🚀 MotivateAI

Your Autonomous Consistency Agent

An intelligent learning companion designed to combat the 95% dropout rate in self-directed learning by adapting to your behavioral patterns in real-time.

---

📖 Overview

Self-directed learning is incredibly difficult. Most learners abandon their goals because they try to follow rigid schedules that don't match their actual working style.

MotivateAI flips this model. Instead of forcing you to adapt to a schedule, it uses Gemini 2.5 Flash and Behavioral Analytics to adapt the schedule to you. It breaks monumental goals into 10-30 minute micro-tasks, monitors your completion rates and break patterns, and dynamically generates future sessions tailored to when and how you learn best.

✨ Core Features

• 🧠 Adaptive AI Session Generation: Uses Gemini 2.5 Flash to deconstruct massive learning goals into highly specific, actionable micro-tasks.
• 📊 Behavioral Analytics Engine: Tracks your focus duration, preferred break times, and drop-off signals to optimize your learning schedule automatically.
• 🎥 Targeted YouTube Integration: Automatically sources short, hyper-relevant YouTube tutorials for your current micro-task to prevent context switching.
• 🔒 Secure Authentication & Identity: Built with custom JWT-based authentication and Next.js Edge Proxy for strict access control and IDOR prevention.
• 🎨 Modern, Responsive UI: A sleek, glassmorphic dark-mode interface built with Tailwind CSS.

🏛️ System Architecture

graph TD
    subgraph Frontend ["Frontend (Client Browser)"]
        UI[React / Tailwind CSS UI]
    end

    subgraph Vercel ["Vercel Serverless Platform"]
        subgraph Edge ["Edge Network"]
            Proxy{Next.js Edge Proxy}
            JWT[JWT Verification via 'jose']
            Proxy <--> JWT
        end
        
        subgraph NodeRuntime ["Node.js Serverless Functions"]
            AppRouter[Next.js 16 App Router]
            AuthAPI[Auth Routes & bcryptjs]
            AgentAPI[Adaptive Session Generator]
            BehaviorAPI[Behavioral Analytics Engine]
            YouTubeRoute[YouTube Integration Route]
        end
        
        Proxy -->|Valid Requests| AppRouter
        AppRouter --> AuthAPI
        AppRouter --> AgentAPI
        AppRouter --> BehaviorAPI
        AppRouter --> YouTubeRoute
    end
    
    subgraph External ["External APIs & Databases"]
        MongoDB[(MongoDB Atlas)]
        Gemini[Google Gemini 2.5 Flash]
        YouTube[YouTube Data API v3]
    end

    UI -->|HTTPS| Proxy
    AuthAPI <-->|Read/Write| MongoDB
    AgentAPI <-->|Prompt/Response| Gemini
    BehaviorAPI <-->|Track Progress| MongoDB
    YouTubeRoute <-->|Search Videos| YouTube

Loading

🛠️ Technology Stack

• Frontend & Backend Framework: Next.js 16 (App Router)
• Language: TypeScript
• Database: MongoDB (Atlas)
• AI Integration: @google/generative-ai (Gemini 2.5 Flash)
• Authentication: jose (JWT Cryptography), bcryptjs (Password Hashing)
• Styling: Tailwind CSS

---

🚀 Getting Started (Local Development)

1. Prerequisites

• Node.js 18+ installed
• A MongoDB Atlas cluster (free tier works perfectly)
• A Google Gemini API Key

2. Installation

Clone the repository and install dependencies:

git clone https://github.com/aaminashihab/MotivateAI.git
cd MotivateAI
npm install

3. Environment Variables

Create a .env.local file in the root directory and add the following keys:

# MongoDB Connection String (must allow 0.0.0.0/0 in Network Access)
MONGODB_URI=mongodb+srv://<username>:<password>@cluster0...

# Cryptographically secure random string for signing JWTs
# Run `node -e "console.log(require('crypto').randomBytes(32).toString('hex'))"` to generate one
JWT_SECRET=your_super_secret_jwt_key_here

# Your Google Gemini API Key
GEMINI_API_KEY=your_gemini_api_key_here

# YouTube Data API v3 Key (Optional: for fetching video tutorials)
YOUTUBE_API_KEY=your_youtube_api_key_here

4. Run the Development Server

npm run dev

Open http://localhost:3000 in your browser to see the application.

---

🔒 Security Architecture & Known Limitations

As a portfolio project designed to demonstrate core product velocity alongside engineering maturity, several explicit architectural tradeoffs were made:

• Edge-Level Authorization: The application utilizes a Next.js proxy.ts (Edge Middleware) to verify JWT signatures on all protected routes. The application is strictly configured to throw a fatal 500 Internal Server Error if JWT_SECRET is missing in production, preventing dangerous unauthenticated fallbacks.
• IDOR Mitigations: Client-provided userId parameters in URLs (e.g., /api/users/[userId]/preferences) are explicitly ignored for authorization. Access control is verified strictly against the server-derived x-user-id header injected by the JWT middleware.
• In-Memory Rate Limiting: The AI and YouTube API routes currently utilize an in-memory Map for rate limiting. In a Vercel serverless environment, this state does not persist across cold starts. A production deployment would swap this for a centralized Redis store (e.g., Upstash).
• Prompt Injection: Agent endpoints employ regex-based filtering to strip basic prompt injection attacks (e.g., "ignore previous instructions"). Enterprise-grade deployment would require output-side validation and strictly scoped, read-only database roles.
• Password Reset Flow: For demonstration purposes, reset links are printed to the console rather than dispatched via an email provider (like Resend).

---

📜 License

This project is licensed under the MIT License.

---

Built with ❤️ for learners everywhere.