🤖 Wawp AI-Ready Documentation (MCP)

At Wawp, we believe that documentation should be as smart as the code it describes. That's why we've implemented the Model Context Protocol (MCP), allowing you to connect our entire API documentation directly to your favorite AI development tools.

🚀 What is MCP?

The Model Context Protocol (MCP) is an open standard that enables AI models (like Claude, GPT-4, or those inside Cursor) to securely and reliably access external data sources. Instead of copying and pasting documentation, the AI can "ask" our server for the exact information it needs.

🛠️ How to use the Wawp MCP Server

You can run the Wawp MCP server locally to give your AI full context of our WhatsApp API.

1. Prerequisites

Node.js (LTS Version): Required to run the MCP server. Download it from nodejs.org.
AI Client: An AI client that supports MCP (like Cursor, Windsurf, or Claude Desktop).

[!IMPORTANT] If you see an error like executable file not found in %path%, it means Node.js is not installed or your terminal needs a restart after installation.

2. Connect Your Favorite AI Tools

Follow these simple steps for your platform:

🖱️ Cursor

Go to Settings > Cursor Settings > Models > MCP and add a new tool:

Name: wawp-api
Type: command
Command:

npx -y @wawp/mcp-server@latest

🌊 Windsurf

Go to Windsurf Settings (Cmd+,) > MCP > Add Server:

Name: wawp-api
Type: command
Command:

npx -y @wawp/mcp-server@latest

🤖 Claude Desktop

Open your claude_desktop_config.json and add:

{
  "mcpServers": {
    "wawp": {
      "command": "npx",
      "args": ["-y", "@wawp/mcp-server@latest"]
    }
  }
}

🚀 AI Agents (Antigravity / Cortex / etc.)

If you are using a command-line agent or a custom AI, use this command to start the MCP server:

npx -y @wawp/mcp-server@latest

⚡ Boost Your AI Agent

Give your AI deep knowledge of Wawp's WhatsApp API by installing our Agent Rules and API Integration Skill directly into your project. Simply tell your AI agent:

"Use the install_agent_config tool to set up my project with Wawp standards."

This will automatically create:

.cursorrules: Strict rules for secure credential management, correct chatId formats, error handling, and anti-spam rate limiting.
.agent/skills/wawp-api-integration: A complete technical guide covering all endpoints, webhook handling, interactive messages, and Node.js code patterns.

Supported operations

Wawp's MCP server exposes the following operations:

search_docs: Find endpoints for specific tasks (e.g., "How do I send a PDF?").
list_endpoints: See all available categories and actions.
get_endpoint_details: Get exact JSON schemas, required parameters, and response examples for any endpoint.
set_config: Configure your instance_id, access_token, and test_number for the current session.
execute_request: Perform real API calls directly from your AI conversation (Logged as MCP).
get_session_health: Instantly check if your WhatsApp instance is connected.
send_local_file: Send a file (image, PDF, etc.) directly from your local computer.
generate_starter_code: Get a fully working script for any task.
install_agent_config: Setup Wawp premium agent rules in one click.

🔑 How to Provide Your Credentials

To make the AI capable of executing real requests, you need to tell it your API details. You can do this by simply telling the AI:

"Set my Wawp configuration: Access Token is YOUR_TOKEN, Instance ID is YOUR_INSTANCE, and Test Number is YOUR_NUMBER."

The AI will then use the set_config tool to remember these details for the rest of your conversation. You can find your Access Token and Instance ID in the Wawp Dashboard.

📄 Automatic Setup (.env)

You can also create a .env file in your project root to load credentials automatically:

WAWP_ACCESS_TOKEN=your_token_here
WAWP_INSTANCE_ID=your_instance_id
WAWP_TEST_NUMBER=recipient_number

🗺️ Your First AI-Powered WhatsApp Task

If you've never written code before, don't worry. Here is how you can start:

Connect the MCP: Follow the steps in section 2.
Provide context: Tell the AI: "I want to send a welcome message to my customers. Use my Wawp credentials."
Generate & Run: Tell the AI: "Use the generate_starter_code tool to write a Node.js script that sends 'Hello' to my test number."
Execute: The AI will give you the code. Just copy, paste, and run it!

💡 Why use this?

Zero Guesswork: The AI will always use the correct parameters and types.
Save Time: No more switching between your editor and the documentation website.
Auto-pilot: Leverage the design system skills to build UI that matches Wawp perfectly.

💾 Export & Downloads

Need the full API specification for your own tools? Visit our Export Data page to download Wawp API in OpenAPI, Markdown, or HTML formats.