Project Description

Source code Compliance Evaluator and Parser for Threat Resolution and Elimination (SCEPTRE) is an AI tool used to help streamline the process of code review through the utilization of static code scanners and large language models. We researched modern AI models such as Large Language Models (LLMs), had meetings with cybersecurity engineers to revise product scope to meet their needs, and developed a pipelined program that takes in code and outputs a checklist with relevant security vulnerabilities. We worked together to research what AI models we wished to use. After landing on LLMs, we split up to develop the model for our needs, to process data coming into the model, and to output the data from our model as a checklist.