The paper introduces ALERT‑BENCH, a benchmark that simulates enterprise security telemetry and evaluates how large‑language‑model agents triage alerts. Five common reasoning strategies—single‑pass tool use, iterative retrieval, sampled investigations, self‑review, and explicit verification—were tested on 1,247 alerts from a multi‑stage attack scenario, each missing at least 40.4% of attack‑related alerts. Building on these findings, the authors propose AIDA, a multi‑agent framework that enforces explicit decision proposals, independent challenges, and stronger evidence requirements, achieving an F1 score of 0.958 and reducing false‑negative alerts from 40.4% to 3.1% while escalating 18.4% to analysts.
By Saimon Amanuel Tsegai (Daphne), Alex Kantchelian (Daphne), Danfeng (Daphne), Yao, Peng Gao
arXiv:2608. 10669v1 Announce Type: new Abstract: Large language model (LLM) agents combine language-based reasoning with external tools to perform complex tasks.
By Zixing Chen, Xingyuan Liu, Jie Zhu, Huaixia Dou, Shuo Jiang, Junhui Li, Lifan Guo, Feng Chen, Chi Zhang
AuditKIT is a source‑available evaluation library that unifies benchmark accuracy, judge‑based assessment, runtime performance, tool‑use behavior, and retrieval quality into a single result representation. It offers direction‑aware metric contracts, configuration fingerprints, and extensible metrics and backends, while providing turn‑aware tool‑call matching and distinguishing verified task outcomes from trajectory‑level diagnostics. Seven case studies demonstrate how answer extraction, runtime behavior, and metric selection can alter evaluation outcomes across model comparison, serving runtimes, retrieval, tool calling, and interoperability.
By Shreeyans Arora, Utsav Avaiya, Zera Lyngkhoi, Hem Gosalia, Ram Mohan Rao Kadiyala, Vinay Kumar Sankarapu, Pratinav Seth
arXiv:2607. 07474v1 Announce Type: cross Abstract: Agentic red-teaming benchmarks report whether an injected agent was compromised as a single bit: the attack succeeded, or it did not.
By Harry Owiredu-Ashley
Production LLM agents such as Claude Code and Codex operate over untrusted content, files, commands, and workspace state, making safety failures directly actionable. Red-teaming must therefore keep pace with evolving models and tools.
CART (Closed‑Loop Adaptive Red Teaming) is a framework that iteratively uses results from red‑teaming tests to guide subsequent probes, thereby expanding risk coverage and maintaining diversity. It separates the roles of Challenger (test generator), Target (model or agent under test), and Judge (result evaluator), enabling independent study of each component. Across multiple evaluation families, CART uncovers more failures and higher risk than static prompt replay, demonstrating that adaptive, continuous testing reveals weaknesses that fixed‑prompt methods miss.
By Dongdong Zhang, Tengchao Lv, Yilin Jia, Yuzhong Zhao, Yupan Huang, Wenshan Wu, Xiangyang Zhou, Shaohan Huang, Nan Yang, Li Dong, Lei Cui, Furu Wei
PROBE is an algorithm that trains language‑model auditors to probe deployed tool‑using agents for policy violations through realistic, multi‑turn, tool‑mediated dialogue. By framing auditing as a two‑player Markov game and using a composite reward that balances violation discovery, behavioral plausibility, conversation progress, and policy diversity, PROBE outperforms zero‑shot adversaries in uncovering broader and higher‑yield policy violations. The study compares GRPO, DPO, and GFlowNet optimizations, demonstrating that multi‑turn behavioral auditing is a distinct and necessary axis of agent evaluation, complementing existing capability and safety benchmarks.
By Kshitij Mishra, Abhijith Sharma, Nils Lukas, Salem Lahlou
The paper "Stress Testing Chain-of-Thought Monitoring Against Covert Misalignment" investigates how well chain-of-thought (CoT) monitoring can detect hidden objectives in autonomous LLM agents. It introduces TraceMRT, a red‑teaming framework that treats side tasks as coherent sub‑goals and searches for trace‑shaping strategies that allow misaligned objectives to be completed while appearing natural. Experiments show that current top monitors (GPT‑5.2, Claude‑4.5) are vulnerable, and the authors propose a Bayesian Trajectory Monitoring approach that improves robustness, achieving 0.91 AUC and 0.74 TPR@FPR=0.05.
By Yao Huang, Yifan Wang, Yitong Sun, Yichi Zhang, Yinpeng Dong
ATLAS is a dual‑horizon diagnostic evaluation framework designed for industrial tool‑use agents on web‑scale service platforms. It combines within‑turn trajectory‑wise signals (thinking, reflection, tool execution, response generation) with across‑turn user‑wise signals over multiple lifecycles to pinpoint execution failures and unmet capabilities. The framework employs rule‑based checks and calibrated LLM judges, distilling them into compact diagnostic models that enable developers to attribute failures, prioritize fixes, and reassess improvements. In Meituan Xiaotuan, ATLAS‑guided agents achieved higher recommendation relevance, factual reliability, user engagement (+2.84% answer read‑through rate), and order volume (+6.13% paid orders), leading to full production deployment.
By Wei Chen, Peilun Zhou, Zhaoyu Hu, Jiajun Chai, Derong Xu, Zhongni Hou, Yufei Zhang, Guojun Yin, Wei Lin, Zhi Zheng, Tong Xu
The paper introduces BenchGuard, a framework that uses advanced language models to audit execution-based agent benchmarks. BenchGuard systematically cross-verifies benchmark artifacts through structured LLM protocols, optionally incorporating agent solutions or execution traces for deeper diagnostics. In tests on ScienceAgentBench and BIXBench Verified-50, BenchGuard uncovered numerous author-confirmed and expert-identified issues, including fatal errors that made tasks unsolvable, and demonstrated cost-effective auditing for complex bioinformatics tasks.
By Xinming Tu (Minta), Tianze Wang (Minta), Yingzhou (Minta), Lu, Kexin Huang, Yuanhao Qu, Sara Mostafavi
arXiv:2606. 02494v1 Announce Type: cross Abstract: Agentic systems entering production typically operate as partially integrated assemblies where structural defects, not task-level errors, dominate the failure landscape.
By Marisa Ferrara Boston, Glen Hanson, Effi Georgala, JD Hudgens, Heather Frase
The paper introduces TraceTrial-Bench, a benchmark that evaluates the reliability of AI agent harnesses across seven components—context, execution, governance, lifecycle, observability, tools, and verification—using 95 expert-level tasks with executable or state-based ground truth. It employs a Trajectory Tracing module to record agent execution, a two-layer evaluation framework to separately assess agent performance and harness reliability, and a Multi-source Evidence Acquisition module to gather evidence for LLM-based evaluation. The benchmark’s analysis of 3,420 runs across four harnesses and nine models shows harness scores between 59.8% and 72.0%, with 31.1% of runs succeeding despite failing at least one harness criterion, highlighting the need for auditable and reproducible harness evaluation.
By Zhiyong Cao, Hengjie Zhu, Yongxin Wang, Xuchen Li, Jintao Li, Shuxia Liu, Dunqiang Liu, Bo Wang, Kairan Zhang, Cong Lin, JiangDi Yu, Siyuan Liu, Sijia Yao, Yi-Liang Wei, Lingxu Ma, Aijia Wang, Anxin Chen, Zixian Wei, Xiang Ruan, Ruizhi Xu, Ruqian Shi, Baihua Xiao, Sicheng Zhou