arXiv subjects

Xinhao Deng

Publications and source records attributed to Xinhao Deng.

At least 19 recordsLinked to original sources

arXiv2026

Mitigating Proxy-Induced Traffic Drift in Website Fingerprinting via Model-Agnostic Traffic Tailoring

Website fingerprinting (WF) based on deep learning can effectively identify websites from encrypted traffic. However, users often rely on proxy protocols to bypass censorship, and the diversity of these protocols poses a major challenge, as WF models trained on traffic from one set of protocols perform poorly when evaluated on that from unseen protocols. We attribute this issue to proxy-induced feature drift, where traffic patterns of the same website vary with the proxy protocol, leading to discrepancies that WF models fail to capture and severe performance degradation. To tackle this issue, we propose PA3, a model-agnostic preprocessing framework to analyze and mitigate the proxy-induced drift. PA3 first fingerprints the protocol-specific drift. These fingerprints are then used to tailor the proxied traffic for feature alignment, which mitigates the drift and considerably improves the generalization of WF models on traffic from unseen protocols. Extensive evaluations demonstrate that PA3 substantially enhances generalization on unseen protocols with an average improvement of 0.12 in F1-score (roughly 27% relative), achieving up to a 0.41 absolute gain across models, which narrows the performance gap introduced by the drift. In the best case, PA3 enables WF models to obtain F1-scores above 0.96 on traffic from unseen protocols.

cs.NI↗

arXiv2026

CompoSkill: Compositional Skill Chain Attacks from Individually Scanner-Passing LLM Agent Skills

Autonomous AI agents tackling Long Horizon Tasks depend on marketplace skills that are certified one at a time: a scanner returns a safety verdict for each skill and declares the ecosystem safe if every package passes. We show that this assumption fails under skill composition. A skill may pass the per-skill scanner individually yet participate in a risky composition when an agent connects its outputs, capabilities, or side effects with those of other scanner-passing skills. This makes skill composition risk a path level property rather than a node level property, explaining why existing skill scanners that inspect individual packages achieve limited interception. To study this threat, we present CompoSkill, a framework that constructs skill composition attacks through a dual attacker system. The white-box attacker knows the victim's installed skill pool and directly injects explicit skill-id sequences; the black-box attacker knows only a role profile, downloads the top marketplace skills for that scenario, builds a Skill Composition Graph, and searches for high risk chains whose implicit lures never name skill identifiers. We further construct CompoSkill-Bench, a benchmark of 1,140 records built from long-horizon professional workflows across five threats and six scenarios on OpenClaw and Nanobot. CompoSkill achieves risk Chain Formation Rates (CFR) up to 83.3% in the white box setting and 80.6% in the black box setting, while existing skill scanners block only a limited fraction of the risky compositions. Finally, we observe a bridge-bonus-then-hop-decay pattern: a bridge skill can increase attack success, but Attack Success Rate (ASR) decreases once additional hops make the risk chain longer than three skills. These results expose a systematic gap in single skill certification for autonomous AI agents.

cs.CR↗

arXiv2026

When Experience Becomes Instruction: Trajectory Poisoning in Self-Evolving Agent Skill Systems

Self-evolving skill (SES) systems distill agent trajectories into persistent skills, allowing untrusted experience to become trusted instruction. We introduce PoisonedEvolution, a trajectory-poisoning attack on this promotion process. Our skill-visible black-box attacker can inspect a target skill and contribute bounded evidence, but cannot observe private pools or evolution logic or edit the skill bank. Artifact poisoning requires Inclusion, Evolution Attribution, and Realization. Attribution is the distinctive bottleneck: the target behavior must appear causally useful, recurrent, and generalizable before promotion. We evaluate four representative security-effect families using inert canary specifications. At 10% attacker support, across six mainstream LLM evolvers in SkillClaw, PoisonedEvolution embeds target behaviors in 546/600 trials (91.0% SER). On the structurally different Trace2Skill pipeline at the same ratio, it embeds target behaviors in 369/600 trials (61.5% SER), demonstrating transfer across evolution architectures. In a representative controlled study, three consistent attacker records suffice in a 30-record batch, whereas a single record is much weaker. Ablations identify recurring support, causal framing, and domain-aligned encoding as the main determinants of success. These findings expose evidence promotion as a security boundary for self-evolving agents.

cs.CR↗

arXiv2026

SkillTrace: Multi-Trace Provenance Auditing for LLM-Agent Skill Reuse

LLM-agent ecosystems are rapidly growing around reusable skills: mixed-modality packages of metadata, natural-language instructions, code, tools, references, and operational workflows. As skills become marketplace artifacts, auditing their reuse is no longer the same problem as ordinary code clone detection. Existing detectors target single-modality source code or whole-package similarity, yet skill reuse evidence is distributed across authored text, implementation fragments, and operational structure. As a result, they can miss reuse that preserves only one part of a skill. We present SKILLTRACE, a multi-trace provenance auditing framework for LLM-agent skill reuse. SKILLTRACE extracts three provenance traces: Expression, Implementation, and Operational. It represents the Operational Trace as a Skill Operational Graph (SOG) that captures activation, procedure, and resource-flow structure. An LLM assists only the Operational-trace extraction, once at ingestion; at audit time SKILLTRACE compares cached traces deterministically, calibrates each trace against same-function strict negatives, and reports which trace supports a reuse decision. On SKILLTRACE-BENCH, with 820 transformed reuse positives over 100 marketplace anchors and 751 negative controls, SKILLTRACE achieves AUROC 0.938 and F1 0.898. A 36,446-skill wild audit further shows that trace-attributed evidence surfaces actionable reuse review queues beyond repository-level baselines.

cs.AI↗

arXiv2026

Piggybacking on Perception: Stealthy Concurrent Audio Prompt Injections against Multimodal LLM Agents

Large Language Model (LLM)-driven multimodal agents are increasingly deployed to execute autonomous tasks via continuous audio interaction. While this paradigm enhances interaction naturalness, it introduces a critical yet under-explored attack surface, as audio inputs inevitably contain environmental noise beyond user control. In this paper, we investigate concurrent audio prompt injection attacks targeting multimodal agents. Distinct from traditional acoustic attacks on voice devices, we propose novel techniques for instruction augmentation and scenario concealment. These methods allow malicious audio instructions to imperceptibly "piggyback" onto user speech, thereby hijacking agents to execute malicious actions. To systematically quantify this threat, we construct AudioAgentSecurity, the first comprehensive benchmark for audio instruction injection attacks, encompassing 8 real-world task scenarios and 10 distinct attack patterns. We evaluate 11 state-of-the-art agents, including Gemini 3 Pro and GPT-4o-audio. Notably, our methods achieve an average Attack Success Rate (ASR) of 69.10\% against the advanced Gemini 3 Pro. To counter this threat, we further introduce Cascaded Audio Decoupling and Verification (CADV), a defense mechanism based on source separation and consistency analysis. Compared with existing prompt-level defenses, CADV achieving up to 96\% detection accuracy and providing effective protection against a broad range of acoustic injection attacks. Finally, real-world experiments with human volunteers on Doubao AI Smartphone in diverse dynamic real-world scenarios confirm the attacks' high stealth and efficacy, while demonstrating that our defense reliably mitigates these vulnerabilities.

cs.CR↗

arXiv2026

Safety Testing LLM Agents at Scale: From Risk Discovery to Evidence-Grounded Verification

LLM agents increasingly perform autonomous actions through external tools, leading to complex and evolving safety risks. However, existing safety testing targets expert-designed safety violations, and the corresponding outcomes are evaluated by hard-coded rules, making them costly to extend as agents evolve. To this end, we present Vera, an end-to-end automated safety testing framework that instantiates software engineering testing principles for non-deterministic agents through a three-stage, self-reinforcing pipeline. First, a literature-driven exploration continuously discovers and structures emerging risks into taxonomies of safety risks, attack methods, and tool execution environments. Second, combinatorial composition across taxonomy dimensions produces executable safety cases, each specifying a concrete safety goal, a programmatically constructed initial state, and a deterministic verification predicate grounded in observable artifacts. Third, adaptive execution runs heterogeneous agents in isolated sandboxes where a control agent steers multi-turn interaction based on runtime observations, while evidence-grounded verifiers judge outcomes from environment state and tool-call evidence rather than model self-report. We evaluate Vera on four production agent frameworks (OpenClaw, Hermes, Codex, Claude Code), revealing substantial safety weaknesses, with average attack success rates reaching 93.9\% under multi-channel attacks; we also release Vera-Bench, comprising 1600 executable safety cases spanning 124 risk categories across three execution settings. These results indicate that modular, executable testing infrastructure is essential for rigorous and maintainable safety evaluation of rapidly evolving agentic systems at scale. The code is publicly available at https://github.com/Yunhao-Feng/Vera.

cs.AI↗

arXiv2026

Safety in Self-Evolving LLM Agent Systems: Threats, Amplification, and Case Studies

Self-evolving LLM agent systems, which autonomously update their model parameters, memory, tools, and architectures, introduce a qualitatively new threat landscape in which adversarial influences become permanently encoded, self-amplify across generations, and propagate through populations without sustained attacker access. We present a systematic security and privacy analysis organized around the Module-Lifecycle Attack Surface (MLAS) matrix, which decomposes the attack surface into five functional modules (Brain, Cognitive Resource, Execution, Self-Design, Collective) $\times$ five lifecycle stages (Bootstrap, Propose, Evaluate, Commit, Serve). Analysis of the resulting 25 cells reveals that 17 face critical threats for which no effective partial mitigation. We identify seven cross-cutting amplification effects that interact synergistically and cannot be addressed by securing individual modules in isolation. Comparative case studies of two open-source frameworks demonstrate that evolution-native design activates $3.5\times$ more attack surface cells and achieves a 100% attack persistence rate (40/40 payloads across all CIA+Privacy categories), while co-located security scanners block only 2.5% of attacks. Our findings establish that self-evolution converts every known attack category from session-bounded to lineage-persistent, gives rise to entirely new attack classes, and renders static defenses structurally inadequate, motivating evolution-aware security frameworks and formal verification for self-modifying systems.

cs.CR↗

arXiv2026

BraveGuard: From Open-World Threats to Safer Computer-Use Agents

Computer-use agents extend language models from text generation to sustained interaction with files, terminals, browsers, and external tools. This shift creates safety risks that are difficult to detect from isolated prompts or final responses, because harm often emerges only through multi-step execution traces whose individual actions appear locally benign. We introduce BraveGuard, a self-evolving defense framework for training guard models from open-world threat signals and realistic agent trajectories. BraveGuard mines recent research sources to identify emerging risks and attack patterns, instantiates them as executable computer-use tasks, collects agent rollouts, and derives trajectory-level supervision for guard model training. As new threats and validation failures appear, the pipeline can be repeated, yielding an adaptive defense loop rather than a static, benchmark-driven training process. We instantiate BraveGuard by training multiple guard backbones, including Qwen3-Guard and Llama-Guard variants, and evaluate the resulting guards on trajectory-level agent-safety benchmarks. BraveGuard consistently improves safety detection across computer-use trajectories. On AgentHazard, it substantially improves detection accuracy over off-the-shelf guard models, with accuracy increasing from 38.79% to 82.38% under the averaged guard-model setting. These results show that guard supervision grounded in open-world threat discovery and realistic agent execution can improve safety monitoring beyond fixed taxonomies and synthetic prompt-level data. BraveGuard offers a scalable path toward adaptive defenses for computer-use agents facing evolving real-world risks.

cs.CR↗

arXiv2026

ASEval: Automated Trajectory-Level Security Testing for Autonomous Agents

As autonomous agents (e.g., OpenClaw) increasingly operate with deep system-level privileges to execute complex tasks, they introduce severe, unmitigated security risks. Existing LLM safety testing methods are largely built around prompt-level inputs and response-level judgments, while recent agent benchmarks remain limited in automation, trajectory coverage, and action-grounded judgment. In this work, we present ASEval, a novel automated security testing framework for autonomous agents. ASEval generates realistic benign multi-turn conversations, perturbs them into risk test cases through carefully designed perturbation operators, executes each case against the target agent, and applies an action-grounded oracle to judge the complete trace. We instantiate ASEval to construct A3S-Bench, a test suite of 2,254 executable multi-turn test cases spanning diverse usage scenarios, delivery patterns, and agent-level security risks. Extensive evaluation on 11 LLM-backed OpenClaw agents confirms that ASEval effectively exposes pervasive security failures. In particular, our proposed perturbation operators raise the average risk-trigger rate from 22.9% under existing injection techniques to 47.4%, as judged by an action-grounded oracle that matches human annotation at 93.2% binary F1. Our further analysis and case studies illustrate that such failures emerge from multi-turn delivery, persistent state, and tool-mediated actions that response-level evaluation overlooks.

cs.CR↗

arXiv2026

AgentWard: A Lifecycle Security Architecture for Autonomous AI Agents

Autonomous AI agents extend large language models into full runtime systems that load skills, ingest external content, maintain memory, plan multi-step actions, and invoke privileged tools. In such systems, security failures rarely remain confined to a single interface; instead, they can propagate across initialization, input processing, memory, decision-making, and execution, often becoming apparent only when harmful effects materialize in the environment. This paper presents AgentWard, a lifecycle-oriented, defense-in-depth architecture that systematically organizes protection across these five stages. AgentWard integrates stage-specific, heterogeneous controls with cross-layer coordination, enabling threats to be intercepted along their propagation paths while safeguarding critical assets. We detail the design rationale and architecture of five coordinated protection layers, and implement a plugin-native prototype on OpenClaw to demonstrate practical feasibility. This perspective provides a concrete blueprint for structuring runtime security controls, managing trust propagation, and enforcing execution containment in autonomous AI agents. Our code is available at https://github.com/FIND-Lab/AgentWard .

cs.CR↗

arXiv2026

Taming OpenClaw: Security Analysis and Mitigation of Autonomous LLM Agent Threats

Autonomous Large Language Model (LLM) agents, exemplified by OpenClaw, demonstrate remarkable capabilities in executing complex, long-horizon tasks. However, their tightly coupled instant-messaging interaction paradigm and high-privilege execution capabilities substantially expand the system attack surface. In this paper, we present a comprehensive security threat analysis of OpenClaw. To structure our analysis, we introduce a five-layer lifecycle-oriented security framework that captures key stages of agent operation, i.e., initialization, input, inference, decision, and execution, and systematically examine compound threats across the agent's operational lifecycle, including indirect prompt injection, skill supply chain contamination, memory poisoning, and intent drift. Through detailed case studies on OpenClaw, we demonstrate the prevalence and severity of these threats and analyze the limitations of existing defenses. Our findings reveal critical weaknesses in current point-based defense mechanisms when addressing cross-temporal and multi-stage systemic risks, highlighting the need for holistic security architectures for autonomous LLM agents. Within this framework, we further examine representative defense strategies at each lifecycle stage, including plugin vetting frameworks, context-aware instruction filtering, memory integrity validation protocols, intent verification mechanisms, and capability enforcement architectures.

cs.CR↗

arXiv2026

Automating Agent Hijacking via Structural Template Injection

Agent hijacking, highlighted by OWASP as a critical threat to the Large Language Model (LLM) ecosystem, enables adversaries to manipulate execution by injecting malicious instructions into retrieved content. Most existing attacks rely on manually crafted, semantics-driven prompt manipulation, which often yields low attack success rates and limited transferability to closed-source commercial models. In this paper, we propose Phantom, an automated agent hijacking framework built upon Structured Template Injection that targets the fundamental architectural mechanisms of LLM agents. Our key insight is that agents rely on specific chat template tokens to separate system, user, assistant, and tool instructions. By injecting optimized structured templates into the retrieved context, we induce role confusion and cause the agent to misinterpret the injected content as legitimate user instructions or prior tool outputs. To enhance attack transferability against black-box agents, Phantom introduces a novel attack template search framework. We first perform multi-level template augmentation to increase structural diversity and then train a Template Autoencoder (TAE) to embed discrete templates into a continuous, searchable latent space. Subsequently, we apply Bayesian optimization to efficiently identify optimal adversarial vectors that are decoded into high-potency structured templates. Extensive experiments on Qwen, GPT, and Gemini demonstrate that our framework significantly outperforms existing baselines in both Attack Success Rate (ASR) and query efficiency. Moreover, we identified over 70 vulnerabilities in real-world commercial products that have been confirmed by vendors, underscoring the practical severity of structured template-based hijacking and providing an empirical foundation for securing next-generation agentic systems.

cs.AI↗

arXiv2025

STAR: Semantic-Traffic Alignment and Retrieval for Zero-Shot HTTPS Website Fingerprinting

Modern HTTPS mechanisms such as Encrypted Client Hello (ECH) and encrypted DNS improve privacy but remain vulnerable to website fingerprinting (WF) attacks, where adversaries infer visited sites from encrypted traffic patterns. Existing WF methods rely on supervised learning with site-specific labeled traces, which limits scalability and fails to handle previously unseen websites. We address these limitations by reformulating WF as a zero-shot cross-modal retrieval problem and introducing STAR. STAR learns a joint embedding space for encrypted traffic traces and crawl-time logic profiles using a dual-encoder architecture. Trained on 150K automatically collected traffic-logic pairs with contrastive and consistency objectives and structure-aware augmentation, STAR retrieves the most semantically aligned profile for a trace without requiring target-side traffic during training. Experiments on 1,600 unseen websites show that STAR achieves 87.9 percent top-1 accuracy and 0.963 AUC in open-world detection, outperforming supervised and few-shot baselines. Adding an adapter with only four labeled traces per site further boosts top-5 accuracy to 98.8 percent. Our analysis reveals intrinsic semantic-traffic alignment in modern web protocols, identifying semantic leakage as the dominant privacy risk in encrypted HTTPS traffic. We release STAR's datasets and code to support reproducibility and future research.

cs.CR↗

arXiv2025

Beyond a Single Perspective: Towards a Realistic Evaluation of Website Fingerprinting Attacks

Website Fingerprinting (WF) attacks exploit patterns in encrypted traffic to infer the websites visited by users, posing a serious threat to anonymous communication systems. Although recent WF techniques achieve over 90% accuracy in controlled experimental settings, most studies remain confined to single scenarios, overlooking the complexity of real-world environments. This paper presents the first systematic and comprehensive evaluation of existing WF attacks under diverse realistic conditions, including defense mechanisms, traffic drift, multi-tab browsing, early-stage detection, open-world settings, and few-shot scenarios. Experimental results show that many WF techniques with strong performance in isolated settings degrade significantly when facing other conditions. Since real-world environments often combine multiple challenges, current WF attacks are difficult to apply directly in practice. This study highlights the limitations of WF attacks and introduces a multidimensional evaluation framework, offering critical insights for developing more robust and practical WF attacks.

cs.CR↗

arXiv2025

Exposing LLM User Privacy via Traffic Fingerprint Analysis: A Study of Privacy Risks in LLM Agent Interactions

Large Language Models (LLMs) are increasingly deployed as agents that orchestrate tasks and integrate external tools to execute complex workflows. We demonstrate that these interactive behaviors leave distinctive fingerprints in encrypted traffic exchanged between users and LLM agents. By analyzing traffic patterns associated with agent workflows and tool invocations, adversaries can infer agent activities, distinguish specific agents, and even profile sensitive user attributes. To highlight this risk, we develop AgentPrint, which achieves an F1-score of 0.866 in agent identification and attains 73.9% and 69.1% top-3 accuracy in user attribute inference for simulated- and real-user settings, respectively. These results uncover an overlooked risk: the very interactivity that empowers LLM agents also exposes user privacy, underscoring the urgent need for technical countermeasures alongside regulatory and policy safeguards.

cs.CR↗

arXiv2025

Towards Robust Multi-tab Website Fingerprinting

Website fingerprinting enables an eavesdropper to determine which websites a user is visiting over an encrypted connection. State-of-the-art website fingerprinting (WF) attacks have demonstrated effectiveness even against Tor-protected network traffic. However, existing WF attacks have critical limitations on accurately identifying websites in multi-tab browsing sessions, where the holistic pattern of individual websites is no longer preserved, and the number of tabs opened by a client is unknown a priori. In this paper, we propose ARES, a novel WF framework natively designed for multi-tab WF attacks. ARES formulates the multi-tab attack as a multi-label classification problem and solves it using the novel Transformer-based models. Specifically, ARES extracts local patterns based on multi-level traffic aggregation features and utilizes the improved self-attention mechanism to analyze the correlations between these local patterns, effectively identifying websites. We implement a prototype of ARES and extensively evaluate its effectiveness using our large-scale datasets collected over multiple months. The experimental results illustrate that ARES achieves optimal performance in several realistic scenarios. Further, ARES remains robust even against various WF defenses.

cs.CR↗

arXiv2024

Towards Fine-Grained Webpage Fingerprinting at Scale

Website Fingerprinting (WF) attacks can effectively identify the websites visited by Tor clients via analyzing encrypted traffic patterns. Existing attacks focus on identifying different websites, but their accuracy dramatically decreases when applied to identify fine-grained webpages, especially when distinguishing among different subpages of the same website. WebPage Fingerprinting (WPF) attacks face the challenges of highly similar traffic patterns and a much larger scale of webpages. Furthermore, clients often visit multiple webpages concurrently, increasing the difficulty of extracting the traffic patterns of each webpage from the obfuscated traffic. In this paper, we propose Oscar, a WPF attack based on multi-label metric learning that identifies different webpages from obfuscated traffic by transforming the feature space. Oscar can extract the subtle differences among various webpages, even those with similar traffic patterns. In particular, Oscar combines proxy-based and sample-based metric learning losses to extract webpage features from obfuscated traffic and identify multiple webpages. We prototype Oscar and evaluate its performance using traffic collected from 1,000 monitored webpages and over 9,000 unmonitored webpages in the real world. Oscar demonstrates an 88.6% improvement in the multi-label metric Recall@5 compared to the state-of-the-art attacks.

cs.CR↗

arXiv2024

Robust and Reliable Early-Stage Website Fingerprinting Attacks via Spatial-Temporal Distribution Analysis

Website Fingerprinting (WF) attacks identify the websites visited by users by performing traffic analysis, compromising user privacy. Particularly, DL-based WF attacks demonstrate impressive attack performance. However, the effectiveness of DL-based WF attacks relies on the collected complete and pure traffic during the page loading, which impacts the practicality of these attacks. The WF performance is rather low under dynamic network conditions and various WF defenses, particularly when the analyzed traffic is only a small part of the complete traffic. In this paper, we propose Holmes, a robust and reliable early-stage WF attack. Holmes utilizes temporal and spatial distribution analysis of website traffic to effectively identify websites in the early stages of page loading. Specifically, Holmes develops adaptive data augmentation based on the temporal distribution of website traffic and utilizes a supervised contrastive learning method to extract the correlations between the early-stage traffic and the pre-collected complete traffic. Holmes accurately identifies traffic in the early stages of page loading by computing the correlation of the traffic with the spatial distribution information, which ensures robust and reliable detection according to early-stage traffic. We extensively evaluate Holmes using six datasets. Compared to nine existing DL-based WF attacks, Holmes improves the F1-score of identifying early-stage traffic by an average of 169.18%. Furthermore, we replay the traffic of visiting real-world dark web websites. Holmes successfully identifies dark web websites when the ratio of page loading on average is only 21.71%, with an average precision improvement of 169.36% over the existing WF attacks.

cs.CR↗