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

{
The left issue is harder because it involves security-sensitive credential handling, tracing the failure path through authentication and logging, and validating that redaction remains correct across related error paths. The right issue appears comparatively localized to defensive parsing and error translation in one provider’s streaming transformation, with narrower regression testing.
}
https://github.com/berriai/litellm/issues/30670 4:1 https://github.com/berriai/litellm/issues/33622