AI-Powered Code Review Assistant

1)     CodeGuard AI combines static code analysis and Artificial Intelligence to identify potential security issues in source code. It provides a structured approach to reviewing code security. 2)     The system uses predefined rules and pattern matching to detect potentially insecure coding patterns. This helps developers recognize possible vulnerabilities during software development.  3)     Gemini AI explains detected issues in simple and understandable language. This helps users understand the causes and possible security impacts of reported findings. 4)     The platform provides remediation suggestions to help developers improve their source code. These recommendations encourage the use of safer coding practices.  5)     CodeGuard AI supports source-code submission for security analysis. This allows users to examine individual code snippets and review potential security weaknesses.  6)     ZIP-based project scanning can help analyze multiple source files within a software project. This provides a convenient approach to reviewing project code, depending on the supported implementation.  7)     The post-fix verification process allows users to rescan modified source code. The new results help users check whether previously detected patterns remain after corrections.  8)     Scan history and reporting help users organize and revisit previous analysis results. This supports the review of findings across multiple scanning sessions.  9)     CodeGuard AI can assist students and beginner developers in learning about software vulnerabilities and secure coding practices. It connects practical code analysis with AI-generated explanations.  10)  The system can be improved by expanding detection rules, supporting more programming languages, and strengthening verification methods. Further testing is required to evaluate its detection capabilities and the reliability of its recommendations.

[1]     H. Pearce, B. Tan, B. Ahmad, R. Karri, and B. Dolan-Gavitt,    “Examining  Zero-Shot Vulnerability Repair with Large Language Models,” Proceedings of the IEEE Symposium on Security and Privacy (SP), 2023. DOI: 10.1109/SP46215.2023.10179324  [2]     X. Zhou, S. Cao, X. Sun, and D. Lo, “Large Language Model for Vulnerability Detection and Repair: Literature Review and the Road Ahead,” arXiv preprint arXiv:2404.02525, 2024. Available:  DOI  : https://arxiv.org/abs/2404.02525.   [3]     D. de Fitero-Dominguez, E. Garcia-Lopez, A. Garcia-Cabot, and J.-J. Martinez-Herraiz, “Enhanced Automated Code Vulnerability Repair Using Large Language Models,” Engineering Applications of Artificial Intelligence, vol. 138, Part A, Article 109291, 2024. DOI: 10.1016/j.engappai.2024.109291.  [4]     X. Zhou, S. Cao, X. Sun, and D. Lo, “Large Language Model for Vulnerability Detection and Repair: Literature Review and the Road Ahead,” ACM Transactions on Software Engineering and Methodology, vol. 34, no. 5, Article 145, pp. 1–31, 2025. DOI: 10.1145/3708522.  [5]     D. Gnieciak and T. Szandala, “Large Language Models Versus Static Code Analysis Tools: A Systematic Benchmark for Vulnerability Detection,” arXiv preprint arXiv:2508.04448, 2025.    Available DOI: https://arxiv.org/abs/2508.04448.  [6]     V. Belozerov, P. J. Barclay, and A. Sami, “Secure Coding with AI – From Detection to Repair,” Empirical Software Engineering, vol. 31, Article 93, 2026. DOI: 10.1007/s10664-026-10812-8