pgvector & RAG

Complete toolkit for building vector search and RAG (Retrieval-Augmented Generation) applications with PostgreSQL. Covers pgvector schema setup, embedding generation with Ollama, similarity search, full RAG pipelines, and agentic-kit integration.

When to Apply

Use this skill when:

• Setting up pgvector: Creating tables, indexes, vector storage schema
• Generating embeddings: Using Ollama to embed documents and chunks
• Similarity search: Querying vectors with cosine/L2/inner product distance
• Building RAG: Combining retrieval with LLM generation
• Ollama integration: Local LLM inference, model management, streaming
• Agentic-kit RAG: Wiring RAG into agentic-kit chat applications

Architecture

Document → Chunking → Embedding → pgvector Storage
                                        ↓