Disclosed weakness · no CVE recorded

LiteLLM Azure streaming errors could reveal internal server paths

LiteLLM · LiteLLM Proxy · Azure OpenAI streaming errors

First documented disclosure · Review checkpoint

Summary

An interrupted Azure stream could expose a Python traceback to an API client and to configured alert recipients.

Severity and exploitation evidence

No CVSS score recorded in this review.

Exploitation not established: The maintainer documented this exposure during stream-failure testing. The reviewed sources do not report malicious use.

Exploitation evidence source

Technical severity is not a measure of observed attacks or of risk to your deployment. Attempts, demonstrations, and successful compromise retain the qualifications in the source.

How the weakness works

The Azure adapter included the server traceback in an exception message shared with the streaming error and alert handlers.

Conditions required

An Azure stream fails during a request on an affected path. Alert exposure requires a configured exception alert destination.

Potential impact

API callers or alert readers could learn internal server paths and stack details.

Affected versions and configurations

The public fix identifies the affected implementation. The full affected version range is unspecified.

Fix and mitigation status

LiteLLM lists the fix in its development release 1.111.0-dev.4.

Practical next steps

Deploy the Azure error-handling fix and review retained exception alerts that may contain internal paths.

Operating lesson

Keep server stack details in restricted diagnostics and return a short connection error to clients.

Evidence and limits

The maintainer documents a correction. The cited sources provide no issue-specific severity rating or CVE.

Sources