Network security testing is a crucial component of modern cybersecurity practices, enabling organizations to identify and address vulnerabilities before they are exploited. Existing penetration testing frameworks, such as Kali Linux, Metasploit, and Burp Suite, provide powerful tools but typically overwhelm beginners, lack contextual prioritization of vulnerabilities, and fail to adapt to localized environments. To address these challenges, a customizable decision-tree- driven toolkit for ethical network security testing is presented. This toolkit integrates reconnaissance, vulnerability scanning, and controlled exploitation with a decision-tree–based prioritization model. The decision tree evaluates vulnerabilities based on factors such as severity scores, exploit availability, and service criticality, thereby guiding ethical hackers to focus on the most critical risks first. The system also includes ethical safeguards such as activity logging, user confirmation before sensitive tests, and restrictions to controlled virtual environments, ensuring responsible use. Unlike existing frameworks, this approach offers both usability and intelligence, making it particularly valuable in developing contexts where cybersecurity expertise and resources are limited. Results from controlled testing environments demonstrate improvements in accuracy, prioritization efficiency, and reduced false positives. This work contributes a practical, ethical, and intelligent toolkit that advances the practice of network security testing and provides a foundation for further research in AI-assisted ethical hacking.