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

{
The left issue is harder because it spans application startup ordering, subprocess readiness detection, failure handling, background task coordination, and deployment-level regression testing. The right issue is more localized to protocol translation state management and stream conformance, with a narrower test surface.
}
https://github.com/berriai/litellm/issues/27704 3:1 https://github.com/berriai/litellm/issues/32357