MAF - Model Application Firewall

Contextual Security for Modern Applications

Get Started • Documentation • Ecosystem

---

🔒 Overview

MAF (Model Application Firewall) is an intelligent security middleware that brings context, continuity, and intent detection to your application requests. Unlike traditional WAFs that block based on static rules, MAF analyzes user behavior to stop attacks before they execute.

✨ Ecosystem

The MAF project consists of three core components working together:

Component	Description	Repository
MAF Package	The core NPM middleware for Node.js/Express.
MAF App	Full-stack service for managing the middleware.
Self-Hosted	A self-hostable version of the MAF system.

🚀 Quick Start

Install the middleware package:

npm install mafai

Integrate with your Express application:

const express = require('express');
const { mafaiExpress } = require('mafai');

const app = express();
app.use(express.json());

// Initialize MAF Middleware
app.use(mafaiExpress({
    apiKey: "YOUR_MAF_APP_KEY",
    engineUrl: "http://your-engine-ip:3001/evaluate"
}));

app.listen(3000, () => console.log('Server protected by MAF'));

🛠️ Features

• Context-Aware Analysis: Understands the intent behind requests, not just the payload.
• Session Continuity: Tracks user behavior across sessions to detect anomalies.
• Zero Latency: Designed for minimal impact on request processing time.
• Plug-and-Play: Drop-in middleware for existing Express applications.

📚 Research & References

MAF's architecture is built upon cutting-edge research in web security and behavioral analysis.

• WAFFLED: Exploiting Parsing Discrepancies to Bypass WAFs

  Reveals over 1200 unique evasion techniques that bypass top WAFs by exploiting parsing mismatches. MAF's application-layer design eliminates this entire class of vulnerability.

• AI-Powered Behavioral Biometrics for Continuous Authentication

  Foundational research for our Session Continuity engine, using deep learning to detect session hijacking with high accuracy.

---

_{Built with ❤️ by TEAM Deniers.}