy to check requests against a guardrail configuration stored on the NeMo server. The policy references the stored configuration by ID with config.guardrails.config_id. The default guarding_mode is INPUT, so the policy checks requests only and doesn’t check responses. Set config.auth.prefix to an empty string when the NeMo model uses engine: openai, because NeMo passes the header value to its OpenAI client as the API key.

Prerequisites

Environment variables

  • OPENAI_API_KEY: The OpenAI API key that the NeMo Guardrails microservice uses to evaluate your rails.

Set up the policy

Attach the policy

Help us make these docs great!

Kong Developer docs are open source. If you find these useful and want to make them better, contribute today!