Chimerai CLI command wires up a full RAG system with one line of code

A single command — 'npx chimerai add rag' — scaffolds a complete Retrieval-Augmented Generation system for Next.js projects, automatically adding the required AI chat module if not already present. The generated backend is a FastAPI service that handles document ingestion, chunking via RecursiveCharacterTextSplitter, and vector storage using a FAISS flat index on 1536-dimensional embeddings. Retrieval works by injecting the top-k matched document chunks into the LLM system prompt, with responses returning both the generated answer and a metadata block listing source documents and similarity scores. The Next.js frontend communicates with the Python backend through proxy routes, while a single 'chimerai dev' command starts both servers. The setup includes graceful fallback handling for environments where FAISS or NumPy cannot be installed, such as Python 3.13 or certain Windows configurations.