StepGuard: Learning Step-Level Guardrails with Scalable Supervision and Safety-Utility Balancing

2026-08-25

AI Summary

为解决LLM代理在工具调用中的安全风险，提出StepGuard模型通过预执行监控和动态平衡安全与非安全动作来审计并检查每一步操作。

Abstract

LLM-based agents can interact with external environments through tool invocation, but this capability also introduces security risks such as file modification, information leakage, and unauthorized actions. Existing guardrails often evaluate completed trajectories, leaving pre-execution monitoring of step-level actions underexplored. We propose StepGuard, a step-level guard model that can audit completed agent trajectories and check tool actions before they are executed. To train StepGuard, we introduce StepGen, an automatic data engine that generates safe and unsafe trajectories with the same context but different actions at the risky step. To further reduce over-defense and under-defense, we propose Balance-GRPO, which dynamically balances learning between safe and unsafe actions based on their observed accuracy. Experiments show that StepGuard achieves the highest average accuracy among open-weight guard models, with performance comparable to GPT-5.4. When used to guard agents on AgentDojo and AgentDyn, StepGuard reduces mean attack success rate by 77.3% relative to the no-guard setting, while mean utility drops by only 2.8 percentage points.

Problem

• LLM-based agents
• security risks
• step-level actions
• pre-execution monitoring
• guardrails

Innovation

• StepGuard
• Step-Level Guardrails
• StepGen
• Safety-Utility Balancing
• Balance-GRPO