cant reach the state with checkpoint_id

[us]

i am trying to using checkpoint_id and thread_id to reach history in specific point.

let say history is following:

checkpoint 1
human 1
checkpoint 2
ai 1
checkpoint 3
human 2
checkpoint 4
ai 2
checkpoint 5
(i can be wrong about the checkpoints order according to langchain but its not the part of issue, its just basic graph)

so what i expect with from following

async def astream_graph(state, config):
    try:
        async with AsyncMongoDBSaver.from_conn_string(os.getenv("MONGODB_URI"),
                                                      db_name=os.getenv("MONGODB_DB_NAME")) as checkpointer:
            graph = build_graph(checkpointer)
            # print("graph.get_state_history(config)", [state for state in graph.get_state_history(config)])
            async for event in graph.astream(state, config, stream_mode=["messages", "debug", "custom"]):
                yield event
    except Exception as e:
        print(f"Error in streaming: {str(e)}")
        raise

state=... (thats ok)
config = {"thread_id": "abc", "checkpoint_id":"checkpoint 3"}
astream_graph(state, config)

i am expecting to return me history until checkpoint 3 like:

checkpoint 1
human 1
checkpoint 2
ai 1
checkpoint 3

but i am just getting full history, so it looks like checkpoint_id is not works like expected.
I am talking about that replay thing https://langchain-ai.github.io/langgraph/concepts/persistence/#replay

here my env:
python3.12

packages:
langchain==0.3.21
langgraph==0.3.20
langgraph-checkpoint-mongodb==0.1.3
langchain-core==0.3.48
langchain-openai==0.3.7

[us]

@blink1073 any updates?

[blink1073]

Hi @us, thanks for the report! I created https://jira.mongodb.org/browse/INTPYTHON-623 and assigned it to @caseyclements. I don't have a good working knowledge of the LangGraph side of things. He's at the LangChain Interrupt conference this week, so please bear with us. 😄