there is a way i can tell to the ContextChatEngine use only the documents / nodes of the current index? because the problem now is that if i use an index example and example has no info about cars the RAG will always reply
Really the only way is controlling with prompts 😅 Or maybe doing some "test retrieval" with the user message, and if the similarity score is too low, don't call the chat engine?
maybe i will study something like more granually index separation and an agent that can understand what index to use, the only problem in my showcase with the agent is that it will reply to any question not only the question related to the context
because not all the times tha agent make a function call (i try ContextAgumentedOpenAiAgent not the normal one with param that force function calling hahha)
