langgraph

LangGraph Integration Guide

Use mcp-memory-service as the persistent memory backend for LangGraph agents and StateGraphs.

Key Differentiator: Cross-Graph Shared Memory

LangGraph's built-in MemorySaver is graph-local and ephemeral — memory is lost between runs and cannot be shared between different StateGraphs.

mcp-memory-service provides persistent shared memory across all graphs, runs, and even separate processes:

Graph A (Researcher)  ──┐
                        ├──→ mcp-memory-service ←──→ All graphs share one store
Graph B (Writer)     ──┘
Graph C (Reviewer)  ──┘

Setup

pip install mcp-memory-service httpx
MCP_ALLOW_ANONYMOUS_ACCESS=true memory server --http

Memory Tools for ReAct Agents

Define memory as @tool functions for use in a ReAct agent:

import httpx
from langchain_core.tools import tool

MEMORY_URL = "http://localhost:8000"

@tool
async def search_memory(query: str) -> str:
    """Search long-term memory for relevant context."""
    async with httpx.AsyncClient() as client:
        response = await client.post(
            f"{MEMORY_URL}/api/memories/search",
            json={"query": query, "limit": 5},
        )
        memories = response.json()["memories"]
        if not memories:
            return "No relevant memories found."
        return "\n".join(f"- {m['content']}" for m in memories)

@tool
async def store_memory(content: str, tags: list[str] = None) -> str:
    """Store a new memory for future retrieval."""
    async with httpx.AsyncClient() as client:
        response = await client.post(
            f"{MEMORY_URL}/api/memories",
            json={"content": content, "tags": tags or []},
        )
        result = response.json()
        return f"Stored memory: {result.get('content_hash', 'unknown')}"

Use in a ReAct agent:

from langgraph.prebuilt import create_react_agent
from langchain_anthropic import ChatAnthropic

llm = ChatAnthropic(model="claude-sonnet-4-6")

agent = create_react_agent(
    llm,
    tools=[search_memory, store_memory],
    state_modifier="You have access to long-term memory. Search memory before answering. Store important findings.",
)