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

{
The right-hand task is harder because it spans provider-specific response translation, shared usage schemas, and both synchronous and streaming aggregation paths, with compatibility and regression risks. The left-hand task is narrower: correcting default scope semantics and adding focused validation and matching tests.
}
https://github.com/berriai/litellm/issues/29998 3:2 https://github.com/berriai/litellm/issues/28904