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

{
The left requires coordinating callback, tool-execution, and multi-turn orchestration across endpoint and provider-specific paths, with broader regression risk. The right is comparatively localized to identifier normalization in one API translation path, with focused compatibility tests.
}
https://github.com/berriai/litellm/issues/25191 4:1 https://github.com/berriai/litellm/issues/25757