mlsanityData to 7 Oct 2026

EnIGMA: Interactive Tools Substantially Assist LM Agents in Finding Security Vulnerabilities

arXiv 2409.16165 ↗First posted 24 Sep 2024cs.AIcited by 51

On the map: LLM agents and tool use → Agent threats, guardrails and governance

Abstract (arXiv)

Although language model (LM) agents have demonstrated increased performance in multiple domains, including coding and web-browsing, their success in cybersecurity has been limited. We present EnIGMA, an LM agent for autonomously solving Capture The Flag (CTF) challenges. We introduce new tools and interfaces to improve the agent's ability to find and exploit security vulnerabilities, focusing on interactive terminal programs. These novel Interactive Agent Tools enable LM agents, for the first time, to run interactive utilities, such as a debugger and a server connection tool, which are essential for solving these challenges. Empirical analysis on 390 CTF challenges across four benchmarks demonstrate that these new tools and interfaces substantially improve our agent's performance, achieving state-of-the-art results on NYU CTF, Intercode-CTF, and CyBench. Finally, we analyze data leakage, developing new methods to quantify it and identifying a new phenomenon we term soliloquizing, where the model self-generates hallucinated observations without interacting with the environment. Our code and development dataset are available at https://github.com/SWE-agent/SWE-agent/tree/v0.7 and https://github.com/NYU-LLM-CTF/NYU_CTF_Bench/tree/main/development respectively.

Key facts. Cited by 51 papers in the map (data through 7 Oct 2026). Placed in Agent threats, guardrails and governance, one of 16 directions in LLM agents and tool use. Builds on 17 of its 75 references that are in the map. First posted on arXiv 24 Sep 2024.

Builds on

17 of 75 references.

References · 17

  1. 2024The Llama 3 Herd of Modelscited by 14,192
  2. 2021Evaluating Large Language Models Trained on Codecited by 8,053
  3. 2024Gemini 1.5: Unlocking multimodal understanding across millions of tokens of contextcited by 3,175
  4. 2022Constitutional AI: Harmlessness from AI Feedbackcited by 2,950
  5. 2023MetaGPT: Meta Programming for A Multi-Agent Collaborative Frameworkcited by 1,761
  6. 2024OpenHands: An Open Platform for AI Software Developers as Generalist Agentscited by 956
  7. 2021CodeXGLUE: A Machine Learning Benchmark Dataset for Code Understanding and Generationcited by 732
  8. 2024Executable Code Actions Elicit Better LLM Agentscited by 537
  9. 2023Cognitive Architectures for Language Agentscited by 422
  10. 2024Large Language Model-Based Agents for Software Engineering: A Surveycited by 173
  11. 2024Cybench: A Framework for Evaluating Cybersecurity Capabilities and Risks of Language Modelscited by 156
  12. 2022GitHub Copilot AI pair programmer: Asset or Liability?cited by 150
5 more
  1. 2023Purple Llama CyberSecEval: A Secure Coding Benchmark for Language Modelscited by 130
  2. 2024Evaluating Frontier Models for Dangerous Capabilitiescited by 121
  3. 2023PentestGPT: An LLM-empowered Automatic Penetration Testing Toolcited by 94
  4. 2024CyberSecEval 2: A Wide-Ranging Cybersecurity Evaluation Suite for Large Language Modelscited by 84
  5. 2024CYBERSECEVAL 3: Advancing the Evaluation of Cybersecurity Risks and Capabilities in Large Language Modelscited by 56

Built on by

Cited by year (2026 to 7 Oct)

Citing papers per year: 2024: 4; 2025: 21; 2026: 26.

51 papers.

Sort:

By area:

  1. 2025Agent Laboratory: Using LLM Agents as Research Assistantscited by 405
  2. 2024SWE-bench Multimodal: Do AI Systems Generalize to Visual Software Domains?cited by 196
  3. 2025CyberGym: Evaluating AI Agents' Real-World Cybersecurity Capabilities at Scalecited by 71
  4. 2025MCP Security Bench (MSB): Benchmarking Attacks Against Model Context Protocol in LLM Agentscited by 37
  5. 2024HackSynth: LLM Agent and Evaluation Framework for Autonomous Penetration Testingcited by 35
  6. 2025Measuring Agents in Productioncited by 35
  7. 2025D-CIPHER: Dynamic Collaborative Intelligent Multi-Agent System with Planner and Heterogeneous Executors for Offensive Securitycited by 19
  8. 2025Estimating Worst-Case Frontier Risks of Open-Weight LLMscited by 17
  9. 2025Training Language Model Agents to Find Vulnerabilities with CTF-Dojocited by 15
  10. 2024Hacking CTFs with Plain Agentscited by 14
  11. 2025CRAKEN: Cybersecurity LLM Agent with Knowledge-Based Executioncited by 14
  12. 2025Comparing AI Agents to Cybersecurity Professionals in Real-World Penetration Testingcited by 14
38 more
  1. 2025FaultLine: Automated Proof-of-Vulnerability Generation Using LLM Agentscited by 11
  2. 2025SecureVibeBench: Benchmarking Secure Vibe Coding of AI Agents via Reconstructing Vulnerability-Introducing Scenarioscited by 11
  3. 2025AutoPentest: Enhancing Vulnerability Management With Autonomous LLM Agentscited by 8
  4. 2025PentestJudge: Judging Agent Behavior Against Operational Requirementscited by 7
  5. 2025PACEbench: A Framework for Evaluating Practical AI Cyber-Exploitation Capabilitiescited by 6
  6. 2025Can Language Models Falsify? Evaluating Algorithmic Reasoning with Counterexample Creationcited by 5
  7. 2025From Trace to Line: LLM Agent for Real-World OSS Vulnerability Localizationcited by 4
  8. 2025AutoBaxBuilder: Bootstrapping Code Security Benchmarkingcited by 4
  9. 2026From SFT to RL: Demystifying the Post-Training Pipeline for LLM-based Vulnerability Detectioncited by 4
  10. 2026OpenSage: Self-programming Agent Generation Enginecited by 4
  11. 2025AgentRec: Agent Recommendation Using Sentence Embeddings Aligned to Human Feedbackcited by 3
  12. 2026CTFusion: A CTF-based Benchmark for LLM Agent Evaluationcited by 3
  13. 2024The Best Defense is a Good Offense: Countering LLM-Powered Cyberattackscited by 2
  14. 2025Running in CIRCLE? A Simple Benchmark for LLM Code Interpreter Securitycited by 2
  15. 2025HackWorld: Evaluating Computer-Use Agents on Exploiting Web Application Vulnerabilitiescited by 2
  16. 2025Empowering Real-World: A Survey on the Technology, Practice, and Evaluation of LLM-driven Industry Agentscited by 2
  17. 2026KryptoPilot: An Open-World Knowledge-Augmented LLM Agent for Automated Cryptographic Exploitationcited by 1
  18. 2026To Defend Against Cyber Attacks, We Must Teach AI Agents to Hackcited by 1
  19. 2026A Content-Based Framework for Cybersecurity Refusal Decisions in Large Language Modelscited by 1
  20. 2026AI In Cybersecurity Education -- Scalable Agentic CTF Design Principles and Educational Outcomescited by 1
  21. 2026Empowering Autonomous Debugging Agents with Efficient Dynamic Analysiscited by 1
  22. 2026Veritas: Grounding LLM Agents for Reliable Vulnerability Reasoning over Stripped Binariescited by 1
  23. 2026Detecting Privilege Escalation in Polyglot Microservices via Agentic Program Analysiscited by 1
  24. 2026Honeyval: A Comprehensive Evaluation Framework for LLM-powered HTTP Honeypotscited by 1
  25. 2026Beyond Success Rate: Cost-Aware Evaluation of Offensive and Defensive Security Agentscited by 1
  26. 2026RECEIPT: Deterministic, Reward-Hacking-Resistant Verification for White-Box Agentic XSS Discoverycited by 1
  27. 2026Trident : How to Break Deep Reinforcement Learning Cyber Defenses (Agentic)cited by 1
  28. 2026Beyond End-to-End Success: Diagnosing Failures in Long-Horizon Security LLM Agentscited by 1
  29. 2026Challenges and Future Directions in Agentic Reverse Engineering Systemscited by 0
  30. 2026Systematic Capability Benchmarking of Frontier Large Language Models for Offensive Cyber Taskscited by 0
  31. 2026Agentic Vulnerability Reasoning on COTS Binariescited by 0
  32. 2026Patch2Vuln: Agentic Reconstruction of Vulnerabilities from Linux Distribution Binary Patchescited by 0
  33. 2026Measuring Safety Alignment Effects in Autonomous Security Agentscited by 0
  34. 2026CyberEvolver: Structured Self-Evolution for Cybersecurity Agents On the Flycited by 0
  35. 2026Determinants and Limits of LLM Security-Tool Orchestration: A Study with HexStrike-AIcited by 0
  36. 2026VEXAIoT: Autonomous IoT Vulnerability EXploitation using AI Agentscited by 0
  37. 2026Reveree: Diagnosing LLM Reverse-Engineering Agentscited by 0
  38. 2026CAIRN: Dynamic Fact-Intent DAGs for Multi-Agent Explorationcited by 0

50 most cited of 51 All 51 (JSON)

Similar, not cited

Close in wording, no citation either way

  1. 2026Autonomous LLM Agents & CTFs: A Second Look
  2. 2026How Do LLM Agents Actually Get the Flag? Trace-Level Provenance for Agentic Offensive Security Evaluation
  3. 2026Capture the Flags: Family-Based Evaluation of Agentic LLMs via Semantics-Preserving Transformations
  4. 2025Towards Effective Offensive Security LLM Agents: Hyperparameter Tuning, LLM as a Judge, and a Lightweight CTF Benchmark
  5. 2026Do Agents Dream of Root Shells? Partial-Credit Evaluation of LLM Agents in Capture the Flag Challenges
  6. 2026Toward Secure LLM Agents: Threat Surfaces, Attacks, Defenses, and Evaluation
  7. 2026CTFExplorer: Evaluating LLM Offensive Agents Through Multi-Target Web CTF Benchmarking
  8. 2026On Understanding, Identifying, and Mitigating Vulnerabilities in Agentic Large Language Models