# MCP (Model Context Protocol)

The Model Context Protocol (MCP) is an emerging open standard that lets AI agents securely access tools, APIs, and external data through a unified interface. Instead of writing custom integrations or backend logic, MCP bridges your local or cloud tools directly to the AI model — giving it controlled, structured access to real-world functions.
\
When combined with Whapi.Cloud’s WhatsApp API, MCP enables agents to send messages, manage groups, post to channels, or trigger workflows instantly. It transforms WhatsApp automation from code-heavy development into a plug-and-run experience.

In short, MCP brings the missing layer of intelligence between your AI agent and Whapi’s powerful WhatsApp API — letting you build, test, and scale automations faster than ever.

***

### Quick start

Requires Node.js 18+.

Install and run via NPX:

```
npx -y whapi-mcp@latest
```

Or set your Whapi API token and run:

```
# PowerShell
$env:API_TOKEN="YOUR_TOKEN"; npx -y whapi-mcp@latest
```

### Using with MCP clients

* Cursor: add to %USERPROFILE%.cursor\mcp.json

```
{
  "mcpServers": {
    "whapi-mcp": {
      "command": "npx",
      "args": ["-y", "whapi-mcp@latest"],
      "env": { "API_TOKEN": "YOUR_TOKEN" }
    }
  }
}
```

* Claude Desktop: add to %APPDATA%\Claude\claude\_desktop\_config.json (same structure as above).

### Tools

The server exposes tools generated from Whapi OpenAPI. Example:

* sendMessageText (required: to, body)

Example call (pseudo):

```
name: sendMessageText
arguments: { "to": "1234567890@s.whatsapp.net", "body": "Hello" }
```

***

{% hint style="success" %}
**Agent Skill available**

If you use AI coding agents (Cursor, Claude Code, GitHub Copilot, etc.), install the **Whapi Agent Skill** alongside the MCP server. It provides your agent with accurate reference material

```bash
npx skills add Whapi-Cloud/whapi-whatsapp-api-skill
```

[Learn more →](https://support.whapi.cloud/help-desk/ai-tools/agent-skills)
{% endhint %}