Skip to content

rishika105/CodeShield-AI

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

33 Commits
 
 
 
 
 
 
 
 
 
 

Repository files navigation

🛡️ CodeShield AI - GenAI Security Platform

CodeShield AI is a GenAI-powered cybersecurity tool built to help developers detect and fix code vulnerabilities in real-time. It uses advanced models like VulBERTa, VulDeePecker, and OpenAI (GPT) to analyze source code, explain risks, and generate secure alternatives — all wrapped in an interactive and gamified experience.

🎯 Built as a submission for CodHer Hackathon 2025 to reimagine how developers learn and apply secure coding practices using AI.


🚀 Features

  • 🔍 Multi-Model Vulnerability Detection
    Integrated with models like VulBERTa, VulDeePecker, and GPT via Hugging Face to identify security flaws with high precision.

  • 🛠️ Fix Suggestions & Secure Code Generation
    Provides intelligent explanations and auto-generated secure versions of vulnerable code.

  • 🎮 Gamification
    Includes XP, quests, badges, and leaderboards to make learning and fixing vulnerabilities fun and engaging.

  • 💻 Interactive Playground
    A live environment to test, visualize, and understand code vulnerabilities and fixes.

  • 🔐 Authentication System
    Role-based login/signup system built with Node.js and Express, ensuring secure and scalable user management.


Video Demo

1744973783887.mp4

🧱 Tech Stack

Component Tech Used
Frontend React.js + Tailwind CSS
Backend Node.js + Express.js (Auth)
AI/ML Models VulBERTa, VulDeePecker, GPT (Hugging Face, Open AI)
Model Backend Python + Flask
Database MongoDB

👩‍💻 Team HackHerWay – CodHer 2025

Made with passion and purpose to make code security interactive, intelligent, and developer-friendly.

About

A GenAI-powered cybersecurity platform for developers that detects code vulnerabilities using models like VulBERTa, VulDeePecker, and GPT, and provides fixes with gamified learning features.

Topics

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

 
 
 

Contributors