AI-Native SMS Assistant (V1)

An AI-powered SMS assistant that integrates with MessageMedia to send, receive, and manage SMS conversations. Built with Node.js and OpenAI’s GPT-4, it provides real-time context-aware messaging and a user-friendly dashboard.

Features

• Send SMS: Send messages via the /send endpoint.
• Receive & Auto-Reply: Handle incoming messages with GPT-generated responses.
• Contextual Insights: Retrieve conversation context through the /context endpoint.
• Dashboard: Visualize message logs and statuses.
• Chat Interface: Interact with the assistant via a web-based chat UI.
• Function Schema: OpenAI-compatible function schema for integration.

Installation

1. Clone the repository:

   git clone https://github.com/larryfang/sms-mcp
   cd sms-mcp
   

2. Install dependencies:

   npm install
   

3. Configure environment variables:

   Create a .env file in the root directory and add the following:

   OPENAI_API_KEY=your_openai_api_key
   MESSAGE_API_KEY=your_messagemedia_api_key
   MESSAGE_API_SECRET=your_messagemedia_api_secret
   MESSAGE_BASE_URL=https://api.messagemedia.com
   MCP_SERVER_URL=http://localhost:3000
   

4. Start the server:

   node index.js
   

   The server will be running at http://localhost:3000.

Usage

• Send SMS:

  POST /send
  {
    "messages": [
      {
        "destination_number": "+61400000000",
        "content": "Hello, this is a test message."
      }
    ]
  }
  

• Retrieve Context:

  POST /context
  {
    "phone_number": "+61400000000"
  }
  

• Access Dashboard:

  Navigate to http://localhost:3000/dashboard to view message logs.

• Chat Interface:

  Open http://localhost:3000/chat.html in your browser to interact with the assistant.

Project Structure

• index.js: Main server file with route implementations.
• public/: Static files including chat.html.
• conversations/: Stored conversation logs per user.
• webhook-log.json: Logs of incoming and outgoing messages.

Architecture

Overview

The AI-Native SMS Assistant is designed to facilitate seamless SMS communication by integrating OpenAI’s GPT-4 for intelligent responses and MessageMedia for message delivery.

Components

1. Express Server (index.js)

   • Endpoints: /send, /context, /webhook/reply, /webhook/delivery, /dashboard, /meta, /function-schema.

2. OpenAI Integration

   • Generates context-aware responses and classifies user intents.

3. MessageMedia Integration

   • Sends and receives SMS messages and delivery reports.

4. Chat Interface (public/chat.html)

   • Web-based interface for interacting with the assistant.

Data Flow

• Sending Messages → via /send
• Receiving & Auto-reply → via /webhook/reply
• Context Retrieval → via /context
• Logs UI → via /dashboard
• Chat + Function Calls → via /chat.html

Demo Script

Duration: ~5 minutes

Introduction

“Welcome to the AI-Native SMS Assistant demo. Today, we’ll showcase how this assistant streamlines SMS communications using OpenAI’s GPT-4 and MessageMedia.”

Sending an SMS

“We’ll use the /send endpoint to dispatch a greeting to a user.”

Receiving and Auto-Replying

“Simulate an incoming message. The assistant processes it and replies intelligently.”

Retrieving Conversation Context

“Use the /context endpoint to view message history and delivery reports.”

Exploring the Dashboard

“Visit /dashboard to explore message logs visually.”

Chat Interface

“Interact naturally via chat.html. It detects phone numbers, triggers GPT tools, and shows results contextually.”

License

MIT License