Node type proposal: Vector Store #23
Description
Interface with vector databases for storage and retrieval. Supported backends: ChromaDB (local), Pinecone, Weaviate, Qdrant, pgvector. Operations: upsert (store embeddings with metadata), query (similarity search with top-k and filters), delete. Pairs with the Embedding node to enable full RAG pipelines: embed documents, store in vector DB, query at runtime, pass context to prompt node. Configuration: connection string, collection/index name, distance metric.