MAC - LINUX SUPPORT

I have only tested that WINDOWS 11 operating system works, Feel free to test mac and linux to check if they work properly with the current version. If you are experiencing issues with your build:

try building locally with electron builder. open a issue for any bugs you face. ; )

The codebase is designed to be adaptable:

• AI Models: Currently supports Groq API. You can modify the code to integrate with other providers like Deepseek, OpenAI, or any model with an API. All integration code is in electron/ProcessingHelper.ts and UI settings are in src/components/Settings/SettingsDialog.tsx.
• Languages: Add support for additional programming languages
• Features: Extend the functionality with new capabilities
• UI: Customize the interface to your preferences

All it takes is modest JavaScript/TypeScript knowledge and understanding of the API you want to integrate.

• 🎯 Invisible Overlay: Undetectable window that bypasses most screen capture methods

• 📸 Smart Screenshot Capture: Capture questions, code, and web designs with a single keystroke

• 🤖 AI Support: Works with Groq API multimodal vision models

• 💡 Auto-Detection: Automatically detects question type (MCQ, Python, Web Dev, Text)

• ⚡ Quick Answer Mode: One-key solution - capture and process automatically (Ctrl+D)

• 🎨 Web Development: Generates complete HTML/CSS matching design screenshots

• 📋 Smart Copy: Separate copy for HTML and CSS, plus auto-type clipboard feature

• 🐛 Conversational Debugging: Fix errors by sending additional screenshots

• 🎨 Advanced Window Management: Freely move, resize, change opacity, and zoom the window

• 🔄 Vision Model: Uses Groq Llama 4 Maverick for fast multimodal analysis

• 📋 Code Clipboard: Copy generated solutions directly to clipboard

• 🗑️ Screenshot Management: Delete screenshots with keyboard shortcuts

• 🔒 Privacy-Focused: Your API key and data never leave your computer except for API calls

The application uses undetectable global keyboard shortcuts that won't be detected by browsers or other applications:

• Toggle Window Visibility: Ctrl/Cmd + B or Ctrl/Cmd + I or Alt + 1
• Take Screenshot: Ctrl/Cmd + H or Ctrl/Cmd + M
• Process Screenshots: Ctrl/Cmd + Enter
• Start New Problem: Ctrl/Cmd + R
• Quit Application: Ctrl/Cmd + Q

• Delete Last Screenshot: Ctrl/Cmd + Backspace
• Copy Code to Clipboard: Ctrl/Cmd + Shift + C

• Move Left: Ctrl/Cmd + Left Arrow
• Move Right: Ctrl/Cmd + Right Arrow
• Move Up: Ctrl/Cmd + Up Arrow
• Move Down: Ctrl/Cmd + Down Arrow

• Decrease Opacity: Ctrl/Cmd + [
• Increase Opacity: Ctrl/Cmd + ]
• Zoom Out: Ctrl/Cmd + -
• Reset Zoom: Ctrl/Cmd + 0
• Zoom In: Ctrl/Cmd + =

The application is invisible to:

• Zoom versions below 6.1.6 (inclusive)
• All browser-based screen recording software
• All versions of Discord
• Mac OS screenshot functionality (Command + Shift + 3/4)

Note: The application is NOT invisible to:

• Zoom versions 6.1.6 and above
  • https://zoom.en.uptodown.com/mac/versions (link to downgrade Zoom if needed)
• Mac OS native screen recording (Command + Shift + 5)

• Node.js (v16 or higher)
• npm or bun package manager
• Groq API Key
• Screen Recording Permission for Terminal/IDE
  • On macOS:
    1. Go to System Preferences > Security & Privacy > Privacy > Screen Recording
    2. Ensure that CheatSheet AI has screen recording permission enabled
    3. Restart CheatSheet AI after enabling permissions
  • On Windows:
    • No additional permissions needed
  • On Linux:
    • May require xhost access depending on your distribution

• Ctrl+D - Quick Answer (Reset → Capture → Process all in one)
• Ctrl+H - Take Screenshot
• Ctrl+Enter - Process Screenshots
• Ctrl+R - Reset/Clear Everything

• Ctrl+B or Alt+1 - Toggle Window Visibility
• Ctrl+I - Toggle Window Visibility (alias)
• Ctrl+Q - Quit Application

• Ctrl+Arrow Keys - Scroll through content
• Ctrl+[ - Decrease Opacity
• Ctrl+] - Increase Opacity
• Ctrl+- - Zoom Out
• Ctrl+0 - Reset Zoom
• Ctrl+= - Zoom In

• Ctrl+Left - Move Window Left
• Ctrl+Right - Move Window Right
• Ctrl+Up - Move Window Up
• Ctrl+Down - Move Window Down

• Uses Groq Llama 4 Maverick vision model

• Ctrl+Shift+C - Copy HTML/Code to Clipboard
• Ctrl+Shift+D - Copy CSS to Clipboard (extracts from <style> tags)
• Ctrl+Shift+V - Type Clipboard Content (bypasses paste restrictions, faster speed)
• *esc - Stop Typing (interrupts clipboard typing)

• Ctrl+M - Take Screenshot (alias for Ctrl+H)
• Ctrl+Backspace - Delete Last Screenshot

1. Clone the repository:

git clone https://github.com/greeneu/cheatsheet-ai.git
cd cheatsheet-ai

2. Install dependencies:

npm install

3. RECOMMENDED: Clean any previous builds:

npm run clean

4. Run the appropriate script for your platform:

For Windows:

stealth-run.bat

For macOS/Linux:

# Make the script executable first
chmod +x stealth-run.sh
./stealth-run.sh

IMPORTANT: The application window will be invisible by default! Use Ctrl+B (or Cmd+B on Mac) to toggle visibility.

To create installable packages for distribution:

For macOS (DMG):

npm run package-mac

For Windows (Installer):

npm run package-win

The packaged applications will be available in the release directory.

What the scripts do:

• Create necessary directories for the application
• Clean previous builds to ensure a fresh start
• Build the application in production mode
• Launch the application in invisible mode

• Window Manager Compatibility: Some window management tools (like Rectangle Pro on macOS) may interfere with the app's window movement. Consider disabling them temporarily.

• API Usage: Be mindful of your API key's rate limits and credit usage. Vision API calls are more expensive than text-only calls.

• LLM Customization: You can easily customize the app to include additional LLMs by modifying the API calls in ProcessingHelper.ts and related UI components.

• Common Issues:

  • Run npm run clean before starting the app for a fresh build
  • Use Ctrl+B/Cmd+B multiple times if the window doesn't appear
  • Adjust window opacity with Ctrl+[/] or Cmd+[/] if needed
  • For macOS: ensure script has execute permissions (chmod +x stealth-run.sh)

• Electron
• React
• TypeScript
• Vite
• Tailwind CSS
• Radix UI Components
• Groq API

1. Initial Setup

   • Launch the invisible window
   • Enter your Groq API key in the settings
   • Choose your preferred AI provider and model for extraction, solution generation, and debugging

2. Capturing Problem

   • Use global shortcut Ctrl/Cmd + H (or Ctrl/Cmd + M) to take screenshots of code problems
   • Screenshots are automatically added to the queue (up to 2)
   • Remove the last screenshot with Ctrl/Cmd + Backspace if needed

3. Processing

   • Press Ctrl/Cmd + Enter to analyze the screenshots
   • AI extracts problem requirements from the screenshots using Vision API
   • The model generates an optimal solution based on the extracted information
   • All analysis is done using your personal API key

4. Solution & Debugging

   • View the generated solutions with detailed explanations
   • Copy code to clipboard with Ctrl/Cmd + Shift + C
   • Use debugging feature by taking more screenshots of error messages or code
   • Get structured analysis with identified issues, corrections, and optimizations
   • Toggle between solutions and queue views as needed

5. Window Management

   • Move window using Ctrl/Cmd + Arrow keys
   • Toggle visibility with Ctrl/Cmd + B or Ctrl/Cmd + I
   • Adjust opacity with Ctrl/Cmd + [ and Ctrl/Cmd + ]
   • Zoom in/out with Ctrl/Cmd + = and Ctrl/Cmd + -, reset with Ctrl/Cmd + 0
   • Window remains invisible to specified screen sharing applications
   • Start a new problem using Ctrl/Cmd + R

6. Model

   • Uses Groq Llama 4 Maverick vision model
   • Your preferences are saved between sessions

7. Language Selection

   • Easily switch between programming languages with a single click
   • Use arrow keys for keyboard navigation through available languages
   • The system dynamically adapts to any languages added or removed from the codebase
   • Your language preference is saved between sessions

This application is built with extensibility in mind. You can easily add support for additional LLMs alongside the existing integrations:

• Uses Groq API Llama 4 Maverick vision model
• The application architecture allows for multiple LLM backends to coexist
• Users have the freedom to choose their preferred AI provider

To add new models, simply extend the API integration in electron/ProcessingHelper.ts and add the corresponding UI options in src/components/Settings/SettingsDialog.tsx. The modular design makes this straightforward without disrupting existing functionality.

• API Key: Your personal API key is stored locally and only used for API calls to your chosen provider
• AI Provider: Uses Groq API
• Model Selection: You can choose different models for each stage of processing:
  • Problem Extraction: Analyzes screenshots to understand the coding problem
  • Solution Generation: Creates optimized solutions with explanations
  • Debugging: Provides detailed analysis of errors and improvement suggestions
• Language: Select your preferred programming language for solutions
• Window Controls: Adjust opacity, position, and zoom level using keyboard shortcuts
• All settings are stored locally in your user data directory and persist between sessions

This project is licensed under the GNU Affero General Public License v3.0 (AGPL-3.0).

• You are free to use, modify, and distribute this software
• If you modify the code, you must make your changes available under the same license
• If you run a modified version on a network server, you must make the source code available to users
• We strongly encourage you to contribute improvements back to the main project

See the LICENSE-SHORT file for a summary of terms or visit GNU AGPL-3.0 for the full license text.

We welcome contributions! Please see our Contributing Guidelines for more information.

This tool is intended as a learning aid and practice assistant. While it can help you understand problems and solution approaches during interviews, consider these ethical guidelines:

• Be honest about using assistance tools if asked directly in an interview
• Use this tool to learn concepts, not just to get answers
• Recognize that understanding solutions is more valuable than simply presenting them
• In take-home assignments, make sure you thoroughly understand any solutions you submit

If you have questions or need support, please open an issue on the GitHub repository.

Remember: This is a community resource. If you find it valuable, consider contributing rather than just requesting features. The project grows through collective effort, not individual demands.