ClawTrap: A MITM-Based Red-Teaming Framework for Real-World OpenClaw Security Evaluation

TL;DR

ClawTrap is a versatile MITM-based red-teaming framework designed to evaluate the security of OpenClaw web agents in real-world network conditions, revealing vulnerabilities and model robustness differences.

Contribution

The paper introduces ClawTrap, a novel, customizable MITM attack framework for practical security testing of OpenClaw agents in live network environments.

Findings

Weaker models trust tampered observations more, leading to unsafe outputs.

Stronger models better attribute anomalies and fallback safely.

Real-world MITM testing reveals security gaps not seen in static benchmarks.

Abstract

Autonomous web agents such as \textbf{OpenClaw} are rapidly moving into high-impact real-world workflows, but their security robustness under live network threats remains insufficiently evaluated. Existing benchmarks mainly focus on static sandbox settings and content-level prompt attacks, which leaves a practical gap for network-layer security testing. In this paper, we present \textbf{ClawTrap}, a \textbf{MITM-based red-teaming framework for real-world OpenClaw security evaluation}. ClawTrap supports diverse and customizable attack forms, including \textit{Static HTML Replacement}, \textit{Iframe Popup Injection}, and \textit{Dynamic Content Modification}, and provides a reproducible pipeline for rule-driven interception, transformation, and auditing. This design lays the foundation for future research to construct richer, customizable MITM attacks and to perform systematic security testing across agent frameworks and model backbones. Our empirical study shows clear model stratification: weaker models are more likely to trust tampered observations and produce unsafe outputs, while stronger models demonstrate better anomaly attribution and safer fallback strategies. These findings indicate that reliable OpenClaw security evaluation should explicitly incorporate dynamic real-world MITM conditions rather than relying only on static sandbox protocols.