Whasapo
WhatsApp MCP server written in Go. Send and read WhatsApp messages from Claude, OpenClaw, or any MCP-compatible client. Single binary, no runtime dependencies.
Whasapo
WhatsApp MCP server. Send and read WhatsApp messages from any AI assistant that supports MCP — Claude desktop, Claude Code, OpenClaw, and more.
Install
macOS / Linux:
curl -sSL https://raw.githubusercontent.com/toloco/whasapo/main/install.sh | bash
Windows (PowerShell):
irm https://raw.githubusercontent.com/toloco/whasapo/main/install.ps1 | iex
This will:
- Download the latest binary for your platform
- Configure the Claude desktop app
- Walk you through linking your WhatsApp account (QR code scan)
After install, restart your AI app.
Setup by app
Claude Desktop
The installer configures this automatically. If you need to do it manually, add to ~/Library/Application Support/Claude/claude_desktop_config.json:
{
"mcpServers": {
"whatsapp": {
"command": "~/.whasapo/whasapo",
"args": ["serve"]
}
}
}
Restart Claude Desktop. The WhatsApp tools will be available immediately.
Claude Code
Add the MCP server to your project or global settings:
claude mcp add whatsapp ~/.whasapo/whasapo serve
Or add to ~/.claude/settings.json:
{
"mcpServers": {
"whatsapp": {
"command": "~/.whasapo/whasapo",
"args": ["serve"]
}
}
}
OpenClaw
In OpenClaw settings, add a new MCP server:
- Name: whatsapp
- Command:
~/.whasapo/whasapo - Arguments:
serve
Or add to your OpenClaw config file:
{
"mcpServers": {
"whatsapp": {
"command": "~/.whasapo/whasapo",
"args": ["serve"]
}
}
}
Any MCP-compatible client
Whasapo is a standard MCP server using stdio transport. Point any MCP client at:
command: ~/.whasapo/whasapo
args: serve
What can you do with it?
Ask your AI assistant things like:
- "Show me my recent WhatsApp messages"
- "Send a WhatsApp message to John saying I'll be 10 minutes late"
- "What messages did I get in the family group?"
- "Find my contact named Sarah"
- "Reply to Mom's last message saying thanks"
- "Summarize what I missed in the work group chat"
Available tools
| Tool | Description |
|---|---|
send_message | Send a text message to a contact or group |
list_chats | List recent chats with last message preview |
get_messages | Get messages, optionally filtered by chat |
search_contacts | Search contacts by name or phone number |
Commands
whasapo pair Link your WhatsApp account (QR code)
whasapo serve Start the MCP server (your AI app does this automatically)
whasapo status Check if everything is working
whasapo update Update to the latest version
whasapo uninstall Remove whasapo completely
whasapo version Print version
Troubleshooting
"Claude doesn't show WhatsApp tools" Restart the app after installing.
"Can't be opened because Apple cannot check it for malicious software" Run this, then try again:
xattr -d com.apple.quarantine ~/.whasapo/whasapo
"Connection failed" or "not paired" Your WhatsApp link may have expired. Re-pair:
whasapo pair
"No messages found" Messages from before the first install won't appear. Once installed, messages are stored persistently and survive restarts.
Uninstall
whasapo uninstall
Or remotely:
curl -sSL https://raw.githubusercontent.com/toloco/whasapo/main/install.sh | bash -s -- --uninstall
Build from source
Requires Go 1.23+.
make build # build for your machine → bin/whasapo
make release # macOS universal binary → dist/whasapo-VERSION-macos.zip
make release-all # all platforms → dist/whasapo-VERSION-{macos,linux,windows}.*
How it works
Whasapo is an MCP server that connects to WhatsApp using the whatsmeow library — the same protocol the official WhatsApp apps use.
Your WhatsApp session is stored locally in ~/.whasapo/session.db. Messages are persisted in SQLite so they survive restarts. No data is sent to any third-party server.
Похожие серверы
BotEmail MCP Server
Give your AI agent its own email address — instant bot inboxes via API, no human setup required.
Human-in-the-Loop Slack MCP Server
Allows AI assistants to request information and receive responses from humans via Slack.
Slack MCP Server
A server for integrating with Slack, enabling communication and automation within your workspace.
AllVoiceLab
An AI voice toolkit with TTS, voice cloning, and video translation, now available as an MCP server for smarter agent integration.
Qiye Wechat MCP
Enables AI assistants to send messages to Enterprise WeChat (Qiye Wechat) groups via webhooks.
better-telegram-mcp
Production-grade MCP server for Telegram with dual-mode Bot API + MTProto, 6 composite tools
Didlogic MCP Server
Interact with the Didlogic API for telecommunication services, including virtual phone numbers and SMS.
BulkPublish
Bulk publish to multiple social media at once.
Prompt for User Input MCP Server
Enables AI models to prompt users for input directly within their code editor for interactive conversations.
Voidly MCP Server
3 tools for global censorship intelligence (119 countries) + E2E encrypted agent-to-agent messaging with Double Ratchet and post-quantum crypto