SEC-bench Pro: Can Language Models Solve Long-Horizon Software Security Tasks?

Fuente: arXiv
Gespeichert in:
Bibliographische Detailangaben
Hauptverfasser: Lee, Hwiwon, Liu, Jiawei, Kim, Dongjun, Zhang, Ziqi, Xia, Chunqiu Steven, Zhang, Lingming
Format: Preprint
Veröffentlicht: 2026
Schlagworte:
Online-Zugang:
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
_version_ 1866918523546107904
author Lee, Hwiwon
Liu, Jiawei
Kim, Dongjun
Zhang, Ziqi
Xia, Chunqiu Steven
Zhang, Lingming
author_facet Lee, Hwiwon
Liu, Jiawei
Kim, Dongjun
Zhang, Ziqi
Xia, Chunqiu Steven
Zhang, Lingming
contents Large language models (LLMs) now support automated software security tasks, including vulnerability discovery and proof-of-concept (PoC) generation. Existing benchmarks do not faithfully evaluate LLMs in real-world bug hunting scenarios because they rely on fuzzing harnesses, target-specific descriptions, or vulnerability-reproduction tasks. We present SEC-bench Pro, a benchmark for measuring agent bug hunting on critical, high-complexity software systems. This work discloses reports with concrete PoC inputs and links fixes into reproducible tasks through a three-phase pipeline for vulnerability collection, environment reconstruction, and oracle-based validation. We instantiate SEC-bench Pro with 183 validated vulnerabilities across V8 and SpiderMonkey, including a V8 subset with more than $1.5 million in cumulative Google Vulnerability Reward Program awards. These instances span memory-safety, sandbox, JIT, and race-condition bugs under browser-grade and runtime-grade execution conditions. Our evaluation shows that coding agents with frontier models remain below 40% success on both evaluated engines. The open-weight Kimi-K2.6 baseline reaches 11.7% on V8, while the strongest frontier configuration reaches 32.0% on V8 and 38.8% on SpiderMonkey. ClaudeCode and Codex solve complementary instance sets, and their two-agent union reaches 37.9% on V8 and 48.8% on SpiderMonkey. SEC-bench Pro provides robust environments for assessing LLM-based security agents and exposes limitations in long-horizon bug hunting tasks.
format Preprint
id arxiv_https___arxiv_org_abs_2605_26548
institution arXiv
publishDate 2026
record_format arxiv
spellingShingle SEC-bench Pro: Can Language Models Solve Long-Horizon Software Security Tasks?
Lee, Hwiwon
Liu, Jiawei
Kim, Dongjun
Zhang, Ziqi
Xia, Chunqiu Steven
Zhang, Lingming
Cryptography and Security
Machine Learning
Large language models (LLMs) now support automated software security tasks, including vulnerability discovery and proof-of-concept (PoC) generation. Existing benchmarks do not faithfully evaluate LLMs in real-world bug hunting scenarios because they rely on fuzzing harnesses, target-specific descriptions, or vulnerability-reproduction tasks. We present SEC-bench Pro, a benchmark for measuring agent bug hunting on critical, high-complexity software systems. This work discloses reports with concrete PoC inputs and links fixes into reproducible tasks through a three-phase pipeline for vulnerability collection, environment reconstruction, and oracle-based validation. We instantiate SEC-bench Pro with 183 validated vulnerabilities across V8 and SpiderMonkey, including a V8 subset with more than $1.5 million in cumulative Google Vulnerability Reward Program awards. These instances span memory-safety, sandbox, JIT, and race-condition bugs under browser-grade and runtime-grade execution conditions. Our evaluation shows that coding agents with frontier models remain below 40% success on both evaluated engines. The open-weight Kimi-K2.6 baseline reaches 11.7% on V8, while the strongest frontier configuration reaches 32.0% on V8 and 38.8% on SpiderMonkey. ClaudeCode and Codex solve complementary instance sets, and their two-agent union reaches 37.9% on V8 and 48.8% on SpiderMonkey. SEC-bench Pro provides robust environments for assessing LLM-based security agents and exposes limitations in long-horizon bug hunting tasks.
title SEC-bench Pro: Can Language Models Solve Long-Horizon Software Security Tasks?
topic Cryptography and Security
Machine Learning
url https://arxiv.org/abs/2605.26548