Twitter Digest AI

An AI-powered Twitter digest system that monitors selected Twitter accounts, summarizes their content using OpenAI's GPT, and delivers periodic email summaries.

Features

• 🐦 Monitors tweets from a configurable list of Twitter users
• 🤖 Uses OpenAI's GPT to generate intelligent summaries
• 📧 Sends periodic email digests (hourly/daily/weekly)
• 🎯 Customizable prompt templates for AI summarization
• 💾 Maintains memory of previous summaries

Prerequisites

• Node.js (v20 or higher)
• OpenAI API key
• Twitter API credentials
• SMTP email server access

Installation

1. Clone the repository:

git clone git@github.com:BaltasarAroso/ai-workshop.git
cd ai-workshop

2. Install dependencies:

npm install

3. Set up environment variables:

cp .env.example .env

Edit the .env file with your credentials and settings. See .env.example for all required variables.

Configuration

Twitter Users

Edit data/twitter_users.txt to specify the Twitter handles you want to monitor (one per line).

Prompt Template

Customize the AI summarization by editing data/prompt_template.txt.

Environment Variables

Configure the following in your .env file:

• OpenAI API credentials
• Twitter API credentials
• SMTP email settings
• Update frequency (hourly/daily/weekly)

Example:

OPENAI_API_KEY=your_openai_api_key

# Twitter Configuration
TWITTER_USERNAME=your_twitter_username
TWITTER_PASSWORD=your_twitter_password
TWITTER_EMAIL=your_twitter_email
TWITTER_2FA_SECRET=your_twitter_2fa_secret

# Email Configuration
SMTP_HOST=smtp.gmail.com
SMTP_PORT=587
SMTP_USER=your_email@gmail.com
SMTP_PASSWORD=your_app_specific_password
EMAIL_FROM=your_email@gmail.com
EMAIL_TO=recipient@email.com

# Update Frequency (in hours)
# Example:
# 15 minutes = 0.25
# 1 hour = 1
# 1 day = 24
# 1 week = 168
# 1 month = 720
UPDATE_FREQUENCY=24```

## Usage

Start the application:

```bash
npm start

The system will:

1. Load the configured Twitter users
2. Fetch their recent tweets
3. Generate AI summaries
4. Send email digests according to the configured schedule

Project Structure

├── src/
│   ├── agents/
│   │   └── TwitterDigestAgent.js    # Main agent logic
│   ├── config/
│   │   └── config.js                # Configuration management
│   └── index.js                     # Application entry point
├── data/
│   ├── twitter_users.txt            # List of Twitter users to monitor
│   └── prompt_template.txt          # AI summarization prompt template
├── database/
│   └── memory.json                  # Persistent storage
├── package.json
├── .env                             # Environment variables
└── README.md

Customization

Update Frequency

Modify the UPDATE_FREQUENCY in your .env file:

• 0.25: Runs every 15 minutes
• 1: Runs every hour
• 24: Runs once per day
• 168: Runs once per week

Email Template

Customize the email format by modifying the formatEmailContent method in utils/utils.js.

Memory Storage

The system maintains state in database/memory.json. This helps track previously processed tweets and maintain context for summaries.

License

This project is licensed under the MIT License - see the LICENSE file for details.