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

{
The right issue requires tracing model registration and provider-specific configuration across persistence, loading, request translation, and streaming behavior, with regression coverage across multiple configuration paths. The left issue is primarily a documentation correction based on already-fixed behavior, requiring minimal engineering work.
}
https://github.com/berriai/litellm/issues/28044 5:1 https://github.com/berriai/litellm/issues/34727