Screen Tracker - Windows Screen Tracking System

A Windows-only screen tracking system that captures screenshots every 5 minutes, analyzes them with local AI, and provides productivity insights through a web dashboard.

Features

• Automated Screenshot Capture: Takes screenshots every 5 minutes on Windows
• Privacy Protection: Auto-blurs sensitive content (passwords, credit cards, emails)
• AI-Powered Analysis: Uses local Ollama with vision models for activity categorization
• Daily Summaries: Automatic generation of daily activity summaries
• Time Allocation Statistics: Detailed breakdown of time spent by app and category
• Productivity Suggestions: AI-generated recommendations for time management improvement
• Web Dashboard: Beautiful interface for viewing all tracked data

Tech Stack

Backend

• Node.js + TypeScript
• Express.js (API server)
• node-screenshots (Windows screenshot capture)
• active-win (active window detection)
• node-cron (scheduling)
• better-sqlite3 (embedded database)
• sharp (image processing)
• Tesseract.js (OCR for privacy detection)
• ollama (local AI integration)

Frontend

• Next.js 15 with App Router
• React 19
• Tailwind CSS
• Lucide React (icons)

Prerequisites

1. Node.js (v18 or higher)
2. Ollama with a vision model installed (e.g., llava or bakllava)

Installing Ollama

# Download Ollama from https://ollama.ai
# Pull a vision model
ollama pull llava

Installation

1. Clone the repository and navigate to the project directory:

cd "D:\Zhuobin Vide Coding\【review screen】"

2. Install all dependencies:

npm run install:all

3. Configure environment variables:

# Copy the example .env file
cp backend/.env.example backend/.env

# Edit backend/.env with your settings
# Default values should work for most cases

Usage

Development Mode

Start both backend and frontend in development mode:

npm run dev

This will start:

• Backend API server on http://localhost:3001
• Frontend dashboard on http://localhost:3000

Production Mode

1. Build the project:

npm run build

2. Start the production server:

npm start

Manual Controls

Start/Stop Tracking

• Pause: Click "Pause" button in the dashboard
• Resume: Click "Resume" button in the dashboard
• Manual Capture: Click "Capture Now" to take an immediate screenshot

Generate Daily Summary

# Via API
POST http://localhost:3001/api/analysis/generate/YYYY-MM-DD

Configuration

Edit backend/.env to customize:

# Server
PORT=3001

# Scheduling
CAPTURE_INTERVAL_MINUTES=5
IDLE_THRESHOLD_MINUTES=10

# Screenshot Quality
SCREENSHOT_QUALITY=85

# Data Retention
MAX_SCREENSHOT_AGE_DAYS=90

# Ollama
OLLAMA_BASE_URL=http://localhost:11434
OLLAMA_MODEL=llava

# Privacy Features
PRIVACY_BLUR_PASSWORDS=true
PRIVACY_BLUR_CREDIT_CARDS=true
PRIVACY_BLUR_EMAILS=true
PRIVACY_AI_DETECTION=true

Dashboard Features

Main Dashboard

• Today's activity overview
• Screenshot count and productivity score
• Time allocation by category
• Top applications used
• Daily summary text
• AI-generated suggestions

API Endpoints

Screenshots

• GET /api/screenshots - List screenshots
• GET /api/screenshots/:id - Get single screenshot
• GET /api/screenshots/:id/image - Get screenshot image
• GET /api/screenshots/today - Get today's screenshots

Analysis

• GET /api/analysis/:date - Get analysis for date
• GET /api/analysis/times/:startDate/:endDate - Get time allocation
• POST /api/analysis/generate/:date - Generate daily summary

Dashboard

• GET /api/dashboard/today - Today's dashboard data
• GET /api/dashboard/stats - Quick statistics
• GET /api/dashboard/suggestions - Productivity suggestions

System

• GET /api/system/status - System status
• POST /api/system/pause - Pause tracking
• POST /api/system/resume - Resume tracking
• POST /api/system/capture - Manual capture

Data Storage

All data is stored locally in the data/ directory:

data/
├── screenshots/
│   ├── original/     # Original screenshots
│   ├── blurred/      # Privacy-blurred versions
│   └── thumbnails/   # Gallery thumbnails
└── exports/          # Exported reports

Database: backend/database/screen-tracker.db (SQLite)

Privacy & Security

• Local-Only: All data stored locally, never uploaded
• Automatic Blur: Sensitive content detected and blurred before storage
• OCR Detection: Text detection for passwords, emails, credit cards
• Pattern Matching: Regex patterns for API keys, SSNs, etc.
• Window Title Detection: Blur known sensitive windows

Troubleshooting

Ollama Not Available

If Ollama is not running:

1. Start Ollama: ollama serve
2. Verify model is installed: ollama list
3. System will use fallback rule-based analysis

Screenshots Not Capturing

1. Check system status in dashboard
2. Verify backend is running on port 3001
3. Check logs in logs/ directory

Database Errors

1. Ensure backend/database/ directory exists
2. Check file permissions
3. Delete .db file and restart to recreate

Project Structure

screen-tracker/
├── backend/              # Node.js backend
│   ├── src/
│   │   ├── core/        # Screenshot, privacy, Ollama, scheduler
│   │   ├── storage/     # Database and file system
│   │   ├── analysis/    # Summarizer, time tracker
│   │   ├── api/         # Express routes
│   │   ├── config/      # Settings and constants
│   │   └── utils/       # Logger, helpers
│   └── database/        # SQLite database
├── frontend/            # Next.js dashboard
│   └── app/            # Pages and components
├── data/               # Screenshot storage
└── logs/               # Application logs

License

MIT

Contributing

This is a personal project for time management and productivity tracking.