WhatsApp MCP Server (Very Good Plugins)
WhatsApp messages, contacts, and media interactions managed directly via AI agents.
Data last scanned today · Reviewed today
Overview
This MCP server bridges WhatsApp communication into an AI-compatible interface. It supports reading and sending text messages, processing various media formats like images and documents, and searching contacts by display name or phone number. Additionally, it can capture call history into a local SQLite database for integration with your workflow. All message data is kept in local SQLite storage, ensuring you maintain control over what information is transmitted to your AI models. The server also provides webhook capabilities to forward incoming messages to external services, expanding its utility beyond simple chat functionality.
Our verdict
With a security score of 45/100, this server is classified as high risk. Because it is community-maintained rather than official, and because critical aspects like authentication and read-only mode have not been editorially reviewed, users should exercise significant caution. It is suitable for those comfortable with self-hosting third-party bridges for personal messaging, but it is not recommended for environments where sensitive data governance is a primary requirement.
- Zero direct dependencies reduces supply chain attack surface
- Twenty-four contributors supporting the codebase
- Active repository with recent daily commits
- High risk score due to lack of editorial review
- Missing clear license terms marked as NOASSERTION
- Authentication and read-only implementations lack external security verification
Tools
| Tool | Description | Risk |
|---|---|---|
| search_contacts | Search contacts by name or phone number. | low |
| get_contact | Resolve a WhatsApp contact name from a phone number, LID, or full JID. | low |
| list_messages | Get messages with filters, date ranges, and sorting. | low |
| send_message | Send a text message to a contact or group, optionally as a quoted reply. | medium |
| send_reaction | Send (or remove) an emoji reaction to a message. | medium |
| send_file | Send a media file (image, video, document). | medium |
| send_audio_message | Send a voice message (automatically converts to Opus .ogg format). | medium |
| download_media | Download media from a received message. | low |
| list_chats | List all chats with metadata. | low |
| get_chat | Get specific chat metadata by JID. | low |
Compatibility
| Client | Local | Docker | Remote | Read-only |
|---|---|---|---|---|
| ChatGPT | ||||
| Claude Desktop | ||||
| Cursor | ||||
| VS Code | ||||
| Windsurf |
Frequently asked questions
›What software prerequisites are required to run the WhatsApp MCP server?
You need Go 1.25+, Python 3.11+, the uv package manager, and either Claude Desktop or Cursor installed. FFmpeg is optionally required if you want support for voice message conversion.
›How do I authenticate the WhatsApp bridge?
Upon starting the bridge for the first time using the go run command, it generates a local REST API token and displays a QR code which you must scan with your phone's WhatsApp application.
›Does updating the server require me to re-pair my WhatsApp account?
No, updating the code does not require re-pairing or deleting your database; your message history and session are preserved unless you explicitly request a full history refresh.
›Where are my WhatsApp messages stored?
All messages are stored locally in an SQLite database, and they are only transmitted to Claude when you explicitly allow it through the MCP interface.
›What should I do if the bridge and MCP server are in different directory locations?
You must set the same WHATSAPP_BRIDGE_TOKEN environment variable in both the bridge and the MCP server environments to ensure they can communicate correctly.
Badge
Maintain this server? Add the live badge to your README.