Abstract
Large language model agents can invoke tools fluently, but enterprise workflows demand more than selecting the right tools: actions must strictly comply with organizational policies, tool feedback often conceals hidden side effects under partial observability, and long-horizon tasks require persistent state tracking across multiple records. To address these challenges, we introduce E-Ledger, a multi-agent harness for safe and persistent execution. E-Ledger employs a code approval layer that checks every proposed action against policy before execution, and maintains a world ledger of verified hidden rules alongside evidence-backed dynamic state. Because hidden rules are typically unknown a priori, we further propose WorldAbduct, an abductive, world-model-driven harness evolution framework. WorldAbduct diagnoses execution trajectories across four complementary views (state consistency, world-observation gap, policy-gate correctness, and goal judgment) to hypothesize latent rules, and verifies them through targeted abductive interactions before integrating them into the ledger. On the enterprise benchmark World of Workflows, E-Ledger with WorldAbduct improves safe task completion across four LLM backbones, outperforming the strongest evolution baseline by 5--15 percentage points. Experiments in ScienceWorld and DiscoveryWorld further show that abductive harness evolution carries over to scientific environments. Our code is available at https://github.com/HKUST-KnowComp/E-LEDGER-WorldAbduct.
Keywords
Subject
Publication details
- Journal
- Not available
- Open access
- Green open access
Cite this article
APA 7
Gao, Y., Guo, Y., Zong, Q., Guo, Y., & Song, Y. (2026). Safe, Persistent, and Evolving Agent Harness for Understanding Partially Observable Worlds. https://omanscience.com/en/articles/safe-persistent-and-evolving-agent-harness-for-understanding-partially-observable-worlds
MLA 9
Gao, Yisen, et al. "Safe, Persistent, and Evolving Agent Harness for Understanding Partially Observable Worlds." https://omanscience.com/en/articles/safe-persistent-and-evolving-agent-harness-for-understanding-partially-observable-worlds.
Chicago (author–date)
Gao, Yisen, Yue Guo, Qing Zong, Yiwen Guo, and Yangqiu Song. 2026. "Safe, Persistent, and Evolving Agent Harness for Understanding Partially Observable Worlds." https://omanscience.com/en/articles/safe-persistent-and-evolving-agent-harness-for-understanding-partially-observable-worlds.
Harvard
Gao, Y., Guo, Y., Zong, Q., Guo, Y. and Song, Y. (2026) 'Safe, Persistent, and Evolving Agent Harness for Understanding Partially Observable Worlds', Available at: https://omanscience.com/en/articles/safe-persistent-and-evolving-agent-harness-for-understanding-partially-observable-worlds.
Vancouver
Gao Y, Guo Y, Zong Q, Guo Y, Song Y. Safe, Persistent, and Evolving Agent Harness for Understanding Partially Observable Worlds. https://omanscience.com/en/articles/safe-persistent-and-evolving-agent-harness-for-understanding-partially-observable-worlds
IEEE
Y. Gao, Y. Guo, Q. Zong, Y. Guo, and Y. Song, "Safe, Persistent, and Evolving Agent Harness for Understanding Partially Observable Worlds," https://omanscience.com/en/articles/safe-persistent-and-evolving-agent-harness-for-understanding-partially-observable-worlds.