Mengxiao Wang

Mengxiao Wang

Ph.D. student, SPIES Lab, Texas A&M University.

Publications

You can also find my articles on my Google Scholar profile.

Under Review / Preprints

SoK: Trading Agents or Market Crashers? Dissecting Robustness and Security Failures in Academic Financial LLM Trading Schemes

Published in Under review, 2026

A systematization-of-knowledge study of the security and robustness failures reported across academic LLM-based financial trading schemes. The paper catalogs attack surfaces — prompt injection, jailbreak, prompt leaking, tool misuse — and evaluates published defenses under a common threat model.

Recommended citation: Mengxiao Wang and Nitesh Saxena. (2026). "SoK: Trading Agents or Market Crashers? Dissecting Robustness and Security Failures in Academic Financial LLM Trading Schemes." Under review.

GuardianPWA: Enhancing Security Throughout the Progressive Web App Installation Lifecycle

Published in arXiv:2509.13561, 2025

A defense that follows a Progressive Web App from install prompt through installation and runtime, closing gaps where a hostile PWA can slip privileges past the user or between browser and OS.

Recommended citation: Mengxiao Wang and Guofei Gu. (2025). "GuardianPWA: Enhancing Security Throughout the Progressive Web App Installation Lifecycle." arXiv preprint arXiv:2509.13561.
Download Paper

PromptSleuth: Detecting Prompt Injection via Semantic Intent Invariance

Published in arXiv:2508.20890, 2025

A detector for prompt-injection attacks that reasons about the semantic intent of the input rather than matching surface patterns. Intent-invariant checks stay robust when attackers paraphrase, encode, or reorder the injected instruction.

Recommended citation: Mengxiao Wang, Yuxuan Zhang, and Guofei Gu. (2025). "PromptSleuth: Detecting Prompt Injection via Semantic Intent Invariance." arXiv preprint arXiv:2508.20890.
Download Paper

Conference Papers

Demystifying Progressive Web Application Permission Systems

Published in 56th IEEE/IFIP International Conference on Dependable Systems and Networks (DSN), 2026

A measurement and analysis of the permission systems Progressive Web Applications inherit from the browser and add on top of it, and how the two layers interact in practice.

Recommended citation: Mengxiao Wang, Jiajun Huang, and Guofei Gu. (2026). "Demystifying Progressive Web Application Permission Systems." Proc. of the 56th IEEE/IFIP International Conference on Dependable Systems and Networks (DSN).

Beyond Visual Confusion: Understanding How Inconsistencies in ENS Normalization Facilitate Homoglyph Attacks

Published in The Web Conference (WWW), 2025

Shows that ENS homoglyph attacks are not only a visual-similarity problem but also a normalization-inconsistency problem: different clients and resolvers disagree on how to canonicalize a name, and the disagreement itself is exploitable.

Recommended citation: Jiajun Huang, Sai Ravan Chintapalli, Mengxiao Wang, and Guofei Gu. (2025). "Beyond Visual Confusion: Understanding How Inconsistencies in ENS Normalization Facilitate Homoglyph Attacks." Proceedings of the ACM on Web Conference 2025.

Journal Articles