MemgraphToolkit

This will help you get started with the Memgraph toolkit.

Tools within MemgraphToolkit are designed for the interaction with the Memgraph database.

Setup

To be able tot follow the steps below, make sure you have a running Memgraph instance on your local host. For more details on how to run Memgraph, take a look at Memgraph docs

If you want to get automated tracing from runs of individual tools, you can also set your LangSmith API key by uncommenting below:

# os.environ["LANGSMITH_API_KEY"] = getpass.getpass("Enter your LangSmith API key: ")
# os.environ["LANGSMITH_TRACING"] = "true"

Installation

This toolkit lives in the langchain-memgraph package:

%pip install -qU langchain-memgraph 

Instantiation

Now we can instantiate our toolkit:

from langchain.chat_models import init_chat_model
from langchain_memgraph import MemgraphToolkit
from langchain_memgraph.graphs.memgraph import MemgraphLangChain

db = MemgraphLangChain(url=url, username=username, password=password)

llm = init_chat_model("gpt-4o-mini", model_provider="openai")

toolkit = MemgraphToolkit(
    db=db,  # Memgraph instance
    llm=llm,  # LLM chat model for LLM operations
)

API Reference:init_chat_model

Tools

View available tools:

toolkit.get_tools()

Invocation

Tools can be individually called by passing an arguments, for QueryMemgraphTool it would be:

from langchain_memgraph.tools import QueryMemgraphTool

# Rest of the code omitted for brevity

tool.invoke({QueryMemgraphTool({"query": "MATCH (n) RETURN n LIMIT 5"})})

Use within an agent

from langgraph.prebuilt import create_react_agent

agent_executor = create_react_agent(llm, tools)

API Reference:create_react_agent

example_query = "MATCH (n) RETURN n LIMIT 1"

events = agent_executor.stream(
    {"messages": [("user", example_query)]},
    stream_mode="values",
)
for event in events:
    event["messages"][-1].pretty_print()

API reference

For more details on API visit Memgraph integration docs

Related

• Tool conceptual guide
• Tool how-to guides