A tiny proxy that sits in front of any OpenAI-compatible LLM backend (Ollama, LiteLLM, etc.) and forces conversation-history compaction before the context window fills up, so local-model coding CLIs don't hard-die from context overflow.
python mc prox llm openai-ap ai-tool local-llm ollam coding-assistan litell context-windo context-managemen
-
Updated
Aug 29, 2026 - Python