NovaMind — AI Chatbot

A sleek, premium AI chatbot powered by the Google Gemini API — featuring a beautiful black & white design system, dynamic API key load-balancing, a complete authentication overlay, Progressive Web App (PWA) offline support, text-to-speech, code copying, and rich media generation.

🌐 Live Website: https://aawaraa.netlify.app/

---

✨ Features

• 🎭 AI Personas / Modes: Switch seamlessly between distinct personalities: Default Assistant, Coding Expert, Creative Writer, and Sarcastic Assistant.
• 🧠 Gemini 3.6 / 3.7 Flash & Auto-Fallback: Powered by the latest gemini-3.6-flash model with intelligent discovery and auto-fallback to prevent downtime.
• 🔐 Authentication Overlay Lockout: The interface is fully locked until users sign up, log in, or choose to enter via the Guest Profile.
• 🗂️ Isolated Session Histories: Logged-in accounts have conversations securely stored and loaded under novamind_sessions_${email}, while guest users run isolated sessions under novamind_sessions_guest.
• 🔑 Zero-Config Key Entry for Users: Users start chatting immediately without needing to configure or supply their own API keys.
• 🔀 API Key Load Balancing & Rotation: Rotates requests randomly across multiple developer-configured Gemini keys in config.js to multiply rate limits (3 keys scale the limit to 45 RPM and 4,500 RPD).
• 🛡️ Demo Token Quota & Live Metrics: Real-time tracker for RPM, RPD, session tokens, and quota consumption in the sidebar.
• 📧 Gmail OTP Reset Simulation: Support for forgotten passwords. Triggers a random 6-digit OTP code, popped up on-screen in a simulated Gmail notification window.
• 🗣️ Text-to-Speech (TTS): Built-in speaker buttons on bot messages, powered by the Web Speech API. Removes markdown markup for smooth, clean read-aloud playback.
• 📋 Code Copying: Renders code fences with headers containing the language name and a click-to-copy utility.
• 📢 Conversation Sharing: Modal triggers to export chat transcript logs directly to WhatsApp, Instagram, Email, or via the Native Mobile Share API.
• 🎨 Image Generation: Detects image generation intent (e.g., "draw/create a picture") and outputs beautifully formatted images rendered via Pollinations.ai.
• 📍 Smart On-Demand Location: Queries browser geolocation permission only when the chat context specifically demands it (e.g. weather, nearby places).
• 📱 Progressive Web App (PWA): Installable application shell with network-first service worker rules for instant updates and offline support.

---

🚀 Getting Started

1. Clone the repo

git clone https://github.com/komallbarhate/Chatbot.git
cd Chatbot

2. Configure API Keys

Copy config.example.js to create config.js:

cp config.example.js config.js

Open config.js and add one or more Gemini API keys:

const GEMINI_API_KEYS = [
  "YOUR_GEMINI_API_KEY_1",
  "YOUR_GEMINI_API_KEY_2"
];
const GEMINI_API_KEY = GEMINI_API_KEYS[0];

(Get keys at → Google AI Studio)

3. Run Locally

Simply open index.html in any browser or launch a local preview server:

# Using python simple server
python -m http.server 8000

---

🗂️ Project Structure

Chatbot/
├── index.html        # Locking login overlays, notifications & chat shell
├── style.css         # Dark monochrome design system
├── chatbot.js        # Auth controllers, dynamic history loading, model fallback, and rotation
├── config.js         # Secret key container (gitignored)
├── config.example.js # Template configuration file
├── manifest.json     # PWA setup metadata
├── sw.js             # Service worker caches (v6 Network-First)
└── icon.svg          # High-resolution vector logo

---

📡 API Limits (With 3 Rotating Keys)

Metric	Single Key	3 Rotating Keys
Requests Per Minute	15 RPM	45 RPM
Requests Per Day	1,500 RPD	4,500 RPD
Max Token Limit	10M tokens / day	30M tokens / day

---

Made with ♥ by Komal