Skip to main content
Glama
HamzaFayaz

whatsapp-mcp-assistant

by HamzaFayaz

WhatsApp MCP Assistant

whatsapp-mcp-assistant — MCP server that exposes a real WhatsApp account to Cursor through safe, approval-gated tools and a local bridge.

This project exposes a real WhatsApp account to Cursor through an MCP server. Cursor is the agent interface; this repo only provides MCP tools and a local WhatsApp bridge.

V1 uses:

  • Python FastMCP server for Cursor.

  • Local Go whatsmeow HTTP bridge for WhatsApp Web login and messaging.

  • Real per-user WhatsApp sessions stored locally under sessions/.

  • No mock WhatsApp data.

Tools

  • get_connection_status()

  • get_login_qr()

  • logout_whatsapp(confirm)

  • get_recent_chats(limit=10, unread_only=false, include_groups=true)

  • search_contact(query, include_groups=true, limit=5)

  • get_messages_from_contact(chat_id, limit=20, include_outgoing=true)

  • draft_reply(chat_id, instruction, recent_messages=[], tone="natural")

  • send_whatsapp_message(chat_id, message, approved)

send_whatsapp_message rejects every request unless approved=true.

Related MCP server: @botozap/mcp

Setup

Install Python dependencies:

python -m pip install -r requirements.txt

Install Go, then install Go bridge dependencies:

cd go-whatsapp-bridge
go mod tidy

Start the WhatsApp bridge:

go run .

Start Cursor with this MCP server config shape:

{
  "mcpServers": {
    "whatsapp-assistant": {
      "command": "python",
      "args": complete path like ["/AI/MCP Server/server.py"]
    }
  }
}

QR Login Flow

  1. Start the Go bridge.

  2. Ask Cursor: Check WhatsApp connection.

  3. If disconnected, ask Cursor: Get WhatsApp login QR.

  4. Scan the returned QR string using WhatsApp > Linked Devices.

  5. Ask Cursor to check connection again.

Every user must run the server locally and scan their own WhatsApp QR. Do not share sessions/, .env, QR codes, tokens, or connected server instances.

Demo Flow

Use a test WhatsApp number.

  1. Send a WhatsApp message from another phone to the connected test account.

  2. Ask Cursor: Show my recent WhatsApp chats.

  3. Ask Cursor: What did Bilal message me today?

  4. Ask Cursor: Draft a reply saying I will come at 6.

  5. After reviewing the draft, say: Okay, send it.

Current Bridge Limitation

The local bridge records real messages it receives while the bridge is running and keeps them in sessions/message_cache.json. WhatsApp Web libraries do not reliably provide full historical chat export through this simple bridge, so test by sending messages after the bridge is connected.

Related MCP Connectors

  • Your own WhatsApp as an MCP server: read, search and send from any MCP client.

  • WhatsMCP connects Claude and other MCP-compatible AI agents directly to WhatsApp. Send and receive text, images, documents, and voice notes; manage groups (create, add/remove members, promote admins); look up contacts and profiles; follow channels; and read call and message history — all through a standard MCP interface. For voice use cases, WhatsMCP offers SIP-based calling plans (inbound-only, or full inbound/outbound) so AI voice agents can answer and place WhatsApp calls, plus low-latency WebSocket integrations with voice agent providers like ElevenLabs. Multiple WhatsApp accounts can be paired and managed per workspace, with webhook support for real-time inbound message delivery to your own infrastructure.

  • Let Claude or ChatGPT search, read and send your WhatsApp messages over MCP. OAuth sign-in.

  • Drive WhatsApp from any MCP client: pair devices, send text and media, manage contacts and groups.

Related MCP Servers

  • A
    license
    Not graded
    quality
    A
    maintenance
    A self-hosted WhatsApp bridge that exposes a stdio MCP server with ~20 tools for reading conversations, sending messages, managing groups, contacts, and aliases, enabling AI agents to operate WhatsApp directly.
    2
    MIT
  • F
    license
    Not graded
    quality
    B
    maintenance
    MCP server that lets Claude read and send WhatsApp messages using a genuine WhatsApp Web client behind a code-enforced human-behavior middleware for safety.
    -
  • A
    license
    Not graded
    quality
    A
    maintenance
    Enables AI agents to interact with WhatsApp through a safety-first MCP server, with tools for sending messages, media, searching chats, and managing drafts, all governed by default-deny allowlists and non-overridable rate limits.
    Apache 2.0