Sorting LiteLLM GitHub issues by difficulty to implement. Higher = harder.

{
The left issue requires diagnosing and safely fixing a concurrent state-update race in latency-aware routing, with cache consistency, async behavior, and regression testing across deployment-selection paths. The right issue is comparatively narrow request-input/header sanitization and error handling.
}
https://github.com/berriai/litellm/issues/24720 5:1 https://github.com/berriai/litellm/issues/27188