QLPro: Automated Code Vulnerability Discovery via LLM and Static Code Analysis Integration

TL;DR

QLPro is a novel framework that combines large language models and static analysis to improve vulnerability detection in open-source code, identifying more vulnerabilities including new zero-day exploits.

Contribution

It introduces QLPro, integrating LLMs with static analysis, and provides a new dataset, JavaTest, to evaluate vulnerability detection capabilities.

Findings

QLPro detects more vulnerabilities than CodeQL.

QLPro identified 6 previously unknown vulnerabilities.

2 of the discovered vulnerabilities are confirmed zero-days.

Abstract

We introduce QLPro, a vulnerability detection framework that systematically integrates LLMs and static analysis tools to enable comprehensive vulnerability detection across entire open-source projects.We constructed a new dataset, JavaTest, comprising 10 open-source projects from GitHub with 62 confirmed vulnerabilities. CodeQL, a state-of-the-art static analysis tool, detected only 24 of these vulnerabilities while QLPro detected 41. Furthermore, QLPro discovered 6 previously unknown vulnerabilities, 2 of which have been confirmed as 0-days.