LangChain + Lumbox
Give a LangChain agent its own inbox, OTP extraction and email send/receive in one call.
PythonLangChain is the most widely used framework for building LLM agents. Lumbox ships a first-class LangChain tool factory: create_langchain_tools() returns ready-to-bind tools so your agent can create an inbox, wait for an OTP, and read, send or reply to email without any glue code.
from lumbox import create_langchain_tools
from langchain.agents import create_react_agent
from langchain_openai import ChatOpenAI
# One call: agent-owned inbox, OTP extraction, send/receive, all as tools
tools = create_langchain_tools(api_key="ak_...")
agent = create_react_agent(ChatOpenAI(model="gpt-4o"), tools)
agent.invoke({
"messages": "Sign up at example.com and confirm the verification email"
})How it works
- 1Install the LangChain extra: pip install "lumbox[langchain]".
- 2Call create_langchain_tools(api_key) to get ready-to-bind LangChain tools.
- 3Pass the tools to your LangChain agent (create_react_agent or AgentExecutor).
- 4The agent can now create an inbox, wait for an OTP, and send or read email on its own.
FAQ
How do I give a LangChain agent an email inbox?
Install lumbox[langchain] and call create_langchain_tools(api_key='ak_...'). It returns LangChain tools for creating an inbox, waiting for an OTP, and sending or reading email. Bind them to your agent and it can handle email autonomously.
Can a LangChain agent read OTP verification codes?
Yes. Lumbox's get_otp / wait_for_email tools long-poll until the email arrives and return the extracted OTP code as JSON, so the LangChain agent never has to parse raw email or run a polling loop.