UConn Course Waitlist Monitor

Automated course availability monitor for UConn Student Admin. Checks the student admin Shopping Cart to determine whether classes are open.

Architecture

Two-Phase Design:

• Phase 1 (agent.py): Selenium-based authentication

  • Handles CAS login
  • Processes Duo MFA
  • Captures session cookies, headers, and state tokens
  • Only runs when session expires or is missing

• Phase 2 (worker.py): HTTP-based polling

  • Uses saved session data
  • Fetches and parses Shopping Cart
  • Detects session expiration
  • Automatically triggers Selenium refresh
  • Sends Discord notifications

Requirements

• Python 3.8+
• Microsoft Edge browser
• Edge WebDriver (must be in PATH)
• Discord webhook URLs

Installation

1. Clone or download this repository

2. Install dependencies

pip install -r requirements.txt

3. Create a Discord server and configure webhooks

Create webhooks

• Go to Server Settings → Integrations → Webhooks
• Click New Webhook
• Select the text channel you want the webhook to post to
• Copy the Webhook URL and paste to creds.json
• Repeat for each channel (you need at least 3 channels/webhooks total)

4. Adjust credentials file at creds/creds.json:

{
    "username": "your_netid",
    "password": "your_password", 
    "duo_webhook_url": "https://discord.com/api/webhooks/...",
    "course_webhook_url": "https://discord.com/api/webhooks/...",
    "error_webhook_url": "https://discord.com/api/webhooks/..."
}

Usage

Manual Execution

Run the worker directly:

python3 worker.py

The worker will:

1. Load or refresh session data
2. Fetch Shopping Cart page
3. Parse course availability
4. Send Discord notification
5. Exit

Automated Monitoring (Cron)

Add to crontab for periodic monitoring:

crontab -e

Example (check every 15 minutes):

*/15 * * * * /usr/bin/python3 /full/path/to/worker.py

Important: Only schedule worker.py in cron. Never schedule agent.py directly.

How It Works

First Run

1. worker.py detects missing session.json
2. Calls agent.py to perform Selenium login
3. Duo verification code sent to Discord (duo_webhook_url)
4. User approves Duo push notification
5. Session data saved to creds/session.json
6. Worker fetches and parses Shopping Cart
7. Results sent to Discord (course_webhook_url)

Subsequent Runs

1. Worker loads existing session from session.json
2. Fetches Shopping Cart via HTTP (no browser needed)
3. Parses course data
4. Sends Discord notification

Session Expiration

When session expires:

1. Worker detects CAS error message in response
2. Automatically triggers Selenium refresh
3. New Duo approval required
4. Session data updated
5. Worker continues polling

Logs

Two separate log files in logs/:

activities.txt

• INFO level only
• Normal workflow, successful fetches, etc.

errors.txt

• ERROR level only
• Selenium failures, HTTP errors, etc.

Both logs use 5MB rotation with 3 backup files

Troubleshooting

"Edge WebDriver not found"

Ensure Edge WebDriver is installed and in PATH: For Linux

which msedgedriver

For Windows

where msedgedriver

Session keeps expiring

• Check system time accuracy
• Verify UConn hasn't changed authentication flow
• Check logs/errors.txt for details

No courses found

• Verify Shopping Cart has courses added in PeopleSoft
• Check that course grid HTML structure hasn't changed
• Review logs/activities.txt for parsing details

Discord webhook failures

• Verify webhook URLs are correct
• Check Discord server permissions
• Ensure webhook hasn't been deleted

Duo not working

• Verify Duo is enabled on your account
• Check duo_webhook_url is correct
• Ensure you approve push notification within timeout (30 seconds)

Security Notes

• creds/creds.json contains plaintext credentials (don't push creds.json onto GitHub)
• Use file permissions to restrict access:

  chmod 600 creds/creds.json

• Consider using environment variables instead of creds

Known Limitations

• Element IDs in PeopleSoft are dynamic and may change
• Only Edge is officially supported

TO DO

• Refactor code for clarity and optimization
• Add feature to switch/add classes when spots become open/available
• Add support for alternative browsers (Chrome, Firefox)
• Optional: Rewrite worker.py in compiled language for faster HTTP requests