LangGraph Memory Example

A LangGraph JavaScript agent with AgentCore Memory for conversation persistence.

Features

• Conversation Memory: Persist and recall conversation history across invocations
• Session Management: Track conversations by session ID
• Memory Expiration: Configurable TTL (30 days in this example)
• LangGraph JS: ReAct agent with memory-enhanced context

How It Works

1. On each invocation, the agent retrieves previous conversation history from AgentCore Memory
2. History is injected as system context for the LLM
3. After generating a response, the conversation turn is stored in memory
4. Subsequent invocations in the same session recall previous context

Quick Start

Deploy

npm install
sls deploy

Test

RUNTIME_ARN=<your-runtime-arn> node test-invoke.js

Remove

sls remove

Related Examples

• ../python/langgraph-memory - Python version