Whatsapp MCP
A WhatsApp MCP (Model Context Protocol) server that connects your WhatsApp account with AI agents, enabling automated messaging, contact search, and chat interaction through LLMs.
WhatsApp MCP + CLI
A WhatsApp integration that works in two ways:
- MCP Server — lets AI assistants like Claude control WhatsApp directly
- CLI Tool — send/read WhatsApp messages from your terminal
Built on whatsapp-web.js.
Features
- Send WhatsApp messages by contact name or phone number
- Read recent messages from any chat
- List all recent chats with previews
- Search contacts by name
- Works as an MCP server (Claude, etc.) or standalone CLI
Setup
1. Install dependencies
npm install
2. Authenticate WhatsApp
node index.js auth
Scan the QR code with your WhatsApp mobile app. Your session is saved locally — you only need to do this once.
MCP Server (for Claude / AI assistants)
The MCP server exposes 4 tools over stdio:
| Tool | Description |
|---|---|
send_message | Send a message to a contact |
read_messages | Read recent messages from a chat |
list_chats | List recent chats with previews |
search_contacts | Search contacts by name |
Configure in Claude Code
Add to your .mcp.json or Claude Code MCP settings:
{
"mcpServers": {
"whatsapp": {
"command": "node",
"args": ["C:/path/to/whatsapp-cli/mcp-server.mjs"]
}
}
}
Run manually
node mcp-server.mjs
CLI Usage
Send a message
node index.js send --to "John Doe" --message "Hey!"
# or by phone number
node index.js send --to 919876543210 --message "Hey!"
Read messages
node index.js read --from "John Doe" --count 20
List chats
node index.js list-chats --count 10
Use Cases
- AI-powered WhatsApp assistant — plug into Claude or any MCP-compatible AI to read and send messages hands-free
- WhatsApp automation — build bots, auto-responders, or notification systems
- Terminal messaging — send/read WhatsApp messages without opening your phone
- Business notifications — send automated WhatsApp alerts from scripts or cron jobs
Project Structure
├── index.js # CLI entry point
├── mcp-server.mjs # MCP server (for AI assistants)
├── src/
│ ├── client.js # WhatsApp client setup
│ └── commands/
│ ├── auth.js # QR code authentication
│ ├── send.js # Send message command
│ ├── read.js # Read messages command
│ └── list-chats.js # List chats command
└── package.json
Requirements
- Node.js 18+
- A WhatsApp account (personal or business)
License
MIT
İlgili Sunucular
MCP Telegram Server
An MCP server for interacting with Telegram. It allows searching, sending messages, and managing chats using the Telegram API.
Confluence
Interact with Confluence to execute CQL queries, retrieve page content, and update pages.
Sendblue
Send iMessage and SMS messages using the Sendblue API.
Slack MCP Server
A server for integrating with Slack, enabling communication and automation within your workspace.
vv-mcp
A text-to-speech (TTS) server using the VOICEVOX engine. Requires a running VOICEVOX instance and is currently macOS only.
LinkedIn Post Generator
Generates professional LinkedIn post drafts from YouTube videos by summarizing their content.
Tangerine
An MCP server for Tangerine, the Convo AI assistant backend.
MCP Meme Sticky
Create AI-generated memes and convert them into stickers for Telegram and WhatsApp.
Wizzypedia MCP Server
Interact with Wizzypedia through the MediaWiki API, supporting both read-only and authenticated operations.
Sooshie
Schedule and publish social media posts across 13+ platforms, manage your content calendar, and generate AI-powered captions through any MCP-compatible AI assistant.