If you like reading LLM output, just talk directly to an LLM. Problem solved.