WhatsApp MCP
A Model Context Protocol server that enables interaction with WhatsApp through local stdio or remote HTTP/SSE connections. It allows users to send messages, manage groups, and access chat history using natural language.
README
<<<<<<< HEAD
whatsapp_MCP
=======
WhatsApp MCP — Remote HTTP Server
A Model Context Protocol (MCP) server for WhatsApp built with whatsapp-web.js.
Supports two modes:
| Mode | File | Use case |
|---|---|---|
| stdio | src/index.ts |
Claude Desktop (local) |
| Remote HTTP/SSE | src/remote.ts |
Claude.ai "Remote MCP URL" / any HTTP client |
Quick Start
1. Install dependencies
npm install
2. Build TypeScript
npm run build
3. Run the remote server
npm run start:remote
# or in dev mode (no build needed):
npm run dev:remote
4. Scan the QR code
Open http://localhost:3000/qr in your browser and scan with WhatsApp → Linked Devices.
Endpoints
| Method | Path | Description |
|---|---|---|
| GET | / or /health |
Health check + status |
| GET | /qr |
QR code page (browser) or ?format=json for raw data |
| GET | /tools |
List all available MCP tools |
| GET | /tools/status |
WhatsApp connection status |
| POST | /tools/send-message |
Send a WhatsApp message |
| GET | /tools/chats |
List recent chats |
| GET | /tools/contacts |
List contacts |
| GET | /tools/messages/:chatId |
Fetch messages from a chat |
| POST | /tools/invoke |
Generic tool invocation |
| GET | /sse |
MCP SSE endpoint (for Claude Remote MCP) |
| POST | /messages |
MCP POST endpoint (used with /sse) |
REST API Examples
Check status
curl http://localhost:3000/tools/status
Send a message
curl -X POST http://localhost:3000/tools/send-message \
-H "Content-Type: application/json" \
-d '{"to": "919876543210", "message": "Hello from MCP!"}'
Get recent chats
curl "http://localhost:3000/tools/chats?limit=10"
Get messages from a chat
curl "http://localhost:3000/tools/messages/919876543210?limit=20"
Generic tool invocation
curl -X POST http://localhost:3000/tools/invoke \
-H "Content-Type: application/json" \
-d '{"tool": "whatsapp_get_groups", "args": {}}'
Optional: Bearer Token Auth
Set the API_KEY environment variable to protect REST endpoints:
API_KEY=mysecretkey npm run start:remote
Then include the header in requests:
curl -H "Authorization: Bearer mysecretkey" http://localhost:3000/tools/status
Testing with ngrok (public HTTPS URL)
- Install ngrok: https://ngrok.com/download
- Start your server:
npm run start:remote - In a new terminal:
ngrok http 3000 - Copy the
https://xxxx.ngrok-free.appURL
Paste this into Claude → Settings → Integrations → Remote MCP:
https://xxxx.ngrok-free.app/sse
Deploy to Render
- Push this repo to GitHub
- Go to https://render.com → New → Web Service
- Connect your repo
- Set:
- Build Command:
npm install && npm run build - Start Command:
npm run start:remote - Environment:
NODE_ENV=production - Optional:
API_KEY=yoursecretkey
- Build Command:
- Deploy → copy your
https://your-app.onrender.comURL
Paste into Claude Remote MCP:
https://your-app.onrender.com/sse
⚠️ Important for Render: WhatsApp session data (
.wwebjs_auth) does not persist across Render deploys by default. Use a Render Disk (persistent storage) or re-scan the QR after each deploy. See Render Disks.
Project Structure
whatsapp-mcp/
├── src/
│ ├── index.ts # stdio MCP server (Claude Desktop)
│ ├── remote.ts # HTTP/SSE MCP server (Remote MCP + REST API)
│ └── send_message.ts # CLI script to send a message
├── dist/ # Compiled JS (after npm run build)
├── .wwebjs_auth/ # WhatsApp session data (gitignore this!)
├── package.json
├── tsconfig.json
└── README.md
Available MCP Tools
| Tool | Description |
|---|---|
whatsapp_status |
Check connection status |
whatsapp_send_message |
Send a text message |
whatsapp_send_media |
Send image/video/doc |
whatsapp_get_contacts |
List contacts |
whatsapp_get_groups |
List groups |
whatsapp_get_group_info |
Group details + members |
whatsapp_create_group |
Create a new group |
whatsapp_get_chats |
Recent chats |
whatsapp_get_messages |
Messages from a chat |
.gitignore recommendations
node_modules/
dist/
.wwebjs_auth/
.wwebjs_cache/
.env
2848b1a (whatsapp mcp)
推荐服务器
Baidu Map
百度地图核心API现已全面兼容MCP协议,是国内首家兼容MCP协议的地图服务商。
Playwright MCP Server
一个模型上下文协议服务器,它使大型语言模型能够通过结构化的可访问性快照与网页进行交互,而无需视觉模型或屏幕截图。
Magic Component Platform (MCP)
一个由人工智能驱动的工具,可以从自然语言描述生成现代化的用户界面组件,并与流行的集成开发环境(IDE)集成,从而简化用户界面开发流程。
Audiense Insights MCP Server
通过模型上下文协议启用与 Audiense Insights 账户的交互,从而促进营销洞察和受众数据的提取和分析,包括人口统计信息、行为和影响者互动。
VeyraX
一个单一的 MCP 工具,连接你所有喜爱的工具:Gmail、日历以及其他 40 多个工具。
graphlit-mcp-server
模型上下文协议 (MCP) 服务器实现了 MCP 客户端与 Graphlit 服务之间的集成。 除了网络爬取之外,还可以将任何内容(从 Slack 到 Gmail 再到播客订阅源)导入到 Graphlit 项目中,然后从 MCP 客户端检索相关内容。
Kagi MCP Server
一个 MCP 服务器,集成了 Kagi 搜索功能和 Claude AI,使 Claude 能够在回答需要最新信息的问题时执行实时网络搜索。
e2b-mcp-server
使用 MCP 通过 e2b 运行代码。
Neon MCP Server
用于与 Neon 管理 API 和数据库交互的 MCP 服务器
Exa MCP Server
模型上下文协议(MCP)服务器允许像 Claude 这样的 AI 助手使用 Exa AI 搜索 API 进行网络搜索。这种设置允许 AI 模型以安全和受控的方式获取实时的网络信息。