WAREX: Web Agent Reliability Evaluation on Existing Benchmarks

Recent advances in browser-based LLM agents have shown promise for automating tasks ranging from simple form filling to hotel booking or online shopping. Current benchmarks measure agent performance in controlled environments, such as containers or stable networks, where websites behave deterministically. However, in the real world, users access websites over networks and HTTPS connections that introduce instability from multiple sources: client-side, server-side issues or broader system failures. Moreover, live websites are prone to web attacks such Cross-Site Scripting, as well as general site modifications which can cause unexpected or malicious pop-ups or improper functionality. To address this gap, we present WAREX, a plug-and-play, network-layer tool that integrates with existing web agent benchmarks by simulating common website failures. We measure the impact of WAREX across three popular benchmarks: WebArena, WebVoyager, and REAL. Our experiments show that introducing WAREX leads to significant drops in task success rates, highlighting the limited robustness of state-of-the-art agents. We demonstrate that WAREX serves as more than a diagnostic tool. By fine-tuning an open-source model (Qwen3-8B) on WAREX-generated "failure-recovery" trajectories, we achieve an 88.9% relative improvement in error recovery rates , validating WAREX as a core component for training the next generation of reliable web agents.