Message Persistence

To learn about how to load previous messages and agent states, check out the Loading Message History and Loading Agent State pages.

To persist LangGraph messages to a database, you can use either AsyncPostgresSaver or AsyncSqliteSaver. Set up the asynchronous memory by configuring the graph within a lifespan function, as follows:

from fastapi import FastAPI
from contextlib import asynccontextmanager
from langgraph.checkpoint.postgres.aio import AsyncPostgresSaver
from copilotkit import LangGraphAGUIAgent
from ag_ui_langgraph import add_langgraph_fastapi_endpoint

graph = None
@asynccontextmanager
async def lifespan(app: FastAPI):
    async with AsyncPostgresSaver.from_conn_string(
        "postgresql://postgres:postgres@127.0.0.1:5432/postgres"
    ) as checkpointer:
        # NOTE: you need to call .setup() the first time you're using your checkpointer
        await checkpointer.setup()
        # Create an async graph
        graph = workflow.compile(checkpointer=checkpointer)
        yield
        # Create SDK with the graph

app = FastAPI(lifespan=lifespan)

add_langgraph_fastapi_endpoint(
    app=app,
    agent=LangGraphAGUIAgent(
        name="research_agent",
        description="Research agent.",
        graph=graph,
    ),
    path="/agents/research_agent"
)

To learn more about persistence in LangGraph, check out the LangGraph documentation.