WAHA MCP Server
Enables AI assistants to interact with WhatsApp through the WAHA (WhatsApp HTTP API) platform. Supports chat management, message operations including sending/receiving messages, and marking chats as read.
README
WAHA MCP Server
A Model Context Protocol (MCP) server that enables AI assistants to interact with WhatsApp through the WAHA (WhatsApp HTTP API) platform.
Features
- Chat Management: Get overview of recent WhatsApp chats
- Message Operations: Retrieve, send, and mark messages as read
- MCP Integration: Full compatibility with MCP clients like Claude Desktop
Installation
-
Clone this repository
-
Install dependencies:
npm install -
Configure your WAHA API connection:
cp .env.example .envThen edit
.envwith your WAHA instance details.
Configuration
Required environment variables in .env:
WAHA_BASE_URL: Your WAHA server URL (e.g.,http://localhost:3000)WAHA_API_KEY: Your WAHA API key for authenticationWAHA_SESSION: WhatsApp session name (default:default)
Development
Build the project
npm run build
Run in development mode with auto-reload
npm run dev
Test with MCP Inspector
The MCP Inspector is a web-based tool to test your MCP server interactively:
npm run inspector
This will:
- Start the WAHA MCP server
- Launch the MCP Inspector web UI
- Open your browser automatically
Use the inspector to:
- Test all 4 tools interactively
- View tool schemas
- See request/response data
- Debug tool calls
Run in production
npm start
Usage with Claude Desktop
Add this to your Claude Desktop MCP configuration:
{
"mcpServers": {
"waha": {
"command": "node",
"args": ["/absolute/path/to/waha-mcp-server/dist/index.js"],
"env": {
"WAHA_BASE_URL": "http://localhost:3000",
"WAHA_API_KEY": "your-api-key-here",
"WAHA_SESSION": "default"
}
}
}
}
Available Tools
waha_get_chats
Get overview of recent WhatsApp chats. Returns chat ID, name, last message preview, and unread count.
Parameters:
limit(optional): Number of chats to retrieve (default: 10, max: 100)offset(optional): Offset for paginationchatIds(optional): Array of specific chat IDs to filter (format:number@c.us)
Example:
{
"limit": 10
}
waha_get_messages
Get messages from a specific WhatsApp chat. Returns message content, sender, timestamp, and status.
Parameters:
chatId(required): Chat ID to get messages from (format:number@c.usfor individual,number@g.usfor group)limit(optional): Number of messages to retrieve (default: 10, max: 100)offset(optional): Offset for paginationdownloadMedia(optional): Download media files (default: false)
Example:
{
"chatId": "1234567890@c.us",
"limit": 10
}
waha_send_message
Send a text message to a WhatsApp chat. Returns message ID and delivery timestamp.
Parameters:
chatId(required): Chat ID to send message to (format:number@c.us)text(required): Message text to sendreplyTo(optional): Message ID to reply tolinkPreview(optional): Enable link preview (default: true)
Example:
{
"chatId": "1234567890@c.us",
"text": "Hello from MCP!"
}
waha_mark_chat_read
Mark messages in a chat as read. Can specify number of recent messages or time range in days.
Parameters:
chatId(required): Chat ID to mark as read (format:number@c.us)messages(optional): Number of recent messages to mark as read (default: 30)days(optional): Mark messages from last N days as read (default: 7)
Example:
{
"chatId": "1234567890@c.us",
"messages": 30
}
Project Structure
waha-mcp-server/
├── src/
│ ├── index.ts # Server entry point
│ ├── config.ts # Configuration management
│ ├── types/ # TypeScript type definitions
│ ├── tools/ # MCP tool implementations
│ └── client/ # WAHA API client
├── dist/ # Built JavaScript output
├── .env # Your configuration (not in git)
├── .env.example # Configuration template
├── package.json
├── tsconfig.json
└── PROJECT_SPEC.md # Full project specification
Chat ID Format
WhatsApp chat IDs have specific formats:
- Individual chats:
<phone_number>@c.us(e.g.,1234567890@c.us) - Group chats:
<group_id>@g.us(e.g.,123456789012345678@g.us)
You can get chat IDs by using the waha_get_chats tool first.
License
ISC
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。