ESCROW is a post‑deployment maintenance framework for large‑language‑model agents that operate under policy‑bound enterprise workflows. It ensures that any skill revisions proposed by the LLM are only deployed after empirical evaluation, using distributed diagnosis, consensus, per‑category non‑regression guards, cross‑cycle anti‑regression checks, and an accuracy–cost Pareto search to produce a versioned, auditable diff for each change. In a real production financial document‑auditing system, ESCROW achieved the best accuracy–cost trade‑off compared with baseline methods and demonstrated transferability on a public $ au$‑bench.
By Ruoqi Shu, Chen Dan, Xuhui Wang, Tianhua Xu, Mengxi Luo, Yanming Mai, Bo Wan
arXiv:2608.24569v2 Announce Type: replace
Abstract: Large language model (LLM) agents build on one another's work through summaries, plans, and other intermediate artifacts. These records let a downs...
By Yiheng Sun, Huifei Wang, Yancheng Zhu, Zhenyu Li, Zebin Zhao, Yifan Yuan
arXiv:2610.14966v1 Announce Type: new
Abstract: Long-horizon agents must complete workflows in which later activities depend on earlier outputs and changes to shared state. A key challenge for evalua...
By Yuxuan Liu, Zhaochen Su, Hangyu Guo, Shijue Huang, Zhixiong Zhang, Kawai Chung, Zhongwei Xie, Tianshi Zheng, Wei Fan, Qing Zong, Haochen Shi, Baixuan Xu, Haoran Li, Yangqiu Song
arXiv:2606. 10241v1 Announce Type: new Abstract: Autonomous improvement loops are hard to trust because the improvement process is usually external scaffolding bolted onto the agent: failures go unlogged, diagnoses cannot be replayed, and promote-or-discard decisions land in a side database rather than the agent's own history.
By Yohei Nakajima
The paper investigates how AI agents that run long workflows using external tools can experience inconsistencies when retries, speculative execution, concurrency, or partial failures occur. It introduces an effect‑history model that distinguishes between actual external events and the agent’s observations, and catalogs eight common external‑effect anomalies. The authors analyze the standard Model Context Protocol tool interface, finding that its annotations are too coarse to fully express the necessary capabilities to prevent these anomalies, thereby motivating the need for reusable transactional contracts at the agent‑tool boundary.
By Artem Trofimov, Boris Novikov
EffectMatch is a runtime system that monitors and validates persistent changes made by large language model agents during software interactions. It collects all side‑effects within a controlled execution boundary and compares them against the application’s approved state, deciding whether to commit the changes and allow subsequent steps. In tests on 206 public business tasks, EffectMatch preserved correct executions and prevented all incorrect commits, with ablation studies showing the importance of each component.
By Haoran Zhang, Hengtong Zhang, Zhiyu Liang, Yu Yan, Decheng Zuo, Hongzhi Wang
arXiv:2608. 19741v1 Announce Type: new Abstract: Recent agent benchmarks increasingly ground evaluation in executable environments, from code repair to web navigation, app APIs, and function calling.
By Zhuochun Li, Youngmin Ko, Ali Keramati, Nicola Ferri, Susana Palmaz Lopez Pelaez, Liang-Chun Tsai, Calvin Wang, Mirco Milletari, Tuhin Kundu, Vadim Smolyakov, Kjartan Olafsson, Tommy Guy
The paper introduces a benchmark of 400 synthetic healthcare workflow episodes to evaluate large language model (LLM) agents’ reliability in escalating care when evidence is missing or conflicting. It finds that while clean decision accuracy is high (0.97), models miss a large proportion of necessary escalations (32–97%) under uncertainty, and even the largest 9.7B model’s performance drops to 0.78 in conflict scenarios. A workflow-aware controller improves utility and reduces missed escalations, but its effectiveness is limited by lexical generalization failures.
By Zihao Zheng (Washington University in St. Louis), Junyi Yao (Washington University in St. Louis), Baichuan Li (Southern Methodist University), Jiayu Long (Washington University in St. Louis)
The paper introduces E‑Ledger, a multi‑agent harness that ensures safe and persistent execution of large language model agents in partially observable environments. It adds a code approval layer to enforce policy compliance and maintains a world ledger of verified hidden rules and dynamic state. To discover unknown rules, the authors propose WorldAbduct, an abductive framework that diagnoses execution trajectories, hypothesizes latent rules, and validates them before integrating them into the ledger, improving safe task completion on enterprise and scientific benchmarks.
By Yisen Gao, Yue Guo, Qing Zong, Yiwen Guo, Yangqiu Song
arXiv:2607. 18316v1 Announce Type: cross Abstract: Tool-augmented language-model agents execute multi-step workflows over external systems, resolving an entity once and then acting on it across subsequent steps.
By Rahul Suresh Babu, Shashank Indukuri
arXiv:2610.34167v1 Announce Type: new
Abstract: Tool-using language model agents are usually evaluated by whether they finish the user's task. That metric misses a common failure: an agent can comple...
By Jiaju Huang, Shaobin Chen, Xinglong Liang, Xinyu Ma, Hao Yang, Wei Ke, Tao Tan, Yue Sun
arXiv:2610.27672v1 Announce Type: new
Abstract: Rerunning every governance control after every change to an artificial intelligence (AI) pipeline is expensive, so teams reuse earlier results. Reuse i...
By Atanu Mukhopadhyay, Shravya Reddy Nalla