Sentinel AI - Universal Enterprise Intelligence Assistant 🛡️

Version: 2.0-Universal | Status: ✅ PRODUCTION READY

A modern, ChatGPT-style enterprise AI assistant powered by Google's Gemini 2.5-flash with universal intelligence capabilities. Now includes Admin Dashboard, VS Code-style IDE, complete data persistence, and Infosys Topaz-like functionality.

🎯 What's New in v2.0

✨ MAJOR UPGRADE (Feb 26, 2026)

• ✅ Persistent Database - All chats stored forever (never lost)
• ✅ Offline Protection - Server-enforced, no caching possible
• ✅ Universal AI - Not just security (full stack, ML, DevOps, etc.)
• ✅ Admin Dashboard - Real-time logs, statistics, activity monitoring
• ✅ VS Code IDE - Professional code editor with Sentinel AI analysis
• ✅ Forensic Logging - Complete audit trail of everything
• ✅ Multi-Provider Ready - Architecture supports any AI provider

📚 Documentation

• 📖 QUICK_START_v2.0.md - Get started in 5 minutes
• 📖 UPGRADE_v2.0.md - Complete feature documentation
• 📖 DEPLOYMENT_COMPLETE.md - Technical details
• 📖 ADMIN_HOST_IMPLEMENTATION.md - Separate admin host summary
• 📖 ADMIN_HOST_SETUP.md - Local & Docker setup guide
• 📖 ADMIN_HOST_PRODUCTION.md - Enterprise deployment patterns

✨ Current Features

Chat & AI

• 💬 Real-time chat with universal AI (like ChatGPT/Claude)
• 🎨 Modern ChatGPT-style UI
• 📝 Full markdown with code highlighting
• 🖼️ Image upload & analysis
• 📋 Complete chat history (persistent)

Intelligence Capabilities

• 💻 Full-stack development - Any programming language
• ☁️ Cloud computing - AWS, Azure, GCP, Kubernetes, Docker
• 🤖 AI & Machine Learning - TensorFlow, PyTorch, Scikit-learn
• 🏗️ System Design - Architecture, microservices, APIs
• 🔐 Security - Encryption, OAuth, MFA, penetration testing
• 📊 Data Science - Analytics, SQL, NoSQL optimization
• 🚀 DevOps - CI/CD, Infrastructure, Container orchestration
• And literally anything else (not limited to security)

Admin Features

• 📊 Admin Dashboard - Real-time statistics & monitoring
• 📋 Activity Logs - Complete forensic audit trail
• 🔍 Advanced Filtering - Search & filter logs
• 📍 IP Tracking - Every action tracked with IP address

Developer Tools

• 💻 Code IDE - VS Code-style editor
• 🔍 Code Analysis - Real-time error detection
• 🤖 Sentinel AI Integration - Analyze code for bugs & improvements
• 📸 Multi-language - Python, JavaScript, Java, Go, Rust, etc.

🚀 Getting Started

Quick Start (5 minutes)

# 1. Navigate to project
cd gemini-ai-chat

# 2. Start server
python app.py

# 3. Access services
Chat:   http://localhost:5000/dashboard
Admin:  http://localhost:5000/admin/dashboard
IDE:    http://localhost:5000/ide

Login

• Google OAuth - Instant login
• Email/Password - Create account
• Phone + OTP - SMS verification

Test Offline Protection

✅ Server running → Chat works
❌ Stop server → "Server offline" error appears
❌ No cached responses, no offline mode

Verify Data Persistence

1. Chat something
2. Stop server (Ctrl+C)
3. Start server again
4. Login again
5. 🎉 Chat history still there!

📁 Project Structure

gemini-ai-chat/
├── app.py                           # Flask backend (UPDATED)
├── sentinel_ai.db                   # SQLite database (persistent)
├── requirements.txt                 # Dependencies
├── api.txt                         # Gemini API keys
├── templates/
│   ├── index.html                 # Main chat interface
│   ├── login.html                 # Login page
│   ├── mfa-setup.html             # MFA configuration
│   ├── admin_dashboard.html       # Admin panel (NEW)
│   └── ide.html                   # Code editor IDE (NEW)
├── README.md                        # This file
├── QUICK_START_v2.0.md             # Quick start guide (NEW)
├── UPGRADE_v2.0.md                 # Feature documentation (NEW)
└── DEPLOYMENT_COMPLETE.md          # Deployment details (NEW)

⚙️ Configuration

Set up your environment variables before running the server:

• SECRET_KEY – Flask secret key for session signing and JWTs.
• OPENAI_API_KEY – API key for the Gemini/OpenAI model you wish to use.
• ADMIN_HOST – optional host (or host:port) where the admin dashboard will be served. When provided, the /admin routes are restricted to requests originating from this host and a separate login page is shown. Useful for running the admin UI on a different domain/port for improved security.

🔐 Architecture

Browser (Cache-Free)
    ↓
Health Check → /api/health (forced server check)
    ↓
Authentication → JWT + Cookies
    ↓
API Requests → No-Cache Headers Applied
    ↓
Sentinel AI Engine (Universal)
    ↓
Gemini 2.5-flash API (with fallback)
    ↓
Database → Permanent Storage
    ↓
Forensic Logs → Complete Audit Trail

📊 Database

Persistent Storage - Nothing Gets Lost

sentinel_ai.db (SQLite)
├── chat_sessions       - Stored conversations
├── chat_messages       - Individual messages (PERMANENT)
├── forensic_logs       - Activity logs (PERMANENT)
├── iam_users          - User accounts
├── iam_audit_logs     - Access logs
└── [+5 more tables]

Retention Policy:

• ⏰ Chat sessions - Until manually deleted
• ⏰ Messages - Until manually deleted
• ⏰ Logs - Until manually deleted
• ⏰ User data - Until account deleted

🔌 API Endpoints

New Endpoints (v2.0)

GET  /api/health            - Server health check
GET  /api/chat-sessions     - Load all chat history
GET  /api/logs              - Get forensic logs
GET  /admin/dashboard       - Admin panel
GET  /ide                   - Code editor IDE

Existing Endpoints

POST /api/chat              - Send message to AI
GET  /api/user              - Get user profile
POST /auth/google-callback  - Google OAuth
POST /auth/logout           - Logout

🛡️ Security Features

• ✅ JWT authentication (24-hour tokens)
• ✅ Password hashing (bcrypt, 12 rounds)
• ✅ Multi-factor authentication (TOTP + SMS)
• ✅ Cache prevention (no offline possible)
• ✅ Forensic logging (complete audit trail)
• ✅ IP tracking (all actions logged)
• ✅ CORS protection
• ✅ Rate limiting
• ✅ CSRF protection
• ✅ Input validation

📈 Performance & Reliability

• ✨ Multi-API-key rotation (3 keys supported)
• ✨ Automatic fallback responses
• ✨ Error recovery & retry logic
• ✨ Session persistence
• ✨ Database auto-backup

🎯 Use Cases

Developers

• Write code in IDE → Get instant analysis
• Ask about any programming topic
• Get architecture recommendations
• Debug issues with AI assistance

Administrators

• Monitor all user activity
• Review forensic logs
• Track system usage
• Analyze patterns & trends

Enterprise Teams

• Universal AI for any question
• Code review automation
• Knowledge base assistant
• Internal support automation

Security Teams

• Network architecture design
• Cloud security setup
• Compliance documentation
• Penetration testing planning

🚀 Deployment

Local Development

python app.py
# Runs on http://localhost:5000

Production (Coming Soon)

• Docker containerization
• AWS/Azure/GCP deployment guides
• Load balancing setup
• SSL/TLS configuration

📞 Support

Documentation:

• 📖 Quick Start: QUICK_START_v2.0.md
• 📖 Features: UPGRADE_v2.0.md
• 📖 Deployment: DEPLOYMENT_COMPLETE.md

Troubleshooting:

• Server offline? python app.py to restart
• Chat not loading? Check /api/health endpoint
• Logs missing? Check Admin Dashboard
• Database issue? Delete sentinel_ai.db (auto-recreates)

🎓 Technical Stack

• Frontend: HTML5, CSS3, JavaScript (Vanilla)
• Backend: Python, Flask, SQLAlchemy
• Database: SQLite3
• AI: Google Gemini 2.5-flash
• Auth: JWT, OAuth 2.0, TOTP, SMS OTP
• Security: bcrypt, CORS, Rate Limiting

📝 License

Private - Sentinel AI

🤝 Contributing

This is a private project. Contact maintainer for contributions.

📞 Contact

For issues, questions, or suggestions, please create an issue in the repository.

---

Status: ✅ PRODUCTION READY | Version: 2.0-Universal | Last Updated: Feb 26, 2026

🚀 Ready to Rock! ├── requirements.txt # Python dependencies ├── api.txt # Your Gemini API key ├── README.md # This file ├── uploads/ # Image/file uploads directory └── templates/ └── index.html # Complete frontend (HTML+CSS+JS)

## 🚀 Quick Start

### Prerequisites
- Python 3.8+
- Google Gemini API Key ([Get one here](https://makersuite.google.com/app/apikey))
- Modern web browser

### Installation

1. **Clone/Setup the project:**
```bash
cd gemini-ai-chat

2. Install dependencies:

pip install -r requirements.txt

3. Add your API key: Edit api.txt and paste your Gemini API key on the first line:

AIzaSy...your_key_here...

4. Run the server:

python app.py

5. Open in browser: Navigate to http://localhost:5000

🎮 Usage

Text Chat

• Type a message in the input field
• Press Ctrl+Enter or Cmd+Enter to send
• Click send button or use keyboard shortcut

Image Analysis

• Click the 📷 Image button
• Select an image from your device
• Add text question about the image
• Send to get AI analysis

Voice Input

• Click the 🎤 Microphone button to record
• Click again to stop recording
• (Transcription service coming soon)

Chat History

• View all previous messages in the left sidebar
• Click any previous conversation to reload
• Create new chats with the "New Chat" button

🔧 API Endpoints

Example Request

curl -X POST http://localhost:5000/api/chat \
  -H "Content-Type: application/json" \
  -d '{
    "message": "Explain this image",
    "session_id": "session_123",
    "image": "data:image/png;base64,..."
  }'

🎨 Customization

Change UI Theme

Edit CSS variables in templates/index.html (lines 18-27):

:root {
    --primary-bg: #ffffff;
    --accent-color: #10a37f;
    /* ... more colors ... */
}

Modify System Instructions

Edit app.py line 64-71 to change AI behavior

Change AI Model

Edit app.py line 32:

MODEL_NAME = 'gemini-2.5-flash'  # or any other available model

🔐 Security

• API key stored locally in api.txt (not in code)
• CORS enabled for local development
• Conversation history stored in memory (not persisted)
• No data sent to external services except Gemini API

📋 Current Limitations & Coming Soon

• Voice transcription (ready, needs speech-to-text service)
• Persistent chat history (database needed)
• File upload support (docs, PDFs, etc.)
• Authentication & user accounts
• Firewall policy builder
• DLP rule generator
• IAM assistant
• MFA setup guide

🛠️ Troubleshooting

Port Already in Use

python app.py  # Will try port 5000
# Change in app.py line 171: app.run(debug=True, port=5001)

API Key Errors

• Verify key is on first line of api.txt
• Check key is valid at Google AI Studio
• Ensure no extra spaces or newlines

Images Not Processing

• Check file is valid PNG/JPG/GIF/WEBP
• Image size should be < 20MB
• Verify Gemini API supports image analysis

Microphone Not Working

• Grant browser microphone permissions
• Check browser security settings
• Try HTTPS or localhost

📊 Model Information

Current Model: gemini-2.5-flash

• Fast and efficient
• Supports text, images, and code
• ~0.075 USD per 1M input tokens
• ~0.3 USD per 1M output tokens

🚀 Deployment

Local Development

python app.py

Production (Gunicorn)

pip install gunicorn
gunicorn -w 4 -b 0.0.0.0:5000 app:app

Docker (Optional)

Create Dockerfile:

FROM python:3.10
WORKDIR /app
COPY . .
RUN pip install -r requirements.txt
EXPOSE 5000
CMD ["python", "app.py"]

📝 License

Open source - modify and use freely for your projects

🤝 Contributing

Suggestions and improvements welcome! The roadmap includes enterprise security features.

📞 Support

• Check README troubleshooting section
• Review API documentation
• Test with sample images first

---

Made with ❤️ for Security Professionals | v1.0 | Built on Gemini API