LedgerAgent: Structured State for Policy-Adherent Tool-Calling Agents

Hugging Face Daily Papers Papers

Summary

LedgerAgent is a method for customer service agents that maintains task states in a separate ledger to improve policy adherence and state management during tool calling. It improves average passk over standard approaches across four domains.

Policy-adherent tool-calling agents in customer-service domains must maintain task states across turns while calling tools and obeying domain policies. Task states consist of relevant facts, identifiers, constraints, and conditions observed through user interaction and tool calls. In standard agents, task states are not represented separately. Observations, tool returns, and policy instructions are placed in the prompt, leaving agents to reconstruct the relevant states from the prompt each time they decide what to do next. This design makes state management implicit, creating two common failure modes. An agent may retrieve the right facts but later ground its decision in stale, missing, or incorrect information; and a syntactically valid tool call may still violate a domain policy that depends on the current task state. We introduce LedgerAgent, an inference-time method for tool-calling agents that maintains observed task states in a separate ledger and renders the states into the prompt. The ledger is also used to check state-dependent policy constraints before environment-changing tool calls are executed, blocking policy violations. Across four customer-service domains and a mixed panel of open- and closed-weight models, LedgerAgent improves average passk over a standard prompt-based tool-calling approach, with the largest gains under stricter multi-trial consistency metrics.
Original Article
View Cached Full Text

Cached at: 06/20/26, 02:26 PM

Paper page - LedgerAgent: Structured State for Policy-Adherent Tool-Calling Agents

Source: https://huggingface.co/papers/2606.20529 Published on Jun 18

·

Submitted byhttps://huggingface.co/sahsaeedi

Amiron Jun 19

Abstract

LEDGERAGENT is a method for customer service agents that maintains task states in a separate ledger to improve policy adherence and state management during tool calling.

Policy-adherenttool-calling agentsin customer-service domains must maintaintask statesacross turns while calling tools and obeyingdomain policies.Task statesconsist of relevant facts, identifiers, constraints, and conditions observed through user interaction and tool calls. In standard agents,task statesare not represented separately. Observations, tool returns, and policy instructions are placed in the prompt, leaving agents to reconstruct the relevant states from the prompt each time they decide what to do next. This design makes state management implicit, creating two common failure modes. An agent may retrieve the right facts but later ground its decision in stale, missing, or incorrect information; and a syntactically valid tool call may still violate a domain policy that depends on the current task state. We introduceLedgerAgent, an inference-time method fortool-calling agentsthat maintains observedtask statesin a separateledgerand renders the states into the prompt. Theledgeris also used to check state-dependentpolicy constraintsbefore environment-changing tool calls are executed, blocking policy violations. Across four customer-service domains and a mixed panel of open- and closed-weight models,LedgerAgent improves average passk over a standard prompt-based tool-calling approach, with the largest gains under strictermulti-trial consistencymetrics.

View arXiv pageView PDFAdd to collection

Get this paper in your agent:

hf papers read 2606\.20529

Don’t have the latest CLI?curl \-LsSf https://hf\.co/cli/install\.sh \| bash

Models citing this paper0

No model linking this paper

Cite arxiv.org/abs/2606.20529 in a model README.md to link it from this page.

Datasets citing this paper0

No dataset linking this paper

Cite arxiv.org/abs/2606.20529 in a dataset README.md to link it from this page.

Spaces citing this paper0

No Space linking this paper

Cite arxiv.org/abs/2606.20529 in a Space README.md to link it from this page.

Collections including this paper0

No Collection including this paper

Add this paper to acollectionto link it from this page.

Similar Articles

PolicyBank: Evolving Policy Understanding for LLM Agents

arXiv cs.CL

PolicyBank proposes a memory mechanism that enables LLM agents to autonomously refine their understanding of organizational policies through iterative interaction and corrective feedback, closing specification gaps that cause systematic behavioral divergence from true requirements. The work introduces a systematic testbed and demonstrates PolicyBank can close up to 82% of policy-gap alignment failures, significantly outperforming existing memory mechanisms.