VulnAgent-X: A Layered Agentic Framework for Repository-Level Vulnerability Detection

Fuente: arXiv
Gespeichert in:
Bibliographische Detailangaben
Hauptverfasser: Meng, Renwei, Wu, Haoyi, Wang, Jingming, Bai, Haoyan
Format: Preprint
Veröffentlicht: 2026
Schlagworte:
Online-Zugang:
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
_version_ 1866910052573511680
author Meng, Renwei
Wu, Haoyi
Wang, Jingming
Bai, Haoyan
author_facet Meng, Renwei
Wu, Haoyi
Wang, Jingming
Bai, Haoyan
contents Software vulnerability detection is critical in software en- gineering as security flaws arise from complex interactions across code structure, repository context, and runtime conditions. Existing meth- ods are limited by local code views, one-shot prediction, and insuffi- cient validation, reducing reliability in realistic repository-level settings. This study proposes VulnAgentX, a layered agentic framework integrat- ing lightweight risk screening, bounded context expansion, specialised analysis agents, selective dynamic verification, and evidence fusion into a unified pipeline. Experiments on function-level and just-in-time vul- nerability benchmarks show VulnAgent-X outperforms static baselines, encoder-based models, and simpler agentic variants, with better local- isation and balanced performance-cost trade-offs. Treating vulnerabil- ity detection as a staged, evidence-driven auditing process improves de- tection quality, reduces false positives, and produces interpretable re- sults for repository-level software security analysis. Code is available at https://github.com/xiaolu-666113/Vlun-Agent-X.
format Preprint
id arxiv_https___arxiv_org_abs_2603_13384
institution arXiv
publishDate 2026
record_format arxiv
spellingShingle VulnAgent-X: A Layered Agentic Framework for Repository-Level Vulnerability Detection
Meng, Renwei
Wu, Haoyi
Wang, Jingming
Bai, Haoyan
Software Engineering
Artificial Intelligence
Software vulnerability detection is critical in software en- gineering as security flaws arise from complex interactions across code structure, repository context, and runtime conditions. Existing meth- ods are limited by local code views, one-shot prediction, and insuffi- cient validation, reducing reliability in realistic repository-level settings. This study proposes VulnAgentX, a layered agentic framework integrat- ing lightweight risk screening, bounded context expansion, specialised analysis agents, selective dynamic verification, and evidence fusion into a unified pipeline. Experiments on function-level and just-in-time vul- nerability benchmarks show VulnAgent-X outperforms static baselines, encoder-based models, and simpler agentic variants, with better local- isation and balanced performance-cost trade-offs. Treating vulnerabil- ity detection as a staged, evidence-driven auditing process improves de- tection quality, reduces false positives, and produces interpretable re- sults for repository-level software security analysis. Code is available at https://github.com/xiaolu-666113/Vlun-Agent-X.
title VulnAgent-X: A Layered Agentic Framework for Repository-Level Vulnerability Detection
topic Software Engineering
Artificial Intelligence
url https://arxiv.org/abs/2603.13384