With the rapid growth of software development, platforms like GitHub have become essential for code sharing and collaboration. However, many developers, especially students and beginners, often upload code without proper security checks, leading to vulnerabilities such as hardcoded credentials, exposed API keys, and insecure coding practices. This project presents an AI-Based GitHub Security Scanner designed to automatically analyze repositories and identify potential security risks. The system integrates with GitHub to scan source code using a combination of static code analysis and AI-driven techniques. It detects common vulnerabilities, misconfigurations, and sensitive data exposure in real time. The AI component enhances detection accuracy by learning patterns from known security issues and suggesting improvements to developers. Additionally, the tool provides detailed reports and recommendations, helping users understand and fix vulnerabilities effectively. By automating security analysis, this project aims to improve coding practices, reduce risks, and promote secure software development. Overall, the proposed system offers a scalable and intelligent solution for early detection of security flaws in GitHub repositories, making it especially useful for students, developers, and organizations.