Another 240papers do not have category tags yet and are not included in filtered results.
提出 COBRA 双 LLM 架构，防御计算机使用 Agent 的分支转向攻击
Abstract提出分支转向防御架构 COBRA，通过事前能力约束与边界代理消除分支转向攻击并兼顾效用。
提出 ToolFence，以能力蓝图和确定性授权拦截 Agent 未授权工具调用
提出 Tracekit，审计编码 Agent 的意图、推理与动作并检测账本篡改
Tracekit 是无第三方依赖的编码智能体审计层：把人类请求、模型自述和已执行动作并排写入可外部锚定的哈希链，执行前用正则门，事后用规则和独立模型互查。
提出可验证操作卡 VAC，把智能体浏览器确认绑定到真实 DOM 操作
提出 KITA，把 Agent 批准绑定到规范动作的密钥隔离阈值签名
KITA 是一套密钥隔离的阈值授权架构，用来把模型评审变成执行外部动作的条件。
提出 SkillGuard，用可达性限制污染后的 Agent 能力
提出 AgentFlow 流策略语言与运行时框架，约束智能体跨步数据流
提出 CAVA，将异构智能体运行时事件规范为可哈希动作并绑定审批与回执
CAVA 是给异构智能体运行时用的动作语义层，用来回答“被批准和被审计的到底是哪一个动作”。
Proposes Prismata, which uses critical-path annotation and mechanical isolation to constrain cross-site prompt injection in web agents.
Reviews threats to autonomous agents and proposes a full-lifecycle defense blueprint, FASA
This is a threat review and defense architecture report using OpenClaw as a case study, not a scored attack or defense evaluation.
Proposes a defense-in-depth aligned with the five stages to mitigate cross-stage threats to autonomous agents
This is a lifecycle threat analysis of OpenClaw, accompanied by a set of layered mitigation designs. The problem addressed is: when autonomous agents connect instant messaging, persistent memory, third-party plugins, and high-privilege execution, attacks are no longer limited to single-turn jailbreaks but can compound across multiple stages from initialization to execution. The authors argue that defenses hardening only individual interfaces cannot stop such cross-temporal attacks.
提出 Aura，用可信内核中介移动 Agent 的身份、感知与动作控制
Aura uses kernel-mediated structured agent interaction, replacing screen-interface mobile assistants..