🚀 Momentum: Adaptive Procrastination Agent

Break the cycle of avoidance with adaptive, context-aware micro-tasks.

Procrastination often stems from tasks feeling too big or overwhelming. Momentum is an intelligent agent designed to bridge the gap between intention and action by breaking down complex goals into hyper-approachable micro-steps that adapt to your success and energy levels.

---

✨ Key Features

• 🧠 Intelligent Decomposition: Uses the Groq Llama 3.3 model to break any task into actionable, logical steps.
• ⚡ Adaptive Feedback Loop:
  • Done ✓: Move seamlessly to the next logical step.
  • Too Hard 🔨: If a step still feels overwhelming, the agent breaks that specific step into even tinier micro-tasks.
• 🕒 Context-Aware Reasoning: Adjusts task sizing (Small/Medium) based on time of day (e.g., smaller tasks for late evenings when energy is lower).
• 💾 Memory System: Logs every interaction in a local SQLite database to track progress and identify patterns.
• 🎨 Modern UI: A responsive React interface styled with Tailwind CSS v4, featuring smooth animations and a structured output layout for maximum clarity.

---

🛠️ Technology Stack

• Backend: Python 3.10+, Flask (REST API), SQLite3, python-dotenv
• AI Core: Groq SDK (Llama 3.3 70B Versatile)
• Frontend: React 19, Vite, Tailwind CSS v4, PostCSS
• Styling: Modern gradients, glassmorphism-inspired components, and responsive grid layouts.

---

📂 Project Structure

adaptive-procrastination-agent/
├── src/                     # Backend Logic
│   ├── app.py               # Flask API & Session Flow
│   ├── groq_client.py       # AI Orchestrator
│   ├── decision_engine.py   # Context Logic
│   ├── db.py                # SQL Storage
│   └── demo_agent.py        # Terminal Demo
├── frontend/                # React Web App
│   ├── src/                 # Components & Styles
│   ├── vite.config.js       
│   └── package.json         
├── prompts/                 # Prompt Engineering Templates
│   ├── decompose.txt        # Task break-down logic
│   └── phrase.txt           # Persona & Tone logic
├── data/                    # Local Storage (SQLite DB)
├── README.md                # Documentation
└── requirements.txt         # Python Dependencies

---

🚀 Getting Started

1. Prerequisites

• Python 3.10+
• Node.js 18+
• Groq API Key (Get yours at console.groq.com)

2. Backend Setup

# Clone the repository
git clone https://github.com/your-username/momentum-agent.git
cd momentum-agent

# Install dependencies
pip install -r requirements.txt

# Configure Environment
# Create a .env file in the root directory
# Add: GROQ_API_KEY=your_key_here

3. Frontend Setup

cd frontend
npm install
npm run dev

4. Run the Application

In one terminal, start the Python server:

python src/app.py

In another terminal, start the Vite dev server:

cd frontend && npm run dev

Open your browser to http://localhost:5173.

---

🤖 Agent Contract

• Observation: The agent observes the user's task input, category, and current time context.
• Reasoning: It determines if the user is in a "high energy" (Day) or "low energy" (Evening) window and adjusts task complexity accordingly.
• Decision: Orchestrates multiple LLM calls to decompose tasks and rephrase them into encouraging, actionable language.
• Learning: Records success/struggle feedback via the "Done" and "Too Hard" buttons to maintain user momentum.

---

📝 License

Distributed under the MIT License. See LICENSE for more information.