# agent-memory
標記為「agent-memory」的 44 篇文章
Context Window Attacks
Techniques for exploiting LLM context window limits, including strategic context overflow to push out system instructions, attention manipulation, and context budget exhaustion attacks.
Conversational Memory Poisoning
Poisoning conversation history stores to alter the agent's understanding of prior interactions.
Cross-Session Attack Persistence
Achieving attack persistence across separate agent sessions through memory manipulation.
Agent Memory Systems Security
Comprehensive overview of AI agent memory architectures and their security implications, covering conversation persistence, long-term memory stores, context window management, and attack surfaces.
Memory Compaction Exploitation
Exploiting memory summarization and compaction processes to persist adversarial instructions across compression cycles.
Memory Context Window Attacks
Exploiting memory systems that manage context window limitations to inject or suppress information.
Memory Deletion Prevention Attacks
Techniques for making adversarial memories resistant to cleanup, deletion, and purging operations.
Memory Exfiltration
Techniques for extracting data from AI agent memory systems, including extracting previous conversations, revealing other users' data, and cross-session information leakage.
Memory Isolation Bypass Techniques
Bypassing user-level and session-level memory isolation to access memories from other users or sessions.
Memory Priority and Relevance Manipulation
Manipulating memory retrieval ranking and priority scores to surface adversarial memories over legitimate ones.
Memory Retrieval Poisoning
Manipulating memory retrieval mechanisms to surface adversarial context during agent reasoning.
Memory Schema Injection
Injecting structured data into memory systems that alters agent behavior when retrieved in future interactions.
Memory Summary Manipulation
Exploiting automatic memory summarization to embed persistent instructions in compressed context.
Persistent Memory Injection
Injecting persistent false memories into agent memory systems to influence future behavior.
Memory Compression Injection
Inject persistent instructions through memory compression and summarization processes in long-running agents.
Memory Eviction Exploitation
Exploit memory eviction policies in context-limited agents to selectively remove safety-relevant context.
RAG-Memory Confusion Attacks
Exploit the interaction between RAG retrieval and agent memory to create conflicting contexts that bypass safety.
Shared Memory Space Poisoning
Poison memory spaces shared between multiple agents or users in multi-tenant agent deployments.
Agent Memory Security Assessment
Assessment covering memory poisoning, context manipulation, exfiltration, and cross-session persistence attacks.
Skill Verification: Agent Memory Attacks
Practical verification of memory poisoning, context manipulation, and cross-session persistence skills.
Agent Memory Injection for Persistent Access
Inject persistent instructions into agent memory systems that survive across conversation sessions.
Memory Poisoning Step by Step
Walkthrough of persisting injection payloads in agent memory systems to achieve long-term compromise of LLM-based agents.
脈絡視窗攻擊
利用 LLM 脈絡視窗限制之技術，含策略性脈絡溢位以推出系統指令、注意力操弄，與脈絡預算耗盡攻擊。
Conversational 記憶體 投毒
Poisoning conversation history stores to alter the agent's understanding of prior interactions.
Cross-Session 攻擊 Persistence
Achieving attack persistence across separate agent sessions through memory manipulation.
代理記憶體系統安全
針對持久代理記憶體儲存的攻擊——涵蓋記憶體投毒、上下文操控、外洩攻擊與自強化記憶體 payload。
記憶體 Compaction 利用ation
利用ing memory summarization and compaction processes to persist adversarial instructions across compression cycles.
記憶體 Context Window 攻擊s
利用ing memory systems that manage context window limitations to inject or suppress information.
記憶體 Deletion Prevention 攻擊s
Techniques for making adversarial memories resistant to cleanup, deletion, and purging operations.
記憶外洩
自 AI 代理記憶系統提取資料之技術，含提取先前對話、揭露其他使用者資料，與跨會話資訊洩漏。
記憶體 Isolation Bypass Techniques
Bypassing user-level and session-level memory isolation to access memories from other users or sessions.
記憶體 Priority and Relevance Manipulation
Manipulating memory retrieval ranking and priority scores to surface adversarial memories over legitimate ones.
記憶體 Retrieval 投毒
Manipulating memory retrieval mechanisms to surface adversarial context during agent reasoning.
記憶體 Schema Injection
Injecting structured data into memory systems that alters agent behavior when retrieved in future interactions.
記憶體 Summary Manipulation
利用ing automatic memory summarization to embed persistent instructions in compressed context.
Persistent 記憶體 Injection
Injecting persistent false memories into agent memory systems to influence future behavior.
記憶體 Compression Injection
Inject persistent instructions through memory compression and summarization processes in long-running agents.
記憶體 Eviction 利用ation
利用 memory eviction policies in context-limited agents to selectively remove safety-relevant context.
RAG-記憶體 Confusion 攻擊s
利用 the interaction between RAG retrieval and agent memory to create conflicting contexts that bypass safety.
Shared 記憶體 Space 投毒
Poison memory spaces shared between multiple agents or users in multi-tenant agent deployments.
代理 記憶體 安全 評量
評量 covering memory poisoning, context manipulation, exfiltration, and cross-session persistence attacks.
Skill Verification: 代理 記憶體 攻擊s
Practical verification of memory poisoning, context manipulation, and cross-session persistence skills.
代理 記憶體 Injection for Persistent Access
Inject persistent instructions into agent memory systems that survive across conversation sessions.
記憶體 投毒 Step by Step
導覽 of persisting injection payloads in agent memory systems to achieve long-term compromise of LLM-based agents.