Enables 70-90% LLM API cost reduction by compressing conversation history via local Gemma 4 models or heuristics, featuring token counting, model routing, and pinned facts for preserving critical context.
Install from
M8ven verifies MCPs across every public registry — install directly from whichever one you prefer.
[](https://m8ven.ai/mcp/woling-dev-promptthrift-mcp-f7a8vp)