Researchers have developed a framework called ARQ that refines CodeQL queries for detecting vulnerabilities in C/C++ programs. ARQ uses synthesized programs to identify weaknesses in the queries and then runs an LLM-based refinement loop to repair them. The authors claim that ARQ-refined queries detect up to 119.8% more true positives than original CodeQL queries, with a precision of at least 98%. They also demonstrate that ARQ can fix unresolved GitHub issues and expose previously undiscovered bugs in real-world libraries. --- Why it matters: This matters because it could improve the accuracy of vulnerability detection tools used by developers and security researchers. By refining CodeQL queries, ARQ has the potential to reduce false positives and false negatives, making it easier to identify actual vulnerabilities in C/C++ code. Source: https://arxiv.org/abs/2608.20637