Simple Telegram Digest Bot

A simplified Telegram bot that monitors channels/chats, processes messages with AI, and sends concise summaries to your Saved Messages.

Quick Start

1. Install dependencies (using the existing virtual environment):

   source .venv/bin/activate
   pip install -r requirements.txt

2. Configure the bot:

   • Edit config.yaml with your settings
   • Edit watchlist.yaml to specify chats to monitor
   • Your API credentials are already loaded from .env

3. Run the bot:

   source .venv/bin/activate
   python main.py

Features

• 📥 Collect - Get recent messages from specified Telegram chats
• 🔍 Filter - Keep only messages from the last X hours (client-side)
• 🤖 Process - Send filtered messages to OpenAI for AI analysis
• 📄 Output - Generate Markdown digest file and send concise summary to Telegram

File Structure

├── main.py           # Entry point and orchestration
├── config.py         # Configuration loading (YAML + .env)
├── telegram.py       # Telegram API interface (Telethon)
├── llm.py           # OpenAI API provider
├── output.py        # Markdown files and Telegram summaries
├── config.yaml      # Bot configuration
├── watchlist.yaml   # Chats to monitor
├── prompt.txt       # System prompt for AI
├── requirements.txt # Dependencies
└── digests/         # Output folder for Markdown files

Configuration

config.yaml

• telegram: API credentials and session settings
• llm: OpenAI model and API key
• settings: Hours back to look, output directory

watchlist.yaml

• chats: List of channels/chats to monitor
• Each chat can be enabled/disabled individually
• Use @username for public channels, chat_id for private chats

Output

• Markdown files: Detailed digest saved to digests/ folder
• Telegram summary: Concise summary sent to your Saved Messages

Modular Design

Each module can be replaced independently:

• Change LLM provider → edit llm.py only
• Change output format → edit output.py only
• Change Telegram client → edit telegram.py only
• Change configuration → edit config.py only

Example Usage

# Activate virtual environment
source .venv/bin/activate

# Run digest generation once
python main.py

# Show help
python main.py --help

The bot will:

1. Load configuration from YAML files and .env
2. Connect to Telegram and collect recent messages
3. Process messages with OpenAI to generate structured digest
4. Save detailed Markdown file to digests/ folder
5. Send concise summary to your Telegram Saved Messages

Troubleshooting

• Authentication: First run will prompt for phone number and verification code
• No messages: Check watchlist.yaml chat names and enable status
• API errors: Verify OpenAI API key in .env file
• Import errors: Make sure virtual environment is activated: source .venv/bin/activate