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

{
#26611 is harder because it requires diagnosing and reconciling multiple observability paths, aggregation semantics, and deployment-level effects before a reliable fix can be defined. #34917 points to a bounded request-reading defect with a clearer code location and a comparatively focused regression test and implementation scope.
}
https://github.com/berriai/litellm/issues/26611 3:1 https://github.com/berriai/litellm/issues/34917