DIGITAL CHEATSHEET

A digital cheatsheet tool to access reference materials during tests. Integrates with Gemini AI for enhanced assistance when connected to the internet.

Has bypassed

• All websites that implement full-screen enforcement and keyboard restrictions (e.g., Infosys Springboard and similar platforms)
• The Hackerearth proctoring browser. Only the keystroke recording method works for this, but I would recommend connecting to a remote computer of yours over VNC. They cannot detect that, apparently.

However

• You shouldn't use it if you are full-screen sharing and they can see your screen or periodically take screenshots. You can create a bypass for that by combining the keystroke recording method (3) and the clipboard pasting method (2). That way, the GUI should never pop up. You will have to modify that yourself, I haven't done it.

AI Features & Keyboard Shortcuts

⚠️ API Key Required: To use AI features, you must have a Gemini API key. Add it to a .env file in the project root.

It doesn't matter if you haven't studied the test, but study these keybinds- or the ones that matter to you!

AI-Powered Shortcuts

Key	Function	Description
Right Arrow	AI Clipboard Analysis	Analyzes text from your clipboard using Gemini AI
0	Screen Analysis	Takes a screenshot and sends it to AI for general question analysis
9	Math Question Analysis	Optimized screenshot analysis for mathematical questions
8	Programming Question Analysis	Optimized screenshot analysis for code questions (defaults to Python)
7	Multiple Choice Analysis	Optimized screenshot analysis for multiple-choice questions
1	Toggle AI Model	Switches between available Gemini models
3	Keystroke Recording	Records keystrokes until = is pressed, then analyzes the captured text

Standard Shortcuts

Key	Function	Description
Up Arrow	Toggle Hotkeys	Enables/disables all keyboard shortcuts
Left Arrow	Search Notes	Opens a search bar to find content in your reference materials
Down Arrow	Exit Program	Closes the application (press additional keys after to ensure exit)
2	Type Clipboard	Types out the current clipboard contents via keyboard simulation
A-Z and others	Access Notes	Various keys mapped to specific reference materials in secret_messages
-	Toggle window visibility	Hides / Shows the hidden window

Configuration Options

Customizing Content and Appearance

Setting	Description	Location
Reference Materials	Modify the secret_messages dictionary	Edit in digital_cheatsheet.py
AI Integration	Add your Gemini API key	Create .env file with GEMINI_API_KEY=your_key_here
Voice To Text	Add your WhisperAPI key	Create .env file with WHISPER_API_KEY=yourkeyhere
UI Customization	Change application background color	Modify self.frame = tk.Frame(self.root, bg='white')
AI Prompts	Customize AI interaction prompts	Edit the PROMPT variables in various AI methods

Stealth Recommendations

• Set application background to match your test paper colour (defaults to white). Replace all instances of 'white' in the code to 'black' or whatever you want.
• Adjust transparency with self.root.attributes('-alpha', 0.8) (lower values = more transparent)
• By default, it's disguised as File Explorer, you can change that by changing the icon and the file name.

Installation & Setup

Prerequisites

• Python 3.7 or higher
• Pip package manager

If you don't know how, just type "python" into your command bar and it will take you to the microsoft store page to install it.

Installing Dependencies

Type this into a console window that's in the same directory as this project or the VSCode terminal.

python -m pip install -r requirements.txt

Environment Setup

1. Create a .env file in the root directory
2. Add your Gemini API key: GEMINI_API_KEY=your_key_here
3. If using Audio Analyzer, add: WHISPER_API_KEY=your_key_here

Building the Application

Creating a Standalone Executable

# Windows
pyinstaller --noconfirm --onefile --windowed --icon "./fileexplorericon.ico" --name "File Explorer" "./digital_cheatsheet.py"

# macOS/Linux (adjust icon format if needed)
pyinstaller --noconfirm --onefile --windowed --icon "./fileexplorericon.ico" --name "File Explorer" "./digital_cheatsheet.py"

Audio Analyzer Component

The project includes a separate audio analysis utility (audio_analyzer.py) application to cheat on listening tests.

Overview

I made this app in about 30 minutes using AI. It uses 2 APIs, both Gemini and LemonFox.ai. Lemonfox is a paid service. You can modify the code to use whisper locally, it's free. However, processing will take longer if you don't have a GPU. You can simply modify the transcribe_mp3 function. Make an issue if you don't know where to start, I'll gladly help.

Technical Details

Feature	Description
Audio Input	Supports direct MP3 URLs or local MP3 files
Image Analysis	Supports clipboard pasting or file upload for question images
Transcription	Uses LemonFox.ai API (paid service) for high-quality audio transcription
Alternative Setup	Can be modified to use local Whisper model for free transcription
Response Quality	Optimized for educational content and listening comprehension tests

Usage Instructions

1. Launch the audio analyzer: python audio_analyzer.py
2. Enter the MP3 URL or file path in the "Audio Source" field
3. Paste (Ctrl+V) or upload question images using the "Upload Image" button
4. Click "Analyze" to process the audio and generate answers
5. Review the AI-generated responses in the "Analysis Results" section

Note

This component functions as a standalone desktop application rather than a discreet overlay. It's designed for situations where you need to quickly analyze listening test materials rather than for use during supervised tests.

Cheating

I don't condone cheating but there is no point of studying something that is barely related to your course. I will never support the Indian Education System.