🤖 AI Stream Agent

This Express.js app connects to OpenAI to power real-time chat completions with Server-Sent Events (SSE). It streams responses seamlessly, keeping conversations smooth and responsive. 🚀

Table of Contents

• Installation
• Usage
• Testing Endpoints

Features

• Real-Time Streaming: Utilizes SSE to stream OpenAI chat completions with low latency.
• Class-Based Controller: Implements a ChatController class for better organization, maintainability, and context binding.
• Input Sanitization: Limits message content length and supports configurable models (gpt-3.5-turbo, gpt-4).
• Robust Validation: Ensures messages are non-empty arrays with valid role and content fields.
• Comprehensive Logging: Tracks request lifecycle, errors, and client disconnections with request IDs.
• Error Handling: Gracefully handles errors with detailed SSE error events and HTTP responses.
• Configurability: Supports environment variables for max tokens, temperature, and port settings.

Installation

1. Clone the repository:

   git clone git@github.com:ivstudio/ai-stream-agent.git
   cd ai-stream-agent

2. Install dependencies:

   npm install

3. Create a .env file in the root directory and add your OpenAI API key:

   OPENAI_API_KEY=your_openai_api_key

Usage

1. Start the server:

   npm run dev

2. The server will start on http://localhost:3001.

Testing Endpoints

You can test the chat completion endpoint using cURL:

curl -N -X POST \
  -H "Content-Type: application/json" \
  -d '{"messages": [{"role": "user", "content": "Hello, how are you?"}]}' \
  http://localhost:3001/api/chat

POST

This endpoint accepts a list of messages and returns a stream of chat completions.

• URL: /api/chat

• Method: POST

• Headers: Content-Type: application/json

• Body:

  {
      "messages": [{ "role": "user", "content": "Hello!" }],
      "model": "gpt-3.5-turbo" // Optional: gpt-3.5-turbo or gpt-4
  }

• Response: Server-Sent Events (SSE) stream with chat completions.

Response: SSE stream with events

data: {"content": "Hi there!", "requestId": "uuid"} (message chunks)
data: {"event": "[DONE]", "requestId": "uuid"} (completion signal)
Errors: data: {"error": "message", "event": "[ERROR]", "requestId": "uuid"}

Error Responses

400 Bad Request: Invalid messages or model.
500 Internal Server Error: Server-side issues.