Show HN: A lightweight LLM proxy to get structured results from most LLMs

  • Looks usefull. Could you explain how it works? you have to chain it after the call from other LLM ?