BerriAI/litellm · error · ValueError

IBM Guardrails auth token is required. Set IBM_GUARDRAILS_AU

Error message

IBM Guardrails auth token is required. Set IBM_GUARDRAILS_AUTH_TOKEN environment variable or pass auth_token parameter.

What it means

Error "IBM Guardrails auth token is required. Set IBM_GUARDRAILS_AUTH_TOKEN environment variable or pass auth_token parameter." thrown in BerriAI/litellm.

Source

Thrown at litellm/proxy/guardrails/guardrail_hooks/ibm_guardrails/ibm_detector.py:58

        base_url: str | None = None,
        detector_id: str | None = None,
        is_detector_server: bool = True,
        detector_params: dict[str, Any] | None = None,
        extra_headers: dict[str, str] | None = None,
        score_threshold: float | None = None,
        block_on_detection: bool = True,
        verify_ssl: bool = True,
        **kwargs,
    ):
        self.async_handler = get_async_httpx_client(
            llm_provider=httpxSpecialProvider.GuardrailCallback,
            params={"ssl_verify": verify_ssl},
        )

        # Set API configuration
        self.auth_token = auth_token or os.getenv("IBM_GUARDRAILS_AUTH_TOKEN")
        if not self.auth_token:
            raise ValueError(
                "IBM Guardrails auth token is required. Set IBM_GUARDRAILS_AUTH_TOKEN environment variable or pass auth_token parameter."
            )

        self.base_url = base_url
        if not self.base_url:
            raise ValueError("IBM Guardrails base_url is required. Pass base_url parameter.")

        self.detector_id = detector_id
        if not self.detector_id:
            raise ValueError("IBM Guardrails detector_id is required. Pass detector_id parameter.")

        self.is_detector_server = is_detector_server
        self.detector_params = detector_params or {}
        self.extra_headers = extra_headers or {}
        self.score_threshold = score_threshold
        self.block_on_detection = block_on_detection
        self.verify_ssl = verify_ssl

View on GitHub (pinned to 77b7c6c40c)

Solutions

1. Export IBM_GUARDRAILS_AUTH_TOKEN in the proxy environment.
2. Alternatively pass auth_token in the guardrail's litellm_params.
3. Verify the token is valid and has access to the IBM detector service.

When it happens

Trigger: Thrown at litellm/proxy/guardrails/guardrail_hooks/ibm_guardrails/ibm_detector.py:58 when the library encounters an invalid state.

Common situations: See trigger scenarios.

AI-assisted analysis of BerriAI/litellm@77b7c6c40c (2026-08-18). Data as JSON: /api/errors/628afcd446273c01. Report an issue: GitHub.