Cover Image

SafeHarbor: Hierarchical Memory-Augmented Guardrail for LLM Agent Safety

Fuente: arXiv

Saved in:

Bibliographic Details
Main Authors:	Liu, Zhe, Ying, Zonghao, Zhang, Wenxin, Zou, Quanchen, Zhang, Deyue, Yang, Dongdong, Zhang, Xiangzheng, Peng, Hao
Format:	Preprint
Published:	2026
Subjects:	Cryptography and Security Artificial Intelligence 68T42
Online Access:	Acceder al recurso
Tags:	Add Tag No Tags, Be the first to tag this record!

Similar Items

DMN: A Compositional Framework for Jailbreaking Multimodal LLMs with Multi-Image Inputs
by: Xu, Wenzhuo, et al.
Published: (2026)

Reasoning-Oriented Programming: Chaining Semantic Gadgets to Jailbreak Large Vision Language Models
by: Zou, Quanchen, et al.
Published: (2026)

Probabilistic Modeling of Jailbreak on Multimodal LLMs: From Quantification to Application
by: Xu, Wenzhuo, et al.
Published: (2025)

Robust Privacy: Inference-Time Privacy through Certified Robustness
by: Jin, Jiankai, et al.
Published: (2026)

Towards Understanding the Safety Boundaries of DeepSeek Models: Evaluation and Findings
by: Ying, Zonghao, et al.
Published: (2025)

Reasoning-Augmented Conversation for Multi-Turn Jailbreak Attacks on Large Language Models
by: Ying, Zonghao, et al.
Published: (2025)

PRISM: Programmatic Reasoning with Image Sequence Manipulation for LVLM Jailbreaking
by: Zou, Quanchen, et al.
Published: (2025)

Two Frames Matter: A Temporal Attack for Text-to-Video Model Jailbreaking
by: Chen, Moyang, et al.
Published: (2026)

Sequential Comics for Jailbreaking Multimodal Large Language Models via Structured Visual Storytelling
by: Zhang, Deyue, et al.
Published: (2025)

Mask-GCG: Are All Tokens in Adversarial Suffixes Necessary for Jailbreak Attacks?
by: Mu, Junjie, et al.
Published: (2025)

AgentBreeder: Mitigating the AI Safety Risks of Multi-Agent Scaffolds via Self-Improvement
by: Rosser, J, et al.
Published: (2025)

Portable Agent Memory: A Protocol for Cryptographically-Verified Memory Transfer Across Heterogeneous AI Agents
by: Ravindran, Santhosh Kumar
Published: (2026)

Multimodal Multi-Agent Ransomware Analysis Using AutoGen
by: Khan, Asifullah, et al.
Published: (2026)

AgentVisor: Defending LLM Agents Against Prompt Injection via Semantic Virtualization
by: Ying, Zonghao, et al.
Published: (2026)

Evolving Deception: When Agents Evolve, Deception Wins
by: Ying, Zonghao, et al.
Published: (2026)

BlockA2A: Towards Secure and Verifiable Agent-to-Agent Interoperability
by: Zou, Zhenhua, et al.
Published: (2025)

SecureWebArena: A Holistic Security Evaluation Benchmark for LVLM-based Web Agents
by: Ying, Zonghao, et al.
Published: (2025)

Demystifying and Detecting Agentic Workflow Injection Vulnerabilities in GitHub Actions
by: Wang, Shenao, et al.
Published: (2026)

SPARK: Jailbreaking T2V Models by Synergistically Prompting Auditory and Recontextualized Knowledge
by: Ying, Zonghao, et al.
Published: (2025)

PoseGuard: Pose-Guided Generation with Safety Guardrails
by: Wang, Kongxin, et al.
Published: (2025)

An Organization-Scoped LLM Agent Runtime Architecture for Regulated Cybersecurity Operations
by: Fatouros, George, et al.
Published: (2026)

Control Physiology: An Agent-Based Model of FAIR-CAM Dynamics
by: Jones, Jack, et al.
Published: (2026)

Tool Receipts, Not Zero-Knowledge Proofs: Practical Hallucination Detection for AI Agents
by: Basu, Abhinaba
Published: (2026)

SafeBench: A Safety Evaluation Framework for Multimodal Large Language Models
by: Ying, Zonghao, et al.
Published: (2024)

Defense Against Indirect Prompt Injection via Tool Result Parsing
by: Yu, Qiang, et al.
Published: (2026)

When the Agent Is the Adversary: Architectural Requirements for Agentic AI Containment After the April 2026 Frontier Model Escape
by: Mitchell, Richard Joseph
Published: (2026)

MemoPhishAgent: Memory-Augmented Multi-Modal LLM Agent for Phishing URL Detection
by: Chen, Xuan, et al.
Published: (2026)

Provably Secure Agent Guardrail
by: Wu, Benlong, et al.
Published: (2026)

Simple Prompt Injection Attacks Can Leak Personal Data Observed by LLM Agents During Task Execution
by: Alizadeh, Meysam, et al.
Published: (2025)

AGENTSAFE: Benchmarking the Safety of Embodied Agents on Hazardous Instructions
by: Ying, Zonghao, et al.
Published: (2025)

AgentSys: Secure and Dynamic LLM Agents Through Explicit Hierarchical Memory Management
by: Wen, Ruoyao, et al.
Published: (2026)

A V2X-based Privacy Preserving Federated Measuring and Learning System
by: Alekszejenkó, Levente, et al.
Published: (2024)

Quantigence: A Multi-Agent AI Framework for Quantum Security Research
by: Alquwayfili, Abdulmalik
Published: (2025)

Poly-Guard: Massive Multi-Domain Safety Policy-Grounded Guardrail Dataset
by: Kang, Mintong, et al.
Published: (2025)

DLP: towards active defense against backdoor attacks with decoupled learning process
by: Ying, Zonghao, et al.
Published: (2024)

NBA: defensive distillation for backdoor removal via neural behavior alignment
by: Ying, Zonghao, et al.
Published: (2024)

A Comparative Evaluation of AI Agent Security Guardrails
by: Li, Qi, et al.
Published: (2026)

Enhancing Guardrails for Safe and Secure Healthcare AI
by: Gangavarapu, Ananya
Published: (2024)

Hoist with His Own Petard: Inducing Guardrails to Facilitate Denial-of-Service Attacks on Retrieval-Augmented Generation of LLMs
by: Suo, Pan, et al.
Published: (2025)

Interpretable LLM Guardrails via Sparse Representation Steering
by: He, Zeqing, et al.
Published: (2025)