CodeShield AI

Problem & Solution Statement Problem Statement Modern software applications can contain security vulnerabilities such as SQL Injection, Hardcoded Secrets, Insecure Input Validation, Cross-Site Scripting (XSS), Path Traversal, Weak Hashing, Eval Injection, and Debug Mode Exposure. Beginners and early-stage developers often find it difficult to identify these vulnerabilities, understand their impact, and apply secure fixes. Existing security tools may report vulnerabilities but can be difficult for learners to understand and may not provide a complete learning and verification experience. Solution CodeShield AI is a security-focused code analysis application that helps users Detect → Explain → Fix → Verify → Remember → Learn from security vulnerabilities. It analyzes submitted code, identifies common vulnerabilities, provides severity levels, explains why the code is insecure, and suggests secure fixes. Users can verify their updated code after applying the fix. The application combines rule-based security detection with IBM Granite AI. Rule-based detectors provide fast and consistent identification of known vulnerability patterns, while IBM Granite acts as an AI Security Coach, providing contextual explanations. Security Memory records detected vulnerability patterns, while Prevention Alerts and Learning Cards help users understand how to prevent similar issues in the future. Unique Innovation Instead of functioning only as a vulnerability scanner, CodeShield AI connects detection with education and verification in one workflow. This makes security feedback more actionable for students and beginner developers. Application and Presentation of Technology The application is built using Python, Streamlit, IBM watsonx.ai/Granite AI, and rule-based security analysis. A structured and interactive interface presents vulnerabilities through clear severity indicators, explanations, secure fixes, verification, prevention guidance, and learning content. Business Potential CodeShield AI can be extended for students, educational institutions, coding platforms, and development teams. Future versions could support additional programming languages, vulnerability patterns, security reports, team dashboards, and development-workflow integrations, creating possibilities for educational subscriptions and institutional licensing. Impact CodeShield AI aims to make secure coding easier to understand and practice by turning security detection into a continuous learning experience.