@agentskit/memory

@agentskit/memory

Persistent and vector memory backends for AgentsKit.

Install

npm install @agentskit/memory

Then install the backend(s) you need:

npm install better-sqlite3  # for sqliteChatMemory
npm install vectra           # for fileVectorMemory (default, pure JS)
npm install redis            # for redisChatMemory / redisVectorMemory

Backends

Factory	Contract	Underlying lib	Native?
sqliteChatMemory({ path })	ChatMemory	better-sqlite3	Yes
redisChatMemory({ url })	ChatMemory	redis	No
redisVectorMemory({ url })	VectorMemory	redis + RediSearch	No
fileVectorMemory({ path })	VectorMemory	vectra	No (pure JS)

Quick example

import { sqliteChatMemory, fileVectorMemory } from '@agentskit/memory'

// Chat persistence
const chatMemory = sqliteChatMemory({ path: './chat.db' })

// Vector search
const vectorMemory = fileVectorMemory({ path: './vectors' })
await vectorMemory.store([{
  id: 'doc-1',
  content: 'AgentsKit is awesome',
  embedding: [0.1, 0.2, 0.3, ...],
}])
const results = await vectorMemory.search([0.1, 0.2, 0.3, ...], { topK: 5 })

Custom vector store

Bring your own vector backend (LanceDB, usearch, Pinecone, etc.):

import type { VectorStore } from '@agentskit/memory'

const myStore: VectorStore = {
  async upsert(docs) { /* your logic */ },
  async query(vector, topK) { /* your logic */ },
  async delete(ids) { /* your logic */ },
}

const memory = fileVectorMemory({ path: './vectors', store: myStore })

Docs

Full documentation