Skip to content
Browse the docs

Frameworks and agents · Published 2026-10-21 · Updated 2026-10-21 · AgentiSend

Send email from LangChain for Python

AgentiSend mail from LangChain for Python goes through one @tool invoked twice, and the second call returns a refusal as code and fix.

To send email from LangChain for Python on AgentiSend, register one @tool whose invoke posts to POST /emails with the agent's own key and returns a refusal as code and fix.

Install

pip install langchain-core

The agentisend package is standard library only, so pip install agentisend adds no other dependency.

Environment

VariableRequiredWhat it is
AGENTISEND_API_KEYyesThe agent's key, with sending_access and a budget on it. Not your own.
MAIL_FROMyesThe From address, on a domain you have verified.
AGENTISEND_BASE_URLnoDefaults to https://api.agentisend.com.

The tool

"""LangChain — one tool that posts to POST /emails, and a scripted agent loop.

The loop is the two calls a model would make. The first sends. The second is
a different message, so a key whose budget is one comes back as
agent_budget_exceeded. No model is called.

Install both packages with `pip install agentisend langchain-core`. The suite
runs this file against the SDK checkout; a fresh project uses the released package.

AGENTISEND_API_KEY is the agent's key, with sending_access and a budget.
MAIL_FROM is an address on a domain you have verified.
AGENTISEND_BASE_URL is optional.
"""

from __future__ import annotations

import json
import os

from langchain_core.tools import tool

from agentisend import AgentiSend, AgentiSendError, idempotency_key

agentisend = AgentiSend()
TO = "langchain-py@example.com"
UPDATE = {
    "to": TO,
    "subject": "Ticket 4182 — we are looking into it",
    "text": "Someone from support will reply within the hour.",
    "purpose": "ticket-4182-update",
}
CLOSED = {
    "to": TO,
    "subject": "Ticket 4182 — closed",
    "text": "The ticket is closed.",
    "purpose": "ticket-4182-closed",
}


@tool
def send_email(to: str, subject: str, text: str, purpose: str) -> dict:
    """Send one email to a person who asked for it.

    Returns the message id, or a refusal with a code and a fix. When the fix
    says a person decides, stop and report it: calling again will not change
    the answer.
    """
    try:
        sent = agentisend.send_email(
            {"from": os.environ["MAIL_FROM"], "to": to, "subject": subject, "text": text},
            idempotency=idempotency_key(purpose, to),
        )
    except AgentiSendError as err:
        return {"sent": False, "code": err.code, "fix": err.fix}
    return {"sent": True, "id": sent["id"]}


def run() -> None:
    """Two tool calls, then stop when the second is refused."""
    for call in (UPDATE, CLOSED):
        result = send_email.invoke(call)
        print(json.dumps(result))
        if result.get("sent") is False:
            return


if __name__ == "__main__":
    run()
examples/langchain-python/tool.py, run by the test suite against a local AgentiSend server

The decorator is @tool from langchain_core.tools, with no display name argument. invoke takes one dict whose keys are to, subject, text, and purpose, and it returns a dict the agent appends as the tool message. The file calls invoke twice and does not call a model. The second dict is a refusal:

{
  "error": {
    "code": "agent_budget_exceeded",
    "message": "This key has used 1 of its 1 recipients for the current monthly period. Each To, CC and BCC recipient counts as one email.",
    "fix": "Wait for the period to reset — get_agent_budget and GET /limits/keys/:id both say when. Raising a budget is a person’s decision, made in the console; a key cannot raise its own.",
    "docs_url": "https://agentisend.com/what-does-agent-budget-exceeded-mean#agent_budget_exceeded",
    "retryable": false
  }
}

idempotency_key(purpose, to) is the header. Reusing one purpose for two different messages replays the first. The JavaScript tool, built with tool() from @langchain/core/tools, is Send email from a LangChain.js agent. The key's ceiling is in An AI agent with a budget and a loop guard. This key cannot list arrivals. GET /emails/receiving needs full access, as Receiving and replying describes.

Next