Langgraph - Add memory to graph using create_openai_functions_agent as an Agent node

[lironezra]

Checked other resources

• I added a very descriptive title to this question.
• I searched the LangChain documentation with the integrated search.
• I used the GitHub search to find a similar question and didn't find it.

Commit to Help

• I commit to help with one of those options 👆

Example Code

#This is my graph

checkpointer = MongoDBSaver(MongoClient(MONGO_URI), "chatbot", "liron_test")

# Define a new graph
workflow = StateGraph(AgentState)

# Define the two nodes we will cycle between
workflow.add_node("agent", run_agent)
workflow.add_node("action", execute_tools)

# Set the entrypoint as `agent`
# This means that this node is the first one called
workflow.add_edge(START, "agent")

# We now add a conditional edge
workflow.add_conditional_edges(
    # First, we define the start node. We use `agent`.
    # This means these are the edges taken after the `agent` node is called.
    "agent",
    # Next, we pass in the function that will determine which node is called next.
    should_continue,
    # Finally we pass in a mapping.
    # The keys are strings, and the values are other nodes.
    # END is a special node marking that the graph should finish.
    # What will happen is we will call `should_continue`, and then the output of that
    # will be matched against the keys in this mapping.
    # Based on which one it matches, that node will then be called.
    {
        # If `tools`, then we call the tool node.
        "continue": "action",
        # Otherwise we finish.
        "end": END,
    },
)

# We now add a normal edge from `tools` to `agent`.
# This means that after `tools` is called, `agent` node is called next.
workflow.add_edge("action", "agent")

# Finally, we compile it!
# This compiles it into a LangChain Runnable,
# meaning you can use it as you would any other runnable
app = workflow.compile(checkpointer=checkpointer)

Description

Hi all,

I am trying to build an agent with tools using Langgraph with openai_function_agent, I am using the exact code from here: https://github.com/langchain-ai/langgraph/blob/main/examples/agent_executor/base.ipynb

I cannot add memory to the graph and agent. Can anyone help me with this? I am using MongoDBSaver from the documentation, and I see that there are values in MongoDB for every message the agent got, but when I am asking second question the chat_history seems to be empty.

Maybe I cannot add memory in this architecture?

System Info

System Information

OS: Windows
OS Version: 10.0.19045
Python Version: 3.11.2 (tags/v3.11.2:878ead1, Feb 7 2023, 16:38:35) [MSC v.1934 64 bit (AMD64)]

Package Information

langchain_core: 0.2.18
langchain: 0.2.1
langchain_community: 0.2.1
langsmith: 0.1.85
langchain_cohere: 0.1.5
langchain_groq: 0.1.5
langchain_openai: 0.1.8
langchain_pinecone: 0.1.1
langchain_text_splitters: 0.2.0
langgraph: 0.1.8

Packages not installed (Not Necessarily a Problem)

The following packages were not found:

langserve

[vbarda]

@lironezra i don't see any conceptual issues with your example. could you please try setting up your graphusing this example https://langchain-ai.github.io/langgraph/#example or a prebuilt ReAct agent and let me know if you're still running into issues? the only difference would be in using prebuilt ToolNode for handling tool calling

[lironezra]

Thanks for the reply @vbarda, use the prebuilt ReAct agent is not an option for us, we want to build the agent like I did with the create_openai_functions_agent.

Here is how I test the graph:
`inputs = {"input": [HumanMessage(content="Hi, I am Liron")], "chat_history": []}
config = {"configurable": {"thread_id": "1"}}

for s in app.stream(inputs, config):
print(list(s.values())[0])
print("----")
# print(app.get_state(config))

inputs = {"input": [HumanMessage(content="What is my name?")]}

for s in app.stream(inputs, config):
print(list(s.values())[0])
print("----")`

The results I got:

1. {'agent_outcome': AgentFinish(return_values={'output': 'Hello Liron! How can I assist you today?'}, log='Hello Liron! How can I assist you today?')}

2. {'agent_outcome': AgentFinish(return_values={'output': "I'm sorry, but as an assistant, I don't have access to personal data about individuals unless it has been shared with me in the course of our conversation. I'm here to provide assistance with your queries. How can I assist you today?"}, log="I'm sorry, but as an assistant, I don't have access to personal data about individuals unless it has been shared with me in the course of our conversation. I'm here to provide assistance with your queries. How can I assist you today?")}

Like you can see, the graph don't save the chat_history on the state. When I debug the code on run_agent function that get the state as a parameter the chat_history is always equal to [].

Any Idea how to solve it? @vbarda