Summary

This PR adds a new tool integration: llama-index-tools-commune, which gives LlamaIndex agents a real email inbox and SMS capability via Commune.

What is Commune?

Commune is email and SMS infrastructure built specifically for AI agents. It provides:

• A real, deliverable inbox that agents can read and write to
• Outbound email and SMS sending with high deliverability
• A simple Python SDK (commune-mail on PyPI)
• Per-message credit billing with no monthly minimums

This is useful for agentic workflows where the agent needs to communicate with humans or other systems through standard channels — customer support triage, automated notifications, multi-agent coordination, and so on.

Real use case: customer support email routing

import os
from llama_index.core.agent import ReActAgent
from llama_index.llms.openai import OpenAI
from llama_index.tools.commune import CommuneToolSpec

tools = CommuneToolSpec(api_key=os.environ["COMMUNE_API_KEY"])

agent = ReActAgent.from_tools(
    tools.to_tool_list(),
    llm=OpenAI(model="gpt-4o"),
    verbose=True,
    system_prompt=(
        "You are a customer support triage agent. Check the inbox for unread "
        "emails, categorize each request (billing, technical, general), and "
        "route them by forwarding to the appropriate team email address with "
        "a brief summary of the issue."
    ),
)

agent.chat("Process any unread support emails in the inbox.")

The agent will call load_inbox(unread_only=True), read each email with get_email, then call send_email to route each one — all without any custom tool wiring.

Spec functions

Package details

• PyPI: pip install llama-index-tools-commune (will be published on merge)
• Upstream SDK: commune-mail on PyPI
• Python: >=3.9
• Depends on: llama-index-core>=0.11.0, commune-mail>=0.2.0
• Author: @shanjai-raj

Files added

llama-index-integrations/tools/llama-index-tools-commune/
├── pyproject.toml
├── README.md
└── llama_index/
    ├── __init__.py
    └── tools/
        ├── __init__.py
        └── commune/
            ├── __init__.py
            └── base.py          ← CommuneToolSpec implementation

Checklist

• Package follows the existing llama-index integration pattern (BaseToolSpec)
• All spec_functions have rich docstrings for LLM consumption
• Error handling returns helpful strings rather than raising exceptions
• README includes installation, setup, and three worked examples
• pyproject.toml includes [tool.llamahub] metadata