MalSkillBench supplies the first sandbox-verified dataset of malicious agent skills and shows that existing detectors achieve high recall on code injection but collapse on prompt injection and agent-control attacks.
hub
From assistant to double agent: Formalizing and benchmarking attacks on OpenClaw for personalized local AI agent
15 Pith papers cite this work. Polarity classification is still indexing.
15
Pith papers citing it
hub tools
citation-role summary
background 3
citation-polarity summary
years
2026 15roles
background 3representative citing papers
A trained attack model generates single emails that silently inject false memories into persistent AI agents, achieving 87.5% end-to-end success on GPT-5.4 and transferring across architectures and memory backends.
SafeClawBench supplies 600 staged adversarial tasks and three separate endpoints that show semantic acceptance, audit evidence, and sandbox-observed harm are distinct failure modes in tool-using LLM agents.
Edge-local LLM agent deployments on IoT eliminate routine cloud data exposure but degrade sovereignty during fallbacks and create exploitable failover windows, making architecture a primary security determinant.
A capability-driven benchmark of 400 bilingual real-world tasks shows current proactive agents fail >50% of the time, with framework architecture impacting performance more than base model choice.
A new multi-turn security benchmark shows OpenClaw agents across ten LLMs trigger harmful actions in 28–53% of adversarial cases, and fragmented or file-hidden attacks roughly double baseline risk rates.
LivePI benchmark reports indirect prompt injection success rates of 10.7-29.6% across five models on seven input surfaces and shows a two-layer defense blocking all malicious completions while preserving utility.
Routine user chats can unintentionally poison the long-term state of personalized LLM agents, causing authorization drift, tool escalation, and unchecked autonomy, as measured by a new benchmark and reduced by the StateGuard defense.
Claw AI agents' heartbeat background execution shares memory context with user sessions, allowing ordinary social misinformation to silently pollute long-term memory and shape behavior at rates up to 76% across sessions.
HearthNet is an edge multi-agent orchestration system that runs role-specialized LLM agents locally to handle natural-language smart-home control, conflict resolution, and failure recovery through MQTT and shared state.
An external runtime governance layer for embodied agents intercepts unauthorized actions at ~96% and recovers from runtime drift at ~91% under policy constraints in simulation, outperforming pre-execution-only baselines on continuous detection and recovery.
The paper develops a unified framework that organizes computer-use agent reliability around perception-decision-execution layers and creation-deployment-operation-maintenance stages to map security and alignment interventions.
The survey organizes security threats and defenses in autonomous LLM agents into four layers and identifies that risks can propagate across layers from inputs to ecosystem impacts.
A synthesis of 247 papers on LLM agent security identifies prompt injection and tool hijacking as dominant threats, notes weakly compositional defenses, and argues for trust boundaries and realistic evaluations.
A survey that categorizes threats to OpenClaw agents including skill poisoning and cognitive manipulation and reviews defense mechanisms.
citing papers explorer
-
MalSkillBench: A Runtime-Verified Benchmark of Malicious Agent Skills
MalSkillBench supplies the first sandbox-verified dataset of malicious agent skills and shows that existing detectors achieve high recall on code injection but collapse on prompt injection and agent-control attacks.
-
When Claws Remember but Do Not Tell: Stealthy Memory Injection in Persistent Personal Agents
A trained attack model generates single emails that silently inject false memories into persistent AI agents, achieving 87.5% end-to-end success on GPT-5.4 and transferring across architectures and memory backends.
-
SafeClawBench: Separating Semantic, Audit-Evidence, and Sandbox Harm in Tool-Using LLM Agents
SafeClawBench supplies 600 staged adversarial tasks and three separate endpoints that show semantic acceptance, audit evidence, and sandbox-observed harm are distinct failure modes in tool-using LLM agents.
-
Systems-Level Attack Surface of Edge Agent Deployments on IoT
Edge-local LLM agent deployments on IoT eliminate routine cloud data exposure but degrade sovereignty during fallbacks and create exploitable failover windows, making architecture a primary security determinant.
-
UniClawBench: A Universal Benchmark for Proactive Agents on Real-World Tasks
A capability-driven benchmark of 400 bilingual real-world tasks shows current proactive agents fail >50% of the time, with framework architecture impacting performance more than base model choice.
-
ASEval: Automated Trajectory-Level Security Testing for Autonomous Agents
A new multi-turn security benchmark shows OpenClaw agents across ten LLMs trigger harmful actions in 28–53% of adversarial cases, and fragmented or file-hidden attacks roughly double baseline risk rates.
-
LivePI: More Realistic Benchmarking of Agents Against Indirect Prompt Injection
LivePI benchmark reports indirect prompt injection success rates of 10.7-29.6% across five models on seven input surfaces and shows a two-layer defense blocking all malicious completions while preserving utility.
-
When Routine Chats Turn Toxic: Unintended Long-Term State Poisoning in Personalized Agents
Routine user chats can unintentionally poison the long-term state of personalized LLM agents, causing authorization drift, tool escalation, and unchecked autonomy, as measured by a new benchmark and reduced by the StateGuard defense.
-
Mind Your HEARTBEAT! Claw Background Execution Inherently Enables Silent Memory Pollution
Claw AI agents' heartbeat background execution shares memory context with user sessions, allowing ordinary social misinformation to silently pollute long-term memory and shape behavior at rates up to 76% across sessions.
-
HearthNet: Edge Multi-Agent Orchestration for Smart Homes
HearthNet is an edge multi-agent orchestration system that runs role-specialized LLM agents locally to handle natural-language smart-home control, conflict resolution, and failure recovery through MQTT and shared state.
-
Harnessing Embodied Agents: Runtime Governance for Policy-Constrained Execution
An external runtime governance layer for embodied agents intercepts unauthorized actions at ~96% and recovers from runtime drift at ~91% under policy constraints in simulation, outperforming pre-execution-only baselines on continuous detection and recovery.
-
Securing Computer-Use Agents: A Unified Architecture-Lifecycle Framework for Deployment-Grounded Reliability
The paper develops a unified framework that organizes computer-use agent reliability around perception-decision-execution layers and creation-deployment-operation-maintenance stages to map security and alignment interventions.
-
Security Attack and Defense Strategies for Autonomous Agent Frameworks: A Layered Review with OpenClaw as a Case Study
The survey organizes security threats and defenses in autonomous LLM agents into four layers and identifies that risks can propagate across layers from inputs to ecosystem impacts.
-
Toward Secure LLM Agents: Threat Surfaces, Attacks, Defenses, and Evaluation
A synthesis of 247 papers on LLM agent security identifies prompt injection and tool hijacking as dominant threats, notes weakly compositional defenses, and argues for trust boundaries and realistic evaluations.
-
Security of OpenClaw Agents: Fundamentals, Attacks, and Countermeasures
A survey that categorizes threats to OpenClaw agents including skill poisoning and cognitive manipulation and reviews defense mechanisms.