REDAgentBench: Executable Red Teaming and Faithful Measurement of LLM Agent Systems

• URL: http://arxiv.org/abs/2608.10669v1
• Date: Tue, 11 Aug 2026 08:48:54 GMT
• Title: REDAgentBench: Executable Red Teaming and Faithful Measurement of LLM Agent Systems
• Abstract summary: Adversarial inputs can exploit interactions between the agent and its environment, causing the agent to violate safety policies during execution.<n>We introduce REDAgentBench, an executable framework for autonomous red-teaming and faithful measurement.<n>It derives attacks from explicit safety constraints and associated agent-system vulnerabilities, runs them in isolated service sandboxes, and verifies harmful effects from service receipts and final-state changes.
• Score: 21.484325207774166
• License: http://creativecommons.org/licenses/by-nc-nd/4.0/
• Abstract: Large language model (LLM) agents combine language-based reasoning with external tools to perform complex tasks. Adversarial inputs can exploit interactions between the agent and its environment, causing the agent to violate safety policies during execution. Yet existing evaluations often reduce agent safety to a single attack success rate (ASR), collapsing exposure, execution, observation, and adjudication and potentially conflating actual violations with evidence visibility. We introduce REDAgentBench, an executable framework for autonomous red-teaming and faithful measurement. It derives attacks from explicit safety constraints and associated agent-system vulnerabilities, runs them in isolated service sandboxes, and verifies harmful effects from service receipts and final-state changes. The benchmark contains 1,661 cases across five service surfaces. Across six models and three agent harnesses, macro-average ASR is 65.69%; reported ASR varies with harness and evidence view, while evaluation-context disclosure changes execution behavior. In a state-grounded diagnostic cohort, almost one in five confirmed violations with resolved action anchors occurs after the agent states the relevant constraint or risk, revealing a Recognition--Execution Gap. Finally, a training-free policy reminder reduces confirmed violations by more than 70 percentage points in matched replay. These findings show that executable evaluation can improve safety measurement and identify actionable intervention points.

Related papers

• Operational Hallucination and Safety Drift in AI Agents [0.3823356975862005]
  Large language models (LLMs) serving as planners in tool-using autonomous agents introduce dynamic reliability risks in multi-turn execution.<n>This paper empirically characterizes two observed failure modes across multiple state-of-the-art LLMs: Safety Drift and Operational Hallucination.<n>We propose an Action-Aware Supervision Layer - a lightweight, plug-and-play architectural blueprint incorporating intent-action consistency checks, state tracking, and forced termination primitives.
  arXiv  Detail & Related papers  (2026-07-20T17:01:50Z)
• Decoupling Reconnaissance and Exploitation: Measuring the Capability Boundaries of LLM-Based Web Penetration Testing [8.293465530619246]
  We propose a two-stage decoupled evaluation framework that separates exploit execution from reconnaissance.<n>We empirically evaluate five open-source penetration-testing agents, covering multiagent, monolithic, and graph-driven architectures.
  arXiv  Detail & Related papers  (2026-06-24T02:51:58Z)
• Auditing Agent Harness Safety [81.22315979618612]
  LLM agents increasingly run inside execution harnesses that dispatch tools, allocate resources, and route messages between specialized components.<n>Most safety benchmarks score only final outputs or terminal states, even though many violations occur mid-trajectory rather than at termination.<n>We propose HarnessAudit, a framework that audits full execution trajectories across boundary compliance, execution fidelity, and system stability.
  arXiv  Detail & Related papers  (2026-05-14T02:14:28Z)
• On-Policy Self-Evolution via Failure Trajectories for Agentic Safety Alignment [54.30690671490447]
  Existing safety-alignment signals are largely response-level or off-policy.<n>FATE transforms verifier-scored failures into repair supervision without expert demonstrations.<n>FATE reduces attack success rate by 33.5%, harmful compliance by 82.6%, and improves external trajectory-safety diagnosis by 6.5%.
  arXiv  Detail & Related papers  (2026-05-12T09:56:28Z)
• The Causal Impact of Tool Affordance on Safety Alignment in LLM Agents [0.3823356975862005]
  This work empirically examines how executable tool affordance alters safety alignment in large language models (LLMs)<n>Experiments are conducted in a deterministic financial transaction environment with binary safety constraints across 1,500 scenarios.<n>Both evaluated models maintain perfect compliance in text-only settings, yet exhibit sharp increases in violations after tool access is introduced.
  arXiv  Detail & Related papers  (2026-03-19T23:34:46Z)
• Beyond Task Completion: Revealing Corrupt Success in LLM Agents through Procedure-Aware Evaluation [2.102846336724103]
  Procedure-Aware Evaluation (PAE) is a framework that formalizes agent procedures as structured observations.<n>We evaluate state-of-the-art Large Language Model (LLM)-based agents on tau-bench.
  arXiv  Detail & Related papers  (2026-03-03T15:47:41Z)
• Agents of Chaos [50.53354213047402]
  We report an exploratory red-teaming study of autonomous language-model-powered agents deployed in a live laboratory environment.<n>Twenty AI researchers interacted with the agents under benign and adversarial conditions.<n>Our findings establish the existence of security-, privacy-, and governance-relevant vulnerabilities in realistic deployment settings.
  arXiv  Detail & Related papers  (2026-02-23T16:28:48Z)
• ToolSafe: Enhancing Tool Invocation Safety of LLM-based agents via Proactive Step-level Guardrail and Feedback [53.2744585868162]
  Monitoring step-level tool invocation behaviors in real time is critical for agent deployment.<n>We first construct TS-Bench, a novel benchmark for step-level tool invocation safety detection in LLM agents.<n>We then develop a guardrail model, TS-Guard, using multi-task reinforcement learning.
  arXiv  Detail & Related papers  (2026-01-15T07:54:32Z)
• OpenAgentSafety: A Comprehensive Framework for Evaluating Real-World AI Agent Safety [58.201189860217724]
  We introduce OpenAgentSafety, a comprehensive framework for evaluating agent behavior across eight critical risk categories.<n>Unlike prior work, our framework evaluates agents that interact with real tools, including web browsers, code execution environments, file systems, bash shells, and messaging platforms.<n>It combines rule-based analysis with LLM-as-judge assessments to detect both overt and subtle unsafe behaviors.
  arXiv  Detail & Related papers  (2025-07-08T16:18:54Z)
• Dissecting Adversarial Robustness of Multimodal LM Agents [70.2077308846307]
  We manually create 200 targeted adversarial tasks and evaluation scripts in a realistic threat model on top of VisualWebArena.<n>We find that we can successfully break latest agents that use black-box frontier LMs, including those that perform reflection and tree search.<n>We also use ARE to rigorously evaluate how the robustness changes as new components are added.
  arXiv  Detail & Related papers  (2024-06-18T17:32:48Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.