These are different "agents" in LLM terms, they have separate contexts and separate training