ai/mcp_supabase at main · tylerprogramming/ai · GitHub

Name		Name	Last commit message	Last commit date
parent directory ..
README.md		README.md
main.py		main.py
telegram_messaging.py		telegram_messaging.py

README.md

Setup

Clone the repository and navigate to the project directory.
Install dependencies:
```
pip install -r requirements.txt
```

Set up environment variables:

Create a .env file in the project root with the following variables:

SUPABASE_ACCESS_TOKEN=your_supabase_access_token
TELEGRAM_BOT_TOKEN=your_telegram_bot_token
TELEGRAM_CHAT_ID=your_telegram_chat_id

Run the script:
```
python main.py
```
The script will run continuously, executing the main task every morning at 6:00 AM.

Customization

Change Schedule:
To run at a different time, modify the hour and minute in scheduler.add_job in main.py.
Change Telegram Message Format:
Edit the formatting logic in the CrewAI task description or in the send_telegram_message function.

How it Works

At the scheduled time, the script:
1. Connects to Supabase using the MCP server.
2. Retrieves the last 5 records from the habits table.
3. Summarizes the habits and selects a quote from "Atomic Habits".
4. Sends the formatted summary to the configured Telegram chat.

Requirements

Python 3.12+
Node.js (for MCP server)
Supabase account and access token
Telegram bot and chat ID