Mastra + MemorySync

Overview

Install memorysync-mastra and add the processors to any Agent: recalled context is injected as a system message before each generate or stream call and the exchange persists verbatim after it. Identity resolves per call from four sources (static config, your own resolver, RequestContext, memory.resource), a zero-config wrapper merges the processors into existing configs, and five structured tools give agents explicit memory. Runs on current @mastra/core 1.x.

Setup status and requirements

Supported version

memorysync-mastra 1.0.0 (npm)

Last setup review

2026-08-23

Permissions

A MemorySync API key with read and write scopes.

Limits

Requires @mastra/core 1.42+ (Node 20+ for the package; current @mastra/core itself requires Node 22+). JavaScript/TypeScript only — Python agents use the LangChain, LangGraph or CrewAI integrations.

Capabilities

• Native inputProcessors/outputProcessors — no model wrapping required
• Four identity sources: static, resolver, RequestContext, memory.resource
• No resolved user means no write — a wrong-scope persist is refused loudly
• Streaming-safe: persists once on finish; a failed stream stores nothing
• Recall modes: query, profile, or full
• Five never-throw agent tools with a read-only mode
• CI-tested against a real Agent on the latest @mastra/core 1.x

Quick Start

import { Agent } from "@mastra/core/agent";
import { openai } from "@ai-sdk/openai";
import { withMemorySync } from "memorysync-mastra";

const agent = new Agent(withMemorySync(
  {
    id: "assistant",
    name: "Assistant",
    instructions: "You are a helpful assistant.",
    model: openai("gpt-4o-mini"),
  },
  { userId: "customer-7" },
));

const { text } = await agent.generate("Book my usual trip.");

Use Cases

Explore More