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

{
The left issue is harder because it involves diagnosing streaming semantics across LiteLLM’s Responses API and Anthropic translation layer, with provider-specific behavior and regression testing across streaming formats. The right issue is comparatively contained: a workflow utility needs more robust pagination parsing plus focused tests.
}
https://github.com/berriai/litellm/issues/20097 4:1 https://github.com/berriai/litellm/issues/32217