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

{
The right issue is harder because it requires changing cache-key semantics across multiple cache backends and request paths, incorporating authenticated tenant context, preserving configurable behavior and compatibility, and validating isolation under both exact and semantic matching. The left issue is comparatively localized to retention cleanup and database consistency, with bounded schema/query and migration considerations.
}
https://github.com/berriai/litellm/issues/29955 4:1 https://github.com/berriai/litellm/issues/29342