When retrieval is eating your context budget and responses are getting cut off.
I have a [N]-token context window for a RAG app. Allocate budget across system prompt, retrieved docs, conversation history, and the response. Tell me what to truncate first under load and why.
Anything in square brackets is meant to be replaced. The builder turns those into fields you can type into.
Tired of re-pasting the same context into every new chat? That is what the Unimatrix MCP server is for.