cs.CLMay 26, 2026

FinHarness: An Inline Lifecycle Safety Harness for Finance LLM Agents

Authors: Haoxuan JiaYang LiuBin ChongYingguang YangYancheng ChenJiayu LiangQian LiHanning Lu+5 more

Organizations: Nanyang Technological University · Tsinghua University · Peking University · University of Science and Technology of China · University of Chinese Academy of Sciences · Soochow University · Beijing University of Posts and Telecommunications · University of Leeds · Fullive Innovation (Beijing) AI Technology Co., Ltd. · Beihang University · University of Illinois Chicago

Abstract

Finance LLM agents must simultaneously block prompt-induced unauthorized actions and approve legitimate multi-step business workflows. However, boundary filters often miss irreversible mid-trajectory tool calls, while post-hoc LLM judges perform auditing only after termination -- too late for intervention and at a computational cost that scales linearly with trace length. We present FinHarness, an inline safety harness that wraps a finance agent end-to-end with three components: a Query Monitor that fuses single-turn intent with cross-turn drift, a Tool Monitor that evaluates each prospective tool call, and a Cascade module that integrates per-step risk and adaptively routes verification between a lightweight and an advanced-tier LLM judge. Fired risk factors are re-injected into the agent input as ex-ante evidence, enabling the agent to refuse, re-plan, or approve on its own. On FinVault, routed FinHarness cuts ASR from 38.3% to 15.0% while largely preserving benign approval (41.1%39.3%41.1\% \to 39.3\%), and uses 4.7×4.7\times fewer advanced-judge calls than an always-advanced ablation.

Explore similar work

CardsList
  1. Auditing Agent Harness Safety

    May 14, 2026Chengzhi Liu, Yichen Guo, Yepeng Liu +8Agent HarnessRuntime Enforcement