Aura Shield 🛡️

Aura Shield is an AI-powered Android application designed to protect your cognitive load by intercepting and analyzing incoming notifications from social apps like WhatsApp and Instagram. It provides real-time sentiment analysis, stress scoring, and smart reply suggestions to help you maintain digital well-being.

Aura Shield serves as an intelligent, zero-trust agentic UX layer that forms a cognitive buffer between you and your digital ecosystem. By intercepting high-friction notification storms (like WhatsApp and Instagram) and processing them through on-device hardware acceleration, Aura shields human attention from micro-decision fatigue.

Explore Architecture • Key Features • Installation • Privacy Protocol

🚀 Key Features

• Notification Interception: Automatically listens to WhatsApp and Instagram messages using a NotificationListenerService.
• AI Sentiment Analysis: Powered by Gemini Nano to analyze the "vibe" of incoming messages.
• Stress Scoring: Calculates a stress level (1-10) for each message. High-stress messages trigger a visual "Safety Ring" pulse animation.
• Smart Replies: Generates empathetic and context-aware reply suggestions.
• Quick Reply: Direct "Send" buttons within the Aura UI to instantly reply without opening the messaging app.
• Persistence: Uses a foreground service (AuraEmpathyService) to ensure the AI monitoring remains active.
• Debug Console: A built-in log viewer to track raw notification events and AI processing steps.

🛠️ Tech Stack

• Language: Kotlin
• UI: Jetpack Compose
• AI: Gemini Nano (Structured JSON Protocol)
• Background Processing: Foreground Services & Notification Listener API
• State Management: Clean Architecture MVVM layered over modern asynchronous StateFlow primitives
• Background Lifecycle: Persistent Android Foreground Services + System Notification Access APIs
• Architecture: MVVM with StateFlow

📸 How it Works

1. Intercept: Aura detects a notification from a supported app.

2. Analyze: The message content is sent to the AI core.

3. Vibe Check: The AI returns a JSON containing:

   {
     "summary": "Short summary of the message",
     "vibe": "Sentiment (e.g., Aggressive, Calm, Urgent)",
     "reply": "Suggested empathetic reply",
     "stress_score": 8
   }

4. Shield: If stress_score > 7, the UI pulses red, alerting you to take a breath before engaging.

5. Intercept: Aura captures a high-friction notification payload at the system layer.

6. Analyze: The raw text payload is pushed into the localized AI pipeline.

7. Structured Vibe Check: The specialized local engine compiles context details into a unified JSON schema:

   {
     "summary": "Short summary of the message",
     "vibe": "Sentiment (e.g., Aggressive, Calm, Urgent)",
     "reply": "Suggested empathetic reply",
     "stress_score": 8
   }

8. Shield Activation: If the calculated stress_score passes a threshold ($> 7$), the UI shifts dynamically into an empathetic, soft-tone warning state, encouraging a micro-pause before engagement.

---

⚙️ Setup & Installation

Environment Configurations

• Minimum Android Environment: API Level 26 (Android 8.0)
• Target Android Environment: API Level 34 (Android 14)
• Required System Clearances:
  • android.permission.BIND_NOTIFICATION_LISTENER_SERVICE (To parse digital messaging friction)
  • android.permission.POST_NOTIFICATIONS (To preserve service lifecycle stability)

Build Instructions

1. Clone the repository.
2. Open in Android Studio.
3. Sync Gradle and build the :app module.
4. Deploy to a physical device or emulator.
5. Important: Grant "Notification Access" to Aura Shield in your device settings.

🛡️ Privacy

Aura Shield processes message data locally using on-device AI models where possible to ensure your private conversations never leave your device.

---