48 runnable RAG and AI agent examples with quick setup

An repository provides working RAG s that are designed to be easy to run, understand, and change. It includes a PDF chatbot, codebase question answering, question answering, and agentic RAG.

Every example comes in both Python and TypeScript, using the same basic layout: setup instructions, code, and a sample environment file. The examples do not depend on shared project components or a complicated setup, so each can be cloned, installed, and run on its own.

They belong to a larger set of 48 AI agent examples covering memory, MCP, voice, , and common agent designs. The focus is on filling practical gaps left by basic demonstrations that only process a few documents and query a .

Key points

  • Examples cover questions about PDFs, codebases, and .
  • An agentic RAG example lets an agent take a more active role in retrieval.
  • Every example is available in Python and TypeScript.
  • Each example has its own setup instructions, code, and sample environment file.
  • The wider repository contains 48 examples for memory, MCP, voice, and .
Read original