A multi-modal AI security system and browser extension that detects phishing, spam, and social engineering attacks in real time. Combining NLP, heuristic intelligence, adaptive learning, and threat visualization to protect users across the web.
SentinelAI is an AI-powered cybersecurity platform and browser extension built to combat modern phishing, spam, and social engineering attacks.
As attackers increasingly leverage AI-generated scams, traditional spam filters struggle to keep up. SentinelAI addresses this challenge by combining natural language analysis, URL heuristics, behavioral signals, and adaptive learning to identify malicious activity before it harms the user.
When a user visits a webpage or receives suspicious content:
- The browser extension analyzes page content, links, and forms.
- The AI threat engine evaluates phishing patterns, spoofed domains, and suspicious language.
- A risk score with confidence metrics is generated.
- The system provides real-time warnings and educational feedback.
- Insights can be visualized in the SentinelAI dashboard website.
This hybrid system combines browser-level protection with centralized analytics dashboards, making it suitable for both individual users and enterprise security teams.
| Feature | Description |
|---|---|
| ๐ง AI-Driven Threat Detection | NLP analysis of phishing phrases, urgency cues, and authority patterns |
| ๐ URL & Domain Spoof Detection | Detects suspicious TLDs, IP-based URLs, and spoofed domains |
| ๐ผ Image Spoof Detection (Planned) | Detect fake logos and deceptive images |
| ๐ Risk Scoring Engine | Calculates threat scores with confidence metrics |
| ๐งฌ Federated Learning Simulation | Adaptive threat detection using anonymized insights |
| ๐ฏ Predictive GAN Simulations | Generates synthetic phishing attacks for defense testing |
| ๐ Security Dashboards | Visualize threats and detection insights |
| ๐ User Awareness Training | Gamified modules to educate users about phishing |
| ๐ข Enterprise Integrations | APIs, compliance reporting, and policy enforcement |
graph TD
A[๐ User Browsing Website] --> B[Browser Extension Content Script]
B --> C[Data Extraction]
C --> D[AI Threat Detection Engine]
D --> E[Text Analysis - NLP]
D --> F[URL & Domain Heuristics]
D --> G[Form & Metadata Inspection]
D --> H[Image Analysis]
E --> I[Threat Scoring Engine]
F --> I
G --> I
H --> I
I --> J[Risk Score + Confidence]
J --> K[User Warning / Notification]
I --> L[SentinelAI Dashboard]
SentinelAI consists of three major components:
- Content scripts monitor web page content.
- Background service worker processes detection requests.
- Local AI engine evaluates phishing threats.
- Chrome APIs handle notifications and browser interaction.
Provides:
- Security dashboards
- Threat analytics
- Documentation
- Enterprise integration modules
Responsible for:
- NLP phishing detection
- Domain spoofing detection
- Suspicious form analysis
- Risk scoring and confidence calculation
| Component | Technology |
|---|---|
| Framework | React |
| Language | TypeScript |
| Build Tool | Vite |
| UI Components | shadcn-ui |
| Styling | Tailwind CSS |
| Component | Technology |
|---|---|
| Extension Framework | Chrome Manifest V3 |
| Language | JavaScript (ES6+) |
| Browser APIs | Chrome Tabs, Runtime, Storage, Notifications |
| Threat Detection Engine | Custom heuristic NLP logic |
| Data Storage | JSON threat patterns |
| Future Integration | TensorFlow.js for image analysis |
SentinelAI/
โ
โโโ extension/ # Chrome Extension (Manifest V3)
โ
โ โโโ background/
โ โ โโโ service-worker.js # Background script for detection logic
โ
โ โโโ content/
โ โ โโโ content.js # Injected scripts analyzing page content
โ
โ โโโ popup/
โ โ โโโ popup.html
โ โ โโโ popup.css
โ โ โโโ popup.js # Popup interface logic
โ
โ โโโ lib/
โ โ โโโ ai-engine.js # AI threat detection engine
โ โ โโโ utils.js # Utility helpers
โ
โ โโโ assets/
โ โ โโโ icons/
โ โ โโโ models/
โ โ โโโ threat-patterns.json
โ โ โโโ logo-database.json
โ โ โโโ blacklist.json
โ
โ โโโ styles/
โ โ โโโ content.css
โ
โ
โโโ website/ # React + Vite dashboard
โ โโโ src/
โ โโโ components/
โ โโโ pages/
โ โโโ public/
โ โโโ tailwind.config.js
โ
โโโ README.md
- Node.js 18+
- Chrome Browser
- npm or yarn
git clone https://github.com/kishorekrrish3/SentinelAI-Phishing-and-Social-Engineering-Detection-Engine-and-Extension.git
cd SentinelAI-Phishing-and-Social-Engineering-Detection-Engine-and-Extensionnpm installnpm run devThe dashboard will be available at:
http://localhost:5173
- Open Chrome and navigate to:
chrome://extensions/
-
Enable Developer Mode.
-
Click Load unpacked.
-
Select the
extension/folder.
The SentinelAI extension will now be active.
Once installed:
1๏ธโฃ Browse any website normally.
2๏ธโฃ SentinelAI analyzes:
- page content
- links
- forms
- domain metadata
3๏ธโฃ If a threat is detected, the extension provides:
โ ๏ธ Warning alerts- ๐ Risk score
- ๐ Threat breakdown
4๏ธโฃ View analytics in the SentinelAI Dashboard.
| Setting | File | Description |
|---|---|---|
| Threat Patterns | threat-patterns.json |
NLP phishing indicators |
| Domain Blacklist | blacklist.json |
Known malicious domains |
| Logo Detection DB | logo-database.json |
Brand logo spoof detection |
| AI Engine Logic | ai-engine.js |
Risk scoring and detection |
Ensure:
- The extension is enabled in chrome://extensions
- The page allows content scripts to run
Verify the Vite server started successfully:
http://localhost:5173
Restart the development server if needed.
- ๐ง Deep learning phishing detection models
- ๐ผ TensorFlow.js image spoof detection
- ๐ Real-time threat intelligence feeds
- ๐ Enterprise SOC dashboards
- ๐ก Cross-browser extension support
- ๐ Large-scale phishing dataset training
We welcome contributions from the community.
- Fork the repository
- Create a feature branch
- Commit your changes
- Open a pull request
Please follow project coding guidelines and check existing issues before submitting.
This project is licensed under the MIT License.
Defending the internet from AI-powered scams with AI itself.
SentinelAI โ intelligent security for the modern web.