Stealth Interview Assistant

A Python desktop application designed for interview assistance with two main components:

1. Stealth Overlay Window: A semi-transparent, frameless window that's always on top but completely invisible to screen capture software (Zoom, Teams, Google Meet)
2. Dual-Channel Audio Capture: Simultaneous capture of microphone input and system audio output using WASAPI loopback

Features

• 🎯 Screen Capture Exclusion: Uses Windows API (WDA_EXCLUDEFROMCAPTURE) to hide the overlay from screen sharing
• 🎤 Microphone Capture: Records your voice (Channel A)
• 🔊 System Audio Capture: Records interviewer's voice via WASAPI loopback (Channel B)
• 📊 Live Monitor Mode: Console output showing real-time audio capture from both channels
• 🎨 Modern UI: Dark theme overlay with audio status indicators
• 🔄 Dual-Stream Processing: Separate queues for independent audio stream handling

System Requirements

• Operating System: Windows 10/11 (required for WDA_EXCLUDEFROMCAPTURE)
• Python: 3.8 or higher
• Audio: Working microphone and system audio output

Installation

1. Clone or download this repository

2. Install Python dependencies:

pip install -r requirements.txt

The main dependencies are:

• customtkinter - Modern UI framework
• pyaudiowpatch - Extended PyAudio with WASAPI loopback support
• pywin32 - Windows API access
• comtypes - COM interface support

Usage

1. GUI Mode (Default)

Run the application with the stealth overlay window:

python main.py

This will:

• Open a semi-transparent overlay window
• Start capturing audio from both microphone and system audio
• Display live transcription placeholders in the overlay
• Show visual indicators for active audio streams

The overlay window:

• Can be dragged around the screen
• Stays on top of all other windows
• Is invisible to screen capture software
• Shows real-time audio capture status

2. Monitor Mode (Console Only)

Test audio capture without the GUI:

python main.py monitor

This will:

• Print live transcription placeholders to the console
• Show sample counts and data sizes from both channels
• Verify that both audio streams are working correctly
• Display statistics when stopped (Ctrl+C)

Example output:

[MIC] 🎤 Sample #1 | Size: 2048 bytes | [Transcribing...]
[SPK] 🔊 Sample #1 | Size: 4096 bytes | [Transcribing...]
[MIC] 🎤 Sample #2 | Size: 2048 bytes | [Transcribing...]
[SPK] 🔊 Sample #2 | Size: 4096 bytes | [Transcribing...]

3. List Audio Devices

Check available audio devices on your system:

python main.py devices

This will display all audio input/output devices with their properties.

4. Help

Display usage information:

python main.py help

How It Works

Stealth Overlay

The overlay window uses multiple Windows API techniques to remain invisible to screen capture:

1. SetWindowDisplayAffinity(WDA_EXCLUDEFROMCAPTURE): Primary method to exclude window from capture
2. DwmSetWindowAttribute: Additional Desktop Window Manager settings
3. Frameless & Transparent: Custom window styling for minimal visibility
4. Always on Top: Stays visible above all applications

The window is visible on your physical screen but appears as a black rectangle or is completely invisible to anyone viewing your screen through Zoom, Teams, Google Meet, or any other screen sharing software.

Dual-Channel Audio Capture

Uses PyAudioWPatch (PyAudio with WASAPI support) to capture two separate audio streams:

Channel A - Microphone:

• Captures from default input device
• Records your voice
• Mono audio stream
• 16kHz sample rate

Channel B - System Audio (Loopback):

• Uses WASAPI loopback device
• Captures all system audio output
• Records interviewer's voice from the meeting
• Stereo audio stream
• Native sample rate (typically 44.1kHz or 48kHz)

Both streams are captured simultaneously in separate threads with independent queues, allowing for real-time processing without blocking.

Architecture

main.py
├── StealthAssistant (Main Controller)
│   ├── Initializes components
│   ├── Manages threads
│   └── Coordinates GUI and audio
│
stealth_overlay.py
├── StealthOverlay (GUI Component)
│   ├── CustomTkinter window
│   ├── Windows API integration
│   ├── Status indicators
│   └── Transcript display
│
audio_capture.py
└── DualChannelAudioCapture (Audio Component)
    ├── PyAudioWPatch integration
    ├── Device detection
    ├── Dual stream management
    └── Monitor mode

Important Notes

Privacy & Ethics

This tool is designed for personal use in situations where you have the right to record. Be aware of:

• Recording laws in your jurisdiction
• Terms of service of the platform you're using
• Consent requirements for recording conversations

Limitations

• Windows Only: The screen capture exclusion feature requires Windows 10/11
• WASAPI Required: System audio capture needs WASAPI loopback support
• Not 100% Guaranteed: Some screen capture methods might still detect the window
• Performance: Running during video calls may impact system performance

Troubleshooting

Overlay is visible in screen share:

• Ensure you're running Windows 10/11
• Try restarting the application
• Check if your screen sharing software uses non-standard capture methods

No audio captured from speakers:

• Verify system audio is playing
• Check that WASAPI loopback device is available
• Run python main.py devices to list available devices
• Ensure no other application is exclusively using the audio device

No audio captured from microphone:

• Check microphone permissions in Windows Settings
• Verify microphone is working in other applications
• Select correct device if multiple microphones are available

Application crashes on startup:

• Verify all dependencies are installed: pip install -r requirements.txt
• Check Python version (3.8+ required)
• Run in monitor mode first to test audio capture: python main.py monitor

Development

Project Structure

Question Assister/
├── main.py                 # Application entry point
├── stealth_overlay.py      # GUI overlay component
├── audio_capture.py        # Audio capture component
├── requirements.txt        # Python dependencies
└── README.md              # This file

Extending the Application

The modular design allows easy extension:

1. Add Transcription: Integrate speech-to-text APIs (Whisper, Google Cloud, Azure)
2. Add LLM Integration: Send transcripts to GPT/Claude for real-time assistance
3. Add Recording: Save audio streams to WAV files
4. Add Hotkeys: Global keyboard shortcuts for controls
5. Add Configuration: User settings for audio devices, window position, etc.

Code Structure

StealthOverlay (stealth_overlay.py):

• __init__: Initialize window and UI
• _configure_window: Set window properties
• _apply_stealth_settings: Apply Windows API exclusion
• _create_ui: Build interface elements
• update_status: Update status messages
• add_transcript: Add text to transcript box

DualChannelAudioCapture (audio_capture.py):

• find_devices: Auto-detect audio devices
• start_capture: Begin dual-stream capture
• stop_capture: Stop all streams
• monitor_mode: Console monitoring
• get_mic_data/get_speaker_data: Retrieve audio from queues

License

This project is provided as-is for educational purposes. Use responsibly and ethically.

Disclaimer

This software is intended for personal use only. Users are responsible for:

• Complying with local recording laws
• Obtaining necessary consent for recordings
• Following platform terms of service
• Using the software ethically and legally

The developers assume no liability for misuse of this software.