NailBiteMenu

A menu bar app that uses your Mac’s camera to detect nail-biting in real time, locally and privately.
⬇️ Download for macOS (DMG)

Table of Contents

• Open Source
• Installation
• Usage
  • Controls
  • Statistics
  • Threshold
  • Model Size
• Alerts
• Privacy
• Requirements
• Feedback

Open Source

The full NailBiteMenu macOS app, including its Swift source and Core ML models, is now open source. You can explore, build, or contribute to it right here:

📂 /app → contains all app source code and models

💡 Pull requests and issues are welcome!

Installation

1. Download the latest release:
   👉 NailBiteMenu.dmg

2. Open the DMG and drag NailBiteMenu.app into your Applications folder.

3. On first launch, macOS may say it’s from an unidentified developer. Go to System Settings → Privacy & Security → Open Anyway.

4. NailBiteMenu now runs quietly from your menu bar.

Tip: If your menu bar is crowded, macOS may hide the icon behind the chevron (⌃). Drag it out from Control Center to pin it.

Usage

Launching the app from Applications shows a Home window that offers quick actions:

• Open Settings: adjust defaults, mute alerts, pick model size.
• Collapse to Menu Bar: close the window while keeping monitoring active.

Controls

Click the status icon to open the compact control panel. You can:

• Start / Pause monitoring
• Adjust detection threshold
• Choose model size (speed vs. accuracy)
• Toggle alerts and debug stats

Monitoring uses the built-in camera at a modest frame rate, keeping CPU usage low while still catching gestures.

Statistics

The Statistics view (accessible via the main window) provides a detailed breakdown of your habit trends over time.

• Range Selection: Quickly view data for the Last 7 Days, Last 30 Days, Last 90 Days, or All Time.
• Key Metrics: See your total detections, daily average, and trends.
• Visualization: Charts display detection frequency by Day, Weekday, and Hour to help you identify your most active times for biting.
• Data Management: The settings section allows you to permanently reset all tracking data if you want a fresh start.

Detection count is stored locally.

Threshold

The threshold controls how sensitive detection is:

Threshold	Behavior
Low (≈45%)	Very sensitive, may flag normal movements
Balanced (≈60%)	Recommended everyday setting
High (≈75%+)	Only triggers on strong evidence

Tune it based on lighting, distance to the camera, and how early you want alerts.

Model Size

Each bundled model trades speed for precision:

Model	Description	Speed	Accuracy
224 px	Fastest, lowest power draw	⚡️⚡️⚡️	⭐️⭐️⭐️
384 px	Balanced	⚡️⚡️	⭐️⭐️⭐️⭐️
512 px	Most precise (higher CPU)	⚡️	⭐️⭐️⭐️⭐️⭐️

Start with 512 px on Apple Silicon Macs. On Intel Macs, 224 px keeps temps and fans quieter.

Alerts

When nail-biting is detected:

• An overlay appears with the detection message, countdown, and “Stay mindful” button.
• A macOS notification can sound if alerts aren’t muted.

Overlays auto-dismiss after three seconds, or immediately when you press Enter or click the button.

Privacy

• All processing happens on-device. The camera feed never leaves your Mac.
• No network calls, analytics, or cloud services.
• You stay in control — pause monitoring any time.

Requirements

• macOS 15.6 Sequoia or later  
• Camera permission (requested on first use)

Feedback

Spotted a bug? Have an idea?

→ Open an issue

  Made on macOS • © 2025 killian31