RAG in 15 lines

createRAG with a file vector store. Working retrieval in under a screen of code.

RAG does not require a vector database, a cluster, or a PhD. Start here, swap pieces later.

import { createRAG } from '@agentskit/rag'
import { openaiEmbedder } from '@agentskit/adapters'
import { fileVectorMemory } from '@agentskit/memory'

const rag = createRAG({
  embed: openaiEmbedder({ apiKey: process.env.OPENAI_API_KEY! }),
  store: fileVectorMemory({ path: './vectors.json' }),
})

await rag.ingest([
  { id: 'streams', content: 'AgentsKit adapters emit streaming chunks.' },
])

const context = await rag.retrieve({
  query: 'how do streams work?',
  messages: [],
})

Tip

Swap fileVectorMemory for another VectorMemory implementation without changing the RAG pipeline.

Note

ingest embeds and stores every supplied chunk. Idempotency and replacement semantics belong to the injected vector store.

Explore nearby

Peer
Cookbook
Copy-paste recipes for the things every agent app needs. Each recipe stands on its own.
Peer
Streaming chat
useChat + abort + back-pressure. The minimum viable streaming chat, production-ready.
Peer
Tools + memory together
The "chat with state and actions" loop — persistent memory plus tool execution.

✎ Edit this page on GitHub·Found a problem? Open an issue →·How to contribute →