langchain-mq9

LangChain tools for the mq9 AI-native async mailbox protocol.

Give your LangChain Agents a persistent, priority-aware inbox — powered by RobustMQ.

Install

pip install langchain-mq9

Quick start

from langchain_mq9 import Mq9Toolkit
from langchain.agents import AgentType, initialize_agent
from langchain_openai import ChatOpenAI

toolkit = Mq9Toolkit(server="nats://demo.robustmq.com:4222")
tools = toolkit.get_tools()

agent = initialize_agent(
    tools,
    ChatOpenAI(model="gpt-4o-mini"),
    agent=AgentType.OPENAI_FUNCTIONS,
    verbose=True,
)

agent.invoke({"input": "Create a mailbox for me to receive task results."})

Tools

CreateMailboxTool

Creates a mq9 mailbox and returns its mail_id.

SendMessageTool

Sends an async message to a mailbox.

GetMessagesTool

Retrieves pending messages from a mailbox.

Mq9Toolkit

from langchain_mq9 import Mq9Toolkit

toolkit = Mq9Toolkit(server="nats://localhost:4222")
tools = toolkit.get_tools()  # returns [CreateMailboxTool, SendMessageTool, GetMessagesTool]

Multi-Agent example

See examples/agent_example.py for a complete example of two Agents communicating asynchronously via mq9.

Why mq9?

• Store-first: messages persist until TTL expires — sender and receiver don't need to be online simultaneously
• Priority queuing: route urgent tasks ahead of background work
• No consumer state: simple, stateless protocol ideal for Agent workflows
• Built on NATS: fast, lightweight, zero broker dependencies