🛡️ AI Security WAF

A real-time AI-powered Web Application Firewall that detects and blocks malicious payloads including SQL injection, XSS, command injection, and LLM-specific attacks like prompt injection and jailbreaks.

Features

• Multi-Layer Detection

  • Signature-based rule engine (regex pattern matching)
  • Entropy/perplexity analysis for encoded or obfuscated payloads
  • Embedding-based semantic detection for system prompt leakage

• Threat Categories Covered

  • SQL Injection (basic, UNION-based, blind/time-based)
  • NoSQL Injection
  • Cross-Site Scripting (XSS)
  • Command Injection
  • Path Traversal
  • Server-Side Template Injection (SSTI)
  • XML External Entity (XXE)
  • Insecure Deserialization
  • Prompt Injection & Jailbreak attempts (DAN, roleplay hijacking, fictional-framing exploits)
  • Delimiter/tag smuggling
  • System prompt extraction

• Payload Normalization Automatically decodes HTML entities, Unicode/hex escapes, URL encoding, and multi-layer Base64 before scanning — catching obfuscated attacks that bypass naive filters.

• Live Dashboard Real-time scan results, threat log, and an interactive attacker sandbox with pre-loaded test payloads for demonstration.

Tech Stack

• Framework: Next.js (App Router)
• Language: TypeScript
• Styling: Tailwind CSS
• Detection Engine: Custom rule engine + entropy scoring + embedding similarity

Getting Started

git clone https://github.com/iqrazulfiqar7373-coder/ai-security-waf.git
cd ai-security-waf
npm install
npm run dev