ILGC Workplace and Distraction

Prerequisites

• Python 3.8 or higher
• Node.js and npm
• Camera permissions enabled

Setup Instructions

Windows

1. Create virtual environment:

   cd windows
   python3 -m venv .ilgc
   source .ilgc/Scripts/activate

2. Install dependencies:

   pip install -r requirements.txt

3. Grant camera permissions:

   • Settings > Privacy & Security > Camera
   • Enable camera access for your terminal/command prompt

4. Start backend services (open 4 terminals, all with venv activated):

   # Terminal 1
   python3 api_server.py
   
   # Terminal 2
   cd src
   python3 watch.py
   
   # Terminal 3
   cd src
   python3 client.py
   
   # Terminal 4
   cd src
   python3 collate_data.py

5. Start frontend:

   cd frontend
   npm install
   npm run dev

6. Open in browser: Go to the URL shown by Vite (usually http://localhost:8080/ilgcresearch/).

Mac

1. Create virtual environment:

   cd Mac
   python3 -m venv .ilgc
   source .ilgc/bin/activate

2. Install dependencies & configure pylsl:

   pip install -r requirements.txt
   cp -r lib/pylsl/* .ilgc/lib/python3.13/site-packages/pylsl/lib

3. Grant permissions:

   • Camera: System Preferences > Security & Privacy > Camera
   • Notifications (for interventions): Test with:

   osascript -e 'display notification "Keep up the good work!" with title "ILGC Research"'

4. Start backend services (open 4 terminals, all with venv activated):

   # Terminal 1
   python3 api_server.py
   
   # Terminal 2
   cd src
   python3 watch.py
   
   # Terminal 3
   cd src
   python3 client.py
   
   # Terminal 4
   cd src
   python3 collate_data.py

5. Start frontend:

   cd frontend
   npm install
   npm run dev

6. Open in browser: Go to the URL shown by Vite (usually http://localhost:8080/ilgcresearch/).

How it Works

1. Fill out the task form with participant details
2. Click "Start Task" - this will:
   • Save participant ID, batch, task category, and task description to ./details/task_details.json
   • Start the interventions.py script automatically
   • The interventions script will load the task details and include them in the GPT analysis prompt

The system will now monitor for distractions based on the specific task you've defined!

API Endpoints

• POST /api/save-task-details → Save task details & start interventions
• POST /api/stop-interventions → Stop interventions
• GET /api/status → Get monitoring status
• GET /api/get-task-details → Get current task details

Troubleshooting

Mac – Interventions:

• Allow Script Editor in Notifications settings
• Ensure /usr/bin/osascript has Accessibility settings (Privacy & Security) permissions