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

{
The left issue likely requires tracing and aligning database-backed model discovery across backend APIs, frontend state, and public hub configuration, with broader integration and regression risk. The right issue appears localized to exception classification logic with focused tests.
}
https://github.com/berriai/litellm/issues/19853 3:1 https://github.com/berriai/litellm/issues/28163