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

{
35543 requires tracing asynchronous proxy request handling, preserving response semantics across call types, and adding integration coverage; 28071 is a localized tokenization compatibility fix with comparatively limited scope.
}
https://github.com/berriai/litellm/issues/35543 3:1 https://github.com/berriai/litellm/issues/28071