resk-caching

Smart caching system for LLM responses - reduce API costs by avoiding redundant calls