Command Palette

Search for a command to run...

UnylyUnyly
Browse all

ScreenMonitorMCP

FreeNot checked

Enables real-time screen monitoring, UI element analysis, and predictive user behavior learning for AI assistants.

GitHubEmbed

About

Enables real-time screen monitoring, UI element analysis, and predictive user behavior learning for AI assistants.

README

CI License: MIT Python 3.9+ MCP Compatible

A REVOLUTIONARY Model Context Protocol (MCP) server! Gives AI real-time vision capabilities, UI intelligence, and predictive behavior learning power. This isn't just screen capture - it gives AI the power to truly "see" and understand your digital world!

🌟 WHY ScreenMonitorMCP?

  • 🔥 First & Only: Real-time continuous screen monitoring feature
  • 🧠 AI Intelligence: AI that understands UI elements and can interact with them
  • 🔮 Predictive: System that learns and predicts user behaviors
  • Proactive: Assistant that offers help before you need it
  • 🎯 Natural: AI that understands commands like "Click the save button"

🔥 REVOLUTIONARY FEATURES

🔄 Real-Time Continuous Monitoring

  • AI's Eyes Never Close: 2-5 FPS continuous screen monitoring
  • Smart Change Detection: Distinguishes between small, major, and critical changes
  • Proactive Analysis: AI automatically analyzes important changes
  • Adaptive Performance: Smart frame rate adjustment

🎯 UI Element Intelligence

  • Computer Vision UI Detection: Automatically recognizes buttons, forms, menus
  • OCR Text Extraction: Reads text from anywhere on the screen
  • Smart Click System: Natural language commands like "Click the save button"
  • Interaction Mapping: AI knows exactly where and how to interact

🧠 Predictive Intelligence

  • Behavior Learning: AI learns your usage patterns and habits
  • Intent Prediction: Predicts what you'll do next based on context
  • Proactive Help: Offers help before you ask
  • Workflow Optimization: Suggests improvements in your work patterns

🛠️ REVOLUTIONARY MCP TOOLS

🔄 Real-Time Monitoring Tools

  • start_continuous_monitoring() - Starts AI's continuous vision capability
  • stop_continuous_monitoring() - Stops continuous monitoring
  • get_monitoring_status() - Real-time status information and statistics
  • get_recent_changes() - Recently detected screen changes

🎯 UI Intelligence Tools

  • analyze_ui_elements() - Recognizes and maps all UI elements on screen
  • smart_click() - Smart clicking with natural language commands ("Click the save button")
  • extract_text_from_screen() - OCR text extraction from screen

🧠 Predictive AI Tools

  • learn_user_patterns() - Learns and analyzes user behavior patterns
  • predict_user_intent() - Predicts user intent based on current context
  • proactive_assistance() - Offers proactive help before user requests
  • record_user_action() - Records user actions and feeds learning system

📸 Traditional Tools

  • capture_and_analyze() - Screen capture and AI analysis (enhanced)
  • list_tools() - MCP standard compliant lists all tools (categorized, detailed information)

🎯 USAGE SCENARIOS

🔍 Real-Time Monitoring

# Start AI's continuous vision capability
await start_continuous_monitoring(fps=3, change_threshold=0.1)

# Check monitoring status
status = await get_monitoring_status()

# View recent changes
changes = await get_recent_changes(limit=5)

🎯 UI Intelligence

# Analyze all UI elements on screen
ui_analysis = await analyze_ui_elements()

# Smart clicking with natural language
await smart_click("Click the save button")

# Extract text from screen
text_data = await extract_text_from_screen()

🧠 Predictive AI

# Learn user behavior patterns
patterns = await learn_user_patterns()

# Predict user intent
intent = await predict_user_intent()

# Get proactive assistance
assistance = await proactive_assistance()

🚀 INSTALLATION

1. Prepare Project Files

# Navigate to project directory
cd ScreenMonitorMCP

# Install required libraries
pip install -r requirements.txt

2. Configure Environment Variables

Edit the .env file:

# Server Configuration
HOST=127.0.0.1
PORT=7777
API_KEY=your_secret_key

# AI Configuration
OPENAI_API_KEY=your_openai_api_key
OPENAI_BASE_URL=https://api.openai.com/v1
DEFAULT_OPENAI_MODEL=gpt-4o

3. Standalone Testing (Optional)

# Test the server
python main.py

# Test revolutionary features
python test_revolutionary_features.py

🔧 MCP CLIENT SETUP

Claude Desktop / MCP Client Configuration

Add the following JSON to your MCP client's configuration file:

🎯 Simple Configuration (Recommended)

{
  "mcpServers": {
    "screenMonitorMCP": {
      "command": "python",
      "args": ["/path/to/ScreenMonitorMCP/main.py"],
      "cwd": "/path/to/ScreenMonitorMCP"
    }
  }
}

🔧 Advanced Configuration

{
  "mcpServers": {
    "screenMonitorMCP": {
      "command": "python",
      "args": [
        "/path/to/ScreenMonitorMCP/main.py"
      ],
      "cwd": "/path/to/ScreenMonitorMCP",
      "env": {
        "OPENAI_API_KEY": "your-api-key-here"
      }
    }
  }
}

🛡️ Secure Configuration

{
  "mcpServers": {
    "screenMonitorMCP": {
      "command": "python",
      "args": [
        "/path/to/ScreenMonitorMCP/main.py",
        "--api-key", "your-secret-key"
      ],
      "cwd": "/path/to/ScreenMonitorMCP"
    }
  }
}

🪟 Windows Example

{
  "mcpServers": {
    "screenMonitorMCP": {
      "command": "python",
      "args": ["C:/path/to/ScreenMonitorMCP/main.py"],
      "cwd": "C:/path/to/ScreenMonitorMCP"
    }
  }
}

⚠️ Important Notes

  1. File Path: Update /path/to/ScreenMonitorMCP/main.py path according to your project directory
  2. Python Path: Make sure Python is in PATH or use full path: "C:/Python311/python.exe"
  3. Working Directory: cwd parameter is important for proper .env file reading
  4. API Keys: All settings are automatically read from .env file

🧪 USAGE EXAMPLES

🔄 Starting Real-Time Monitoring

# Start AI's continuous vision capability
result = await start_continuous_monitoring(
    fps=3,
    change_threshold=0.1,
    smart_detection=True
)

# Check monitoring status
status = await get_monitoring_status()

# View recent changes
changes = await get_recent_changes(limit=10)

# Stop monitoring
await stop_continuous_monitoring()

🎯 Using UI Intelligence

# Analyze all UI elements on screen
ui_elements = await analyze_ui_elements(
    detect_buttons=True,
    extract_text=True,
    confidence_threshold=0.7
)

# Smart clicking with natural language
await smart_click("Click the save button", dry_run=False)

# Extract text from specific region
text_data = await extract_text_from_screen(
    region={"x": 100, "y": 100, "width": 500, "height": 300}
)

🧠 Predictive Intelligence

# Learn user behavior patterns
patterns = await learn_user_patterns()

# Predict user intent
intent = await predict_user_intent(
    current_context={"current_app": "VSCode"}
)

# Get proactive assistance
assistance = await proactive_assistance()

# Record user action
await record_user_action(
    action_type="click",
    target="save_button",
    app_context="VSCode"
)

📸 Traditional Screen Capture

# Enhanced screen capture and analysis
result = await capture_and_analyze(
    capture_mode="all",
    analysis_prompt="What do you see on this screen?",
    max_tokens=500
)

# List all tools
tools = await list_tools()

🚀 REVOLUTIONARY CAPABILITIES

This MCP server gives AI the following capabilities:

  • 👁️ Continuous Vision: AI can monitor the screen non-stop
  • 🧠 Smart Understanding: Recognizes UI elements and interacts with them
  • 🔮 Future Prediction: Learns and predicts user behaviors
  • Proactive Help: Offers help before you need it
  • 🎯 Natural Interaction: Understands commands like "Click the save button"

🔧 TROUBLESHOOTING

Common Issues and Solutions

  1. Unicode/Encoding Error (Windows)

    UnicodeEncodeError: 'charmap' codec can't encode character
    

    Solution: ✅ This error is fixed! Server automatically uses UTF-8 encoding.

  2. JSON Configuration Error

    // ❌ Wrong
    {
      "command": "python",
      "args": ["path/to/main.py",]  // Trailing comma is wrong
    }
    
    // ✅ Correct
    {
      "command": "python",
      "args": ["path/to/main.py"]
    }
    
  3. Python Path Issue

    {
      "command": "C:/Python311/python.exe",  // Use full path
      "args": ["C:/path/to/ScreenMonitorMCP/main.py"]
    }
    
  4. Missing Dependencies

    cd ScreenMonitorMCP
    pip install -r requirements.txt
    
  5. OCR Issues

    # Install Tesseract (optional)
    # EasyOCR installs automatically
    
  6. MCP Connection Closed Error

    MCP error -32000: Connection closed
    

    Solution: Check file paths and add cwd parameter.

📝 LICENSE

This project is licensed under the MIT License.


🚀 Revolutionary MCP server that gives AI real "eyes"! 🔥 Next-generation AI-human interaction starts here!

from github.com/ideook/ScreenMonitorMCP

Installing ScreenMonitorMCP

This server has no published package — it is built from source. Open the repository and follow its README.

▸ github.com/ideook/ScreenMonitorMCP

FAQ

Is ScreenMonitorMCP MCP free?

Yes, ScreenMonitorMCP MCP is free — one-click install via Unyly at no cost.

Does ScreenMonitorMCP need an API key?

No, ScreenMonitorMCP runs without API keys or environment variables.

Is ScreenMonitorMCP hosted or self-hosted?

Self-hosted: the server runs locally on your machine via the install command above.

How do I install ScreenMonitorMCP in Claude Desktop, Claude Code or Cursor?

Open ScreenMonitorMCP on unyly.org, pick your client tab (Claude Desktop, Claude Code, Cursor) and press Install — the config is generated automatically, no JSON editing.

Related MCPs

Compare ScreenMonitorMCP with

Not sure what to pick?

Find your stack in 60 seconds

Author?

Embed badge for your README

Browse similar

All design MCPs