Fares Trad

I am Fares Trad, a MASc student at York University, supervised by Dr. Gias Uddin. My research interests include software security, secure code analysis, automated program repair, and large language models for software engineering. I am particularly interested in building and evaluating AI-based approaches for detecting, explaining, and mitigating security vulnerabilities in code.

My recent work includes SafeScript, a VS Code extension for detecting security vulnerabilities in generated and handwritten code using static-analysis-inspired checks, vulnerability mapping, and AI-assisted feedback to help produce safer code. I have also worked on adversarial benchmarking for automated program repair systems and on comparing static analysis tools with AI-based methods for software vulnerability detection.

Before and during my graduate studies, I have been involved in teaching assistantship duties, software engineering research, and applied work related to security, static analysis, software testing, and developer tools.

Homepage: HOMEPAGE

Research Interests

• Software security
• Secure code analysis
• Software testing
• Large language models for software engineering
• Automated program repair
• AI-assisted developer tools

Publications & Research Contributions

SafeScript-X: Structured Vulnerability-Focused Explanations for Secure Code Analysis
Ongoing research on structured prompting techniques for vulnerability-focused explanation and secure code analysis using large language models.

SafeScript
Ongoing work on a VS Code extension for detecting and explaining security vulnerabilities in generated and handwritten code through static-analysis-inspired checks and AI-assisted feedback.

Adversarial Benchmarking for Automated Program Repair
Ongoing research on adversarial issue construction and robustness evaluation for APR systems, including attack categories designed to expose weaknesses in software engineering agents.