multi-intent
The pre-LLM reflex layer for AI agents. library with a live training server, microsecond latency, learns from LLM corrections — System 1 to your LLM's System 2.
