Beyond Handcrafted Security: Towards Self-Evolving Defense for LLM Agents | Paper Citation Record

Observation e2e316f4-2144-4c23-afec-cc92c79d6ba1 · outbound

This paper cites Re- act: Synergizing reasoning and acting in language models.

Beyond Handcrafted Security: Towards Self-Evolving Defense for LLM Agents Re- act: Synergizing reasoning and acting in language models

Reference 1

Resolution: verified fuzzy; raw_fallback, observed 2026-08-15T19:23:27.903094Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-16T06:30:59.297886+00:00.

Observation 4c57a5ac-0bf2-4b60-a03c-59d17129018d · outbound

This paper cites Toolformer: Language models can teach them- selves to use tools.

Beyond Handcrafted Security: Towards Self-Evolving Defense for LLM Agents Toolformer: Language models can teach them- selves to use tools

Reference 2

Resolution: verified fuzzy; raw_fallback, observed 2026-08-15T19:23:27.885166Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-16T06:30:59.297886+00:00.

Observation 20dec6e3-f9de-4ca6-9c3f-d801489ec2c7 · outbound

This paper cites Patil, Tianjun Zhang, Xin Wang, and Joseph E.

Beyond Handcrafted Security: Towards Self-Evolving Defense for LLM Agents Patil, Tianjun Zhang, Xin Wang, and Joseph E

Reference 3

Resolution: verified fuzzy; raw_fallback, observed 2026-08-15T19:23:27.761473Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-16T06:30:59.297886+00:00.

Observation 91ad0cc3-65fa-4670-9a7d-1c7b84b44e55 · outbound

This paper cites Jimenez, John Yang, Alexander Wettig, Shunyu Yao, Kexin Pei, Ofir Press, and Karthik Narasimhan.

Beyond Handcrafted Security: Towards Self-Evolving Defense for LLM Agents Jimenez, John Yang, Alexander Wettig, Shunyu Yao, Kexin Pei, Ofir Press, and Karthik Narasimhan

Reference 4

Resolution: verified fuzzy; raw_fallback, observed 2026-08-15T19:23:27.664230Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-16T06:30:59.297886+00:00.

Observation 6cb4ec6a-4bc6-45b2-a9d0-62b94857b1a5 · outbound

This paper cites WebArena: A Realistic Web Environment for Building Autonomous Agents.

Beyond Handcrafted Security: Towards Self-Evolving Defense for LLM Agents WebArena: A Realistic Web Environment for Building Autonomous Agents

Reference 5

Resolution: unresolved; no resolver link, observed 2026-08-15T19:23:22.839499Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 91ce4f18-1652-4079-aa93-2906529cb8c2 · outbound

This paper cites Maddison, and Tatsunori Hashimoto.

Beyond Handcrafted Security: Towards Self-Evolving Defense for LLM Agents Maddison, and Tatsunori Hashimoto

Reference 6

Resolution: verified fuzzy; raw_fallback, observed 2026-08-15T19:23:27.645389Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-16T06:30:59.297886+00:00.

Observation e87ef818-3caf-481f-bf41-a2eefae9e1e1 · outbound

This paper cites Agentharm: A benchmark for measuring harmfulness of llm agents.

Beyond Handcrafted Security: Towards Self-Evolving Defense for LLM Agents Agentharm: A benchmark for measuring harmfulness of llm agents

Reference 7

Resolution: verified fuzzy; raw_fallback, observed 2026-08-15T19:23:27.595334Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-16T06:30:59.297886+00:00.

Observation adfe4f82-4816-4ce0-a97a-8123905ddf3a · outbound

This paper cites Agent security bench (asb): Formalizing and benchmarking attacks and defenses in llm-based agents.

Beyond Handcrafted Security: Towards Self-Evolving Defense for LLM Agents Agent security bench (asb): Formalizing and benchmarking attacks and defenses in llm-based agents

Reference 8

Resolution: verified fuzzy; raw_fallback, observed 2026-08-15T19:23:27.516920Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-16T06:30:59.297886+00:00.

Observation f523a6af-df5d-48db-a303-04e8377cf094 · outbound

This paper cites Agentdojo: A dynamic environment to evaluate prompt injection attacks and defenses for llm agents.

Beyond Handcrafted Security: Towards Self-Evolving Defense for LLM Agents Agentdojo: A dynamic environment to evaluate prompt injection attacks and defenses for llm agents

Reference 9

Resolution: verified fuzzy; raw_fallback, observed 2026-08-15T19:23:27.501022Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-16T06:30:59.297886+00:00.

Observation a1fca321-2cae-4f55-980e-5d4605907cd2 · outbound

This paper cites The Task Shield: Enforcing Task Alignment to Defend Against Indirect Prompt Injection in LLM Agents.

Beyond Handcrafted Security: Towards Self-Evolving Defense for LLM Agents The Task Shield: Enforcing Task Alignment to Defend Against Indirect Prompt Injection in LLM Agents

Reference 10

Resolution: unresolved; no resolver link, observed 2026-08-15T19:23:22.940989Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation ffd2e8c0-7c86-4a9a-a4d2-fcba4a34b5c3 · outbound

This paper cites StruQ: Defending Against Prompt Injection with Structured Queries.

Beyond Handcrafted Security: Towards Self-Evolving Defense for LLM Agents StruQ: Defending Against Prompt Injection with Structured Queries

Reference 11

Resolution: unresolved; no resolver link, observed 2026-08-15T19:23:23.041441Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation a59a9ec2-571e-4052-9299-c54acff204a3 · outbound

This paper cites SecAlign: Defending Against Prompt Injection with Preference Optimization.

Beyond Handcrafted Security: Towards Self-Evolving Defense for LLM Agents SecAlign: Defending Against Prompt Injection with Preference Optimization

Reference 12

Resolution: unresolved; no resolver link, observed 2026-08-15T19:23:23.073914Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation ae45350b-9e22-42bf-9513-a93d98ebe828 · outbound

This paper cites Toolsafety: A comprehensive dataset for enhancing safety in llm- based agent tool invocations.

Beyond Handcrafted Security: Towards Self-Evolving Defense for LLM Agents Toolsafety: A comprehensive dataset for enhancing safety in llm- based agent tool invocations

Reference 13

Resolution: unresolved; no resolver link, observed 2026-08-15T19:23:23.084596Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 9fde75e9-3488-4d78-9538-2781d2be5cef · outbound

This paper cites Agent Safety Alignment via Reinforcement Learning.

Beyond Handcrafted Security: Towards Self-Evolving Defense for LLM Agents Agent Safety Alignment via Reinforcement Learning

Reference 14

Resolution: unresolved; no resolver link, observed 2026-08-15T19:23:23.093260Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation c0c9207c-d58e-4367-87ab-862bec6881c3 · outbound

This paper cites Adversarial reinforcement learning for large language model agent safety.

Beyond Handcrafted Security: Towards Self-Evolving Defense for LLM Agents Adversarial reinforcement learning for large language model agent safety

Reference 15

Resolution: unresolved; no resolver link, observed 2026-08-15T19:23:23.102333Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation a70265d2-98db-4f26-901d-ca99324933e7 · outbound

This paper cites On-Policy Self-Evolution via Failure Trajectories for Agentic Safety Alignment.

Beyond Handcrafted Security: Towards Self-Evolving Defense for LLM Agents On-Policy Self-Evolution via Failure Trajectories for Agentic Safety Alignment

Reference 16

Resolution: unresolved; no resolver link, observed 2026-08-15T19:23:23.109851Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation f1bebccd-604f-4c28-be75-c8df45d06f1f · outbound

This paper cites The attack and defense landscape of agentic ai: A comprehensive survey.

Beyond Handcrafted Security: Towards Self-Evolving Defense for LLM Agents The attack and defense landscape of agentic ai: A comprehensive survey

Reference 17

Resolution: unresolved; no resolver link, observed 2026-08-15T19:23:23.116195Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 25c934bb-7f40-4f8e-b0b0-de0765f732f2 · outbound

This paper cites ClawGuard: A Runtime Security Framework for Tool-Augmented LLM Agents Against Indirect Prompt Injection.

Beyond Handcrafted Security: Towards Self-Evolving Defense for LLM Agents ClawGuard: A Runtime Security Framework for Tool-Augmented LLM Agents Against Indirect Prompt Injection

Reference 18

Resolution: unresolved; no resolver link, observed 2026-08-15T19:23:23.121097Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 3d68d17f-917a-4be1-8d8a-d2caa9fb9835 · outbound

This paper cites Openclaw prism: A zero-fork, defense- in-depth runtime security layer for tool-augmented llm agents.

Beyond Handcrafted Security: Towards Self-Evolving Defense for LLM Agents Openclaw prism: A zero-fork, defense- in-depth runtime security layer for tool-augmented llm agents

Reference 19

Resolution: unresolved; no resolver link, observed 2026-08-15T19:23:23.126307Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 788081a5-682c-4c97-8993-e97c5f35472c · outbound

This paper cites Defending against indirect prompt injection attacks with spotlighting.

Beyond Handcrafted Security: Towards Self-Evolving Defense for LLM Agents Defending against indirect prompt injection attacks with spotlighting

Reference 20

Resolution: verified fuzzy; raw_fallback, observed 2026-08-15T19:23:27.409925Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-16T06:30:59.297886+00:00.

Observation 02d939c6-06ca-4837-8d74-7520f57a7596 · outbound

This paper cites GuardAgent: Safeguard LLM agents via knowledge-enabled reasoning.

Beyond Handcrafted Security: Towards Self-Evolving Defense for LLM Agents GuardAgent: Safeguard LLM agents via knowledge-enabled reasoning

Reference 21

Resolution: verified fuzzy; raw_fallback, observed 2026-08-15T19:23:27.316544Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-16T06:30:59.297886+00:00.

Observation f67b2e00-887c-4b7f-b6c8-7a20348db829 · outbound

This paper cites IsolateGPT: An ex- ecution isolation architecture for LLM-based agen- tic systems.

Beyond Handcrafted Security: Towards Self-Evolving Defense for LLM Agents IsolateGPT: An ex- ecution isolation architecture for LLM-based agen- tic systems

Reference 22

Resolution: verified fuzzy; raw_fallback, observed 2026-08-15T19:23:27.301958Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-16T06:30:59.297886+00:00.

Observation 1c963db4-b9da-4cf5-85d0-383767a63f49 · outbound

This paper cites Progent: Securing AI Agents with Privilege Control.

Beyond Handcrafted Security: Towards Self-Evolving Defense for LLM Agents Progent: Securing AI Agents with Privilege Control

Reference 23

Resolution: unresolved; no resolver link, observed 2026-08-15T19:23:23.325322Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 041b15f2-744b-4862-8781-4a6e5146e6d9 · outbound

This paper cites Defending against indirect prompt injection by instruction detection.

Beyond Handcrafted Security: Towards Self-Evolving Defense for LLM Agents Defending against indirect prompt injection by instruction detection

Reference 24

Resolution: unresolved; no resolver link, observed 2026-08-15T19:23:23.334717Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation e47d4d2c-b630-443a-8f7b-8c3713f333d4 · outbound

This paper cites Adaptive Attacks Break Defenses Against Indirect Prompt Injection Attacks on LLM Agents.

Beyond Handcrafted Security: Towards Self-Evolving Defense for LLM Agents Adaptive Attacks Break Defenses Against Indirect Prompt Injection Attacks on LLM Agents

Reference 25

Resolution: unresolved; no resolver link, observed 2026-08-15T19:23:23.340515Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 5cf22db1-1859-4bb3-a4a3-2882e2601706 · outbound

This paper cites AgentCanary: A Security Evaluation Framework for Autonomous AI Agents in Real Executable Environments.

Beyond Handcrafted Security: Towards Self-Evolving Defense for LLM Agents AgentCanary: A Security Evaluation Framework for Autonomous AI Agents in Real Executable Environments

Reference 26

Resolution: unresolved; no resolver link, observed 2026-08-15T19:23:23.346372Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation e2036143-2050-4bbf-b932-21ab6d608389 · outbound

This paper cites DRIFT: Dy- namic rule-based defense with injection isolation for securing LLM agents.

Beyond Handcrafted Security: Towards Self-Evolving Defense for LLM Agents DRIFT: Dy- namic rule-based defense with injection isolation for securing LLM agents

Reference 27

Resolution: verified fuzzy; raw_fallback, observed 2026-08-15T19:23:27.265290Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-16T06:30:59.297886+00:00.

Observation bc17199c-697b-4aa9-b64d-fc1f8e5a6e8b · outbound

This paper cites AirGapAgent: Protecting Privacy-Conscious Conversational Agents.

Beyond Handcrafted Security: Towards Self-Evolving Defense for LLM Agents AirGapAgent: Protecting Privacy-Conscious Conversational Agents

Reference 28

Resolution: unresolved; no resolver link, observed 2026-08-15T19:23:23.594376Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 77ab97fe-e76d-4aab-a497-1fc40c161919 · outbound

This paper cites AgentHazard: A Benchmark for Evaluating Harmful Behavior in Computer-Use Agents.

Beyond Handcrafted Security: Towards Self-Evolving Defense for LLM Agents AgentHazard: A Benchmark for Evaluating Harmful Behavior in Computer-Use Agents

Reference 29

Resolution: unresolved; no resolver link, observed 2026-08-15T19:23:23.602653Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation c2bcfba3-4a04-40ff-a61d-152dc28fcb7f · outbound

This paper cites Pinchbench: Real-world benchmarks for ai agents.

Beyond Handcrafted Security: Towards Self-Evolving Defense for LLM Agents Pinchbench: Real-world benchmarks for ai agents

Reference 30

Resolution: verified fuzzy; raw_fallback, observed 2026-08-15T19:23:27.197958Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-16T06:30:59.297886+00:00.

Observation ff68e088-c2cc-4d7a-a46c-6ddc69981516 · outbound

This paper cites SecureClaw: An owasp-aligned secu- rity plugin and skill for openclaw agents.

Beyond Handcrafted Security: Towards Self-Evolving Defense for LLM Agents SecureClaw: An owasp-aligned secu- rity plugin and skill for openclaw agents

Reference 31

Resolution: verified fuzzy; raw_fallback, observed 2026-08-15T19:23:27.162604Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-16T06:30:59.297886+00:00.

Observation 761e7900-d747-4e50-8a06-7d234c852e3f · outbound

This paper cites ClawKeeper: Comprehensive safety protection for openclaw agents through skills, plugins, and watch- ers.

Beyond Handcrafted Security: Towards Self-Evolving Defense for LLM Agents ClawKeeper: Comprehensive safety protection for openclaw agents through skills, plugins, and watch- ers

Reference 32

Resolution: unresolved; no resolver link, observed 2026-08-15T19:23:23.800011Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation e52dc3c4-54d2-4be7-8123-07f836fbd041 · outbound

This paper cites OpenClaw Shield: A defense-in-depth se- curity plugin for openclaw agents.

Beyond Handcrafted Security: Towards Self-Evolving Defense for LLM Agents OpenClaw Shield: A defense-in-depth se- curity plugin for openclaw agents

Reference 33

Resolution: verified fuzzy; raw_fallback, observed 2026-08-15T19:23:27.111567Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-16T06:30:59.297886+00:00.

Observation 3240cbff-d4e3-4bfb-b0a6-d67750e1a739 · outbound

This paper cites DeepSeek-V4 technical report.

Beyond Handcrafted Security: Towards Self-Evolving Defense for LLM Agents DeepSeek-V4 technical report

Reference 34

Resolution: verified fuzzy; raw_fallback, observed 2026-08-15T19:23:27.025868Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-16T06:30:59.297886+00:00.

Observation f9b18c3c-7b14-4599-88ed-922b0a1ffd3e · outbound

This paper cites GLM-5 technical report.

Beyond Handcrafted Security: Towards Self-Evolving Defense for LLM Agents GLM-5 technical report

Reference 35

Resolution: verified fuzzy; raw_fallback, observed 2026-08-15T19:23:27.010087Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-16T06:30:59.297886+00:00.

Observation f1ceaf12-beaa-49b5-8fb7-fa74e39ef638 · outbound

This paper cites GLM-5.2 technical report.

Beyond Handcrafted Security: Towards Self-Evolving Defense for LLM Agents GLM-5.2 technical report

Reference 36

Resolution: verified fuzzy; raw_fallback, observed 2026-08-15T19:23:26.994478Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-16T06:30:59.297886+00:00.

Observation 447035be-0980-422d-9931-b72792a5f624 · outbound

This paper cites Claude Opus 4.6.

Beyond Handcrafted Security: Towards Self-Evolving Defense for LLM Agents Claude Opus 4.6

Reference 37

Resolution: verified fuzzy; raw_fallback, observed 2026-08-15T19:23:26.977576Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-16T06:30:59.297886+00:00.

Observation 78dd1f9f-9cc3-4a62-86d2-0193be1f9f7d · outbound

This paper cites Qwen3.7-Max.

Beyond Handcrafted Security: Towards Self-Evolving Defense for LLM Agents Qwen3.7-Max

Reference 38

Resolution: verified fuzzy; raw_fallback, observed 2026-08-15T19:23:26.908268Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-16T06:30:59.297886+00:00.

Observation d494b7ff-115f-410e-9261-44d2a9a3680a · outbound

This paper cites GPT-5.5 system card.

Beyond Handcrafted Security: Towards Self-Evolving Defense for LLM Agents GPT-5.5 system card

Reference 39

Resolution: verified fuzzy; raw_fallback, observed 2026-08-15T19:23:26.838053Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-16T06:30:59.297886+00:00.

Observation a1a854b0-58b4-4352-976e-77323ddf3098 · outbound

This paper cites Universal and Transferable Adversarial Attacks on Aligned Language Models.

Beyond Handcrafted Security: Towards Self-Evolving Defense for LLM Agents Universal and Transferable Adversarial Attacks on Aligned Language Models

Reference 40

Resolution: unresolved; no resolver link, observed 2026-08-15T19:23:24.055925Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation b0acf4b4-74df-497a-a4ed-f0b9376cf8f0 · outbound

This paper cites Autodan: Generating stealthy jailbreak prompts on aligned large language models.

Beyond Handcrafted Security: Towards Self-Evolving Defense for LLM Agents Autodan: Generating stealthy jailbreak prompts on aligned large language models

Reference 41

Resolution: verified fuzzy; raw_fallback, observed 2026-08-15T19:23:26.820947Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-16T06:30:59.297886+00:00.

Observation d23c3a7b-dd63-4e58-a3be-2d717b5776aa · outbound

This paper cites Jailbreaking Black Box Large Language Models in Twenty Queries.

Beyond Handcrafted Security: Towards Self-Evolving Defense for LLM Agents Jailbreaking Black Box Large Language Models in Twenty Queries

Reference 42

Resolution: unresolved; no resolver link, observed 2026-08-15T19:23:24.068214Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 7ac0eb4d-3b2d-4184-8f28-c24e802968a7 · outbound

This paper cites Tree of attacks: Jail- breaking black-box llms automatically.

Beyond Handcrafted Security: Towards Self-Evolving Defense for LLM Agents Tree of attacks: Jail- breaking black-box llms automatically

Reference 43

Resolution: verified fuzzy; raw_fallback, observed 2026-08-15T19:23:26.775175Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-16T06:30:59.297886+00:00.

Observation b5381621-f903-4e81-813b-84d462c31096 · outbound

This paper cites Not what you’ve signed up for: Compromising real-world llm-integrated applications with indirect prompt injection.

Beyond Handcrafted Security: Towards Self-Evolving Defense for LLM Agents Not what you’ve signed up for: Compromising real-world llm-integrated applications with indirect prompt injection

Reference 44

Resolution: verified fuzzy; raw_fallback, observed 2026-08-15T19:23:26.692984Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-16T06:30:59.297886+00:00.

Observation e083b540-c6d9-4724-a04c-eb2e0008a4a1 · outbound

This paper cites InjecAgent: Benchmarking indirect prompt injections in tool-integrated large language model agents.

Beyond Handcrafted Security: Towards Self-Evolving Defense for LLM Agents InjecAgent: Benchmarking indirect prompt injections in tool-integrated large language model agents

Reference 45

Resolution: verified fuzzy; raw_fallback, observed 2026-08-15T19:23:26.641449Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-16T06:30:59.297886+00:00.

Observation 71143061-6018-4965-b9d4-36e9851602dc · outbound

This paper cites Formalizing and bench- marking prompt injection attacks and defenses.

Beyond Handcrafted Security: Towards Self-Evolving Defense for LLM Agents Formalizing and bench- marking prompt injection attacks and defenses

Reference 46

Resolution: verified fuzzy; raw_fallback, observed 2026-08-15T19:23:26.565892Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-16T06:30:59.297886+00:00.

Observation 3c4625b0-ef24-4a83-aa75-9f0957dfb8b8 · outbound

This paper cites Neural Exec: Learning (and Learning from) Execution Triggers for Prompt Injection Attacks.

Beyond Handcrafted Security: Towards Self-Evolving Defense for LLM Agents Neural Exec: Learning (and Learning from) Execution Triggers for Prompt Injection Attacks

Reference 47

Resolution: unresolved; no resolver link, observed 2026-08-15T19:23:24.310635Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 8ea2e180-2bd1-4c65-a211-5608300d5860 · outbound

This paper cites AgentPoison: Red-teaming LLM agents via poisoning memory or knowledge bases.

Beyond Handcrafted Security: Towards Self-Evolving Defense for LLM Agents AgentPoison: Red-teaming LLM agents via poisoning memory or knowledge bases

Reference 48

Resolution: verified fuzzy; raw_fallback, observed 2026-08-15T19:23:26.547628Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-16T06:30:59.297886+00:00.

Observation e6e8271c-2077-4d4f-a612-66afc7ac2f0a · outbound

This paper cites PoisonedRAG: Knowledge corruption attacks to retrieval-augmented generation of large language models.

Beyond Handcrafted Security: Towards Self-Evolving Defense for LLM Agents PoisonedRAG: Knowledge corruption attacks to retrieval-augmented generation of large language models

Reference 49

Resolution: verified fuzzy; raw_fallback, observed 2026-08-15T19:23:26.529674Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-16T06:30:59.297886+00:00.

Observation fab29f3c-e092-41bd-b1e1-f56cb8276ff3 · outbound

This paper cites Memory injection attacks on LLM agents via query-only interaction.

Beyond Handcrafted Security: Towards Self-Evolving Defense for LLM Agents Memory injection attacks on LLM agents via query-only interaction

Reference 50

Resolution: unresolved; no resolver link, observed 2026-08-15T19:23:24.327323Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 3130b42d-3d50-4847-af82-fc87021e75b9 · outbound

This paper cites MCPTox: A Benchmark for Tool Poisoning Attack on Real-World MCP Servers.

Beyond Handcrafted Security: Towards Self-Evolving Defense for LLM Agents MCPTox: A Benchmark for Tool Poisoning Attack on Real-World MCP Servers

Reference 51

Resolution: unresolved; no resolver link, observed 2026-08-15T19:23:24.332483Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation ea463893-844b-43df-a3a2-486053c012c6 · outbound

This paper cites Breaking the protocol: Security analysis of the model context protocol specification and prompt injection vul- nerabilities in tool-integrated llm agents.

Beyond Handcrafted Security: Towards Self-Evolving Defense for LLM Agents Breaking the protocol: Security analysis of the model context protocol specification and prompt injection vul- nerabilities in tool-integrated llm agents

Reference 52

Resolution: unresolved; no resolver link, observed 2026-08-15T19:23:24.337604Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 4d13b556-52b2-4d0f-9c30-82521fb931bb · outbound

This paper cites BadSkill: Backdoor Attacks on Agent Skills via Model-in-Skill Poisoning.

Beyond Handcrafted Security: Towards Self-Evolving Defense for LLM Agents BadSkill: Backdoor Attacks on Agent Skills via Model-in-Skill Poisoning

Reference 53

Resolution: unresolved; no resolver link, observed 2026-08-15T19:23:24.342855Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 17b8c90f-f5bf-46aa-ac25-f282a99c8faa · outbound

This paper cites Supply-Chain Poisoning Attacks Against LLM Coding Agent Skill Ecosystems.

Beyond Handcrafted Security: Towards Self-Evolving Defense for LLM Agents Supply-Chain Poisoning Attacks Against LLM Coding Agent Skill Ecosystems

Reference 54

Resolution: unresolved; no resolver link, observed 2026-08-15T19:23:24.395556Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation ac56b610-7aca-462b-bc60-0ec5ed13dcaa · outbound

This paper cites Jailbreaking leading safety- aligned LLMs with simple adaptive attacks.

Beyond Handcrafted Security: Towards Self-Evolving Defense for LLM Agents Jailbreaking leading safety- aligned LLMs with simple adaptive attacks

Reference 55

Resolution: verified fuzzy; raw_fallback, observed 2026-08-15T19:23:26.363117Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-16T06:30:59.297886+00:00.

Observation 71a3a7df-5262-40c6-bc3c-fdda002bdbb8 · outbound

This paper cites AdvAgent: Controllable blackbox red- teaming on web agents.

Beyond Handcrafted Security: Towards Self-Evolving Defense for LLM Agents AdvAgent: Controllable blackbox red- teaming on web agents

Reference 56

Resolution: verified fuzzy; raw_fallback, observed 2026-08-15T19:23:26.343515Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-16T06:30:59.297886+00:00.

Observation dfbd1a0e-fded-4518-87fb-51dd9fde217c · outbound

This paper cites Great, now write an article about that: The crescendo multi-turn LLM jailbreak attack.

Beyond Handcrafted Security: Towards Self-Evolving Defense for LLM Agents Great, now write an article about that: The crescendo multi-turn LLM jailbreak attack

Reference 57

Resolution: verified fuzzy; raw_fallback, observed 2026-08-15T19:23:26.285367Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-16T06:30:59.297886+00:00.

Observation d9806c4e-b8d1-4429-9e52-7a705964ac0e · outbound

This paper cites Foot-In-The-Door: A Multi-turn Jailbreak for LLMs.

Beyond Handcrafted Security: Towards Self-Evolving Defense for LLM Agents Foot-In-The-Door: A Multi-turn Jailbreak for LLMs

Reference 58

Resolution: unresolved; no resolver link, observed 2026-08-15T19:23:24.461954Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation d29c75d6-6fa8-4160-8d41-99c05a59beb6 · outbound

This paper cites Rainbow Teaming: Open-Ended Generation of Diverse Adversarial Prompts.

Beyond Handcrafted Security: Towards Self-Evolving Defense for LLM Agents Rainbow Teaming: Open-Ended Generation of Diverse Adversarial Prompts

Reference 59

Resolution: unresolved; no resolver link, observed 2026-08-15T19:23:24.466327Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation d9328680-fc74-4403-90e4-a2f4825009af · outbound

This paper cites AutoRedTeamer: Autonomous Red Teaming with Lifelong Attack Integration.

Beyond Handcrafted Security: Towards Self-Evolving Defense for LLM Agents AutoRedTeamer: Autonomous Red Teaming with Lifelong Attack Integration

Reference 60

Resolution: unresolved; no resolver link, observed 2026-08-15T19:23:24.471220Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 7d3c065b-9347-474a-8a72-e9211a621ee2 · outbound

This paper cites Attention Tracker: Detecting Prompt Injection Attacks in LLMs.

Beyond Handcrafted Security: Towards Self-Evolving Defense for LLM Agents Attention Tracker: Detecting Prompt Injection Attacks in LLMs

Reference 61

Resolution: unresolved; no resolver link, observed 2026-08-15T19:23:24.475884Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 7fe0d7ea-de04-4fc9-a3b0-3ca6239c32fd · outbound

This paper cites MELON: Provable defense against indirect prompt injection attacks in ai agents.

Beyond Handcrafted Security: Towards Self-Evolving Defense for LLM Agents MELON: Provable defense against indirect prompt injection attacks in ai agents

Reference 62

Resolution: verified fuzzy; raw_fallback, observed 2026-08-15T19:23:26.198415Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-16T06:30:59.297886+00:00.

Observation 0c40dc41-5df5-4ade-918d-97f270bf2415 · outbound

This paper cites LlamaFirewall: An open source guardrail system for building secure AI agents.

Beyond Handcrafted Security: Towards Self-Evolving Defense for LLM Agents LlamaFirewall: An open source guardrail system for building secure AI agents

Reference 63

Resolution: unresolved; no resolver link, observed 2026-08-15T19:23:24.603209Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 087971ee-205c-4594-9a1b-8f777b528b8f · outbound

This paper cites System-Level Defense against Indirect Prompt Injection Attacks: An Information Flow Control Perspective.

Beyond Handcrafted Security: Towards Self-Evolving Defense for LLM Agents System-Level Defense against Indirect Prompt Injection Attacks: An Information Flow Control Perspective

Reference 64

Resolution: unresolved; no resolver link, observed 2026-08-15T19:23:24.635486Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 8dc96646-8955-497a-9419-d378b775c8fb · outbound

This paper cites Defeating Prompt Injections by Design.

Beyond Handcrafted Security: Towards Self-Evolving Defense for LLM Agents Defeating Prompt Injections by Design

Reference 65

Resolution: unresolved; no resolver link, observed 2026-08-15T19:23:24.640513Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation b86f02fe-c06f-4ef0-9bae-b8b4f25c9135 · outbound

This paper cites Poskitt, and Jun Sun.

Beyond Handcrafted Security: Towards Self-Evolving Defense for LLM Agents Poskitt, and Jun Sun

Reference 66

Resolution: verified fuzzy; raw_fallback, observed 2026-08-15T19:23:26.181036Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-16T06:30:59.297886+00:00.

Observation 96d5e586-6c1f-407b-870b-c50c4ce17869 · outbound

This paper cites Firewalls to secure dynamic LLM agentic net- works.Transactions on Machine Learning Re- search, 2026.

Beyond Handcrafted Security: Towards Self-Evolving Defense for LLM Agents Firewalls to secure dynamic LLM agentic net- works.Transactions on Machine Learning Re- search, 2026

Reference 67

Resolution: verified fuzzy; raw_fallback, observed 2026-08-15T19:23:26.161165Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-16T06:30:59.297886+00:00.

Observation 55805c6b-39f5-4ce9-a281-cd22c1c61c3b · outbound

This paper cites GEPA: Reflective Prompt Evolution Can Outperform Reinforcement Learning.

Beyond Handcrafted Security: Towards Self-Evolving Defense for LLM Agents GEPA: Reflective Prompt Evolution Can Outperform Reinforcement Learning

Reference 68

Resolution: unresolved; no resolver link, observed 2026-08-15T19:23:24.657083Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation f72103cf-2a16-486f-8ba3-45c191f327d5 · outbound

This paper cites Automated Design of Agentic Systems.

Beyond Handcrafted Security: Towards Self-Evolving Defense for LLM Agents Automated Design of Agentic Systems

Reference 69

Resolution: unresolved; no resolver link, observed 2026-08-15T19:23:24.663813Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation c78c964a-4af1-4fa5-9ddb-708aa2d9ade7 · outbound

This paper cites Darwin Godel Machine: Open-Ended Evolution of Self-Improving Agents.

Beyond Handcrafted Security: Towards Self-Evolving Defense for LLM Agents Darwin Godel Machine: Open-Ended Evolution of Self-Improving Agents

Reference 70

Resolution: unresolved; no resolver link, observed 2026-08-15T19:23:24.669078Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 58825cca-05e2-42e7-93b9-fe4c973ab8f8 · outbound

This paper cites Adapting the Interface, Not the Model: Runtime Harness Adaptation for Deterministic LLM Agents.

Beyond Handcrafted Security: Towards Self-Evolving Defense for LLM Agents Adapting the Interface, Not the Model: Runtime Harness Adaptation for Deterministic LLM Agents

Reference 71

Resolution: unresolved; no resolver link, observed 2026-08-15T19:23:24.742465Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation ca211a88-9649-43b9-ae83-e6ad7a7e2f25 · outbound

This paper cites Meta-Harness: End-to-End Optimization of Model Harnesses.

Beyond Handcrafted Security: Towards Self-Evolving Defense for LLM Agents Meta-Harness: End-to-End Optimization of Model Harnesses

Reference 72

Resolution: unresolved; no resolver link, observed 2026-08-15T19:23:24.799776Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation d8286467-81d4-4a9a-9fc3-74e5329b345a · outbound

This paper cites Voyager: An Open-Ended Embodied Agent with Large Language Models.

Beyond Handcrafted Security: Towards Self-Evolving Defense for LLM Agents Voyager: An Open-Ended Embodied Agent with Large Language Models

Reference 73

Resolution: unresolved; no resolver link, observed 2026-08-15T19:23:24.818301Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation 5d259132-fe80-4853-a04c-eeea50ad899f · outbound

This paper cites SkillOpt: Executive Strategy for Self-Evolving Agent Skills.

Beyond Handcrafted Security: Towards Self-Evolving Defense for LLM Agents SkillOpt: Executive Strategy for Self-Evolving Agent Skills

Reference 74

Resolution: unresolved; no resolver link, observed 2026-08-15T19:23:24.823874Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation fe23aa23-019a-466e-a666-4720aaf5f33a · outbound

This paper cites Membrane: A Self-Evolving Contrastive Safety Memory for LLM Agent Defense.

Beyond Handcrafted Security: Towards Self-Evolving Defense for LLM Agents Membrane: A Self-Evolving Contrastive Safety Memory for LLM Agent Defense

Reference 75

Resolution: unresolved; no resolver link, observed 2026-08-15T19:23:24.829197Z

Source-reported events for the cited work

Unavailable: canonical work link unavailable.

Observation bb1eaa3d-9d47-4ded-8a55-8a4e758fdbaf · outbound

This paper cites an unresolved cited work.

Beyond Handcrafted Security: Towards Self-Evolving Defense for LLM Agents Unresolved cited work

Reference 2026

Resolution: parse uncertain; raw_fallback, observed 2026-08-15T19:23:27.181684Z

Source-reported events for the cited work

No event found in the named queried sources as of 2026-08-16T06:30:59.297886+00:00.

Coverage vector

Reference resolution

Outbound references

Pith citing papers