T-MAP: Red-Teaming LLM Agents with Trajectory-aware Evolutionary Search

• URL: http://arxiv.org/abs/2603.22341v1
• Date: Sat, 21 Mar 2026 12:33:34 GMT
• Title: T-MAP: Red-Teaming LLM Agents with Trajectory-aware Evolutionary Search
• Abstract summary: We propose a trajectory-aware evolutionary search method, T-MAP, which leverages execution trajectories to guide the discovery of adversarial prompts.<n>Our approach enables the automatic generation of attacks that not only bypass safety guardrails but also reliably realize harmful objectives through actual tool interactions.
• Score: 49.99459363244884
• License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/
• Abstract: While prior red-teaming efforts have focused on eliciting harmful text outputs from large language models (LLMs), such approaches fail to capture agent-specific vulnerabilities that emerge through multi-step tool execution, particularly in rapidly growing ecosystems such as the Model Context Protocol (MCP). To address this gap, we propose a trajectory-aware evolutionary search method, T-MAP, which leverages execution trajectories to guide the discovery of adversarial prompts. Our approach enables the automatic generation of attacks that not only bypass safety guardrails but also reliably realize harmful objectives through actual tool interactions. Empirical evaluations across diverse MCP environments demonstrate that T-MAP substantially outperforms baselines in attack realization rate (ARR) and remains effective against frontier models, including GPT-5.2, Gemini-3-Pro, Qwen3.5, and GLM-5, thereby revealing previously underexplored vulnerabilities in autonomous LLM agents.

Related papers

• Hybrid Analysis for Secure MCP Tool Use in LLM Agents [77.68135853475535]
  We propose a hybrid analysis-based defense framework designed to safeguard the use of MCP tools in large language model (LLM) agents.<n> MTGuard effectively mitigates multiple categories of harmful tool use across different LLM agents while maintaining performance on benign user tasks.
  arXiv  Detail & Related papers  (2026-07-28T05:17:34Z)
• PI-Hunter: Automated Red-Teaming for Exposing and Localizing Prompt Injections [51.8246149916068]
  Large Language Models (LLMs) are rapidly evolving into agentic systems that interact with external tools and environments.<n>We propose PI-Hunter, an automated agentic auditing framework for proactive vulnerability exposure in LLM agents.
  arXiv  Detail & Related papers  (2026-06-10T22:57:02Z)
• Assessing Automated Prompt Injection Attacks in Agentic Environments [36.488797554085345]
  We present a comprehensive empirical evaluation of automated prompt injection attacks against LLM agents.<n>We adapt both white-box (GCG) and black-box (TAP) methods to the agentic setting within the AgentDojo framework.
  arXiv  Detail & Related papers  (2026-06-09T07:54:58Z)
• When the Manual Lies: A Realistic Benchmark to Evaluate MCP Poisoning Attacks for LLM Agents [43.702131498259384]
  This paper systematically investigates Tool Description Poisoning (TDP), a novel semantic attack.<n>In TDP, malicious instructions are not embedded in a tool's executable code, but rather covertly injected into its descriptive metadata.<n>This work provides the first specialized security benchmark tailored for TDP, offering essential insights for securing the cognitive and planning layers of advanced agentic systems.
  arXiv  Detail & Related papers  (2026-05-22T08:34:48Z)
• Invisible Threats from Model Context Protocol: Generating Stealthy Injection Payload via Tree-based Adaptive Search [19.570062155058135]
  Recent advances in the Model Context Protocol (MCP) have enabled large language models (LLMs) to invoke external tools with unprecedented ease.<n>This creates a new class of powerful and tool augmented agents.<n>Existing techniques for indirect prompt injection that target MCP suffer from high deployment costs, weak semantic coherence, or heavy white box requirements.<n>We propose Tree structured Injection for Payloads (TIP), a novel black-box attack which generates natural payloads to reliably seize control of MCP enabled agents even under defense.
  arXiv  Detail & Related papers  (2026-03-25T11:24:47Z)
• Sponge Tool Attack: Stealthy Denial-of-Efficiency against Tool-Augmented Agentic Reasoning [58.432996881401415]
  Recent work augments large language models (LLMs) with external tools to enable agentic reasoning.<n>We propose Sponge Tool Attack (STA), which disrupts agentic reasoning solely by rewriting the input prompt.<n>STA generates benign-looking prompt rewrites from the original one with high semantic fidelity.
  arXiv  Detail & Related papers  (2026-01-24T19:36:51Z)
• OpenRT: An Open-Source Red Teaming Framework for Multimodal LLMs [36.57820295876294]
  We introduce OpenRT, a unified, modular, and high- throughput red-teaming framework for MLLM safety evaluation.<n>At its core, OpenRT architects a paradigm shift in automated red-teaming by introducing an adversarial kernel that enables modular separation across five dimensions.<n>Our framework integrates 37 diverse attack methodologies, spanning white-box gradients, multi-modal perturbations, and sophisticated multi-agent evolutionary strategies.
  arXiv  Detail & Related papers  (2026-01-04T16:41:33Z)
• Securing the Model Context Protocol: Defending LLMs Against Tool Poisoning and Adversarial Attacks [8.419049623790618]
  This work analyzes three classes of semantic attacks on MCP-integrated systems.<n>We introduce a layered security framework with three components: RSA-based manifest signing to enforce descriptor integrity, LLM-on-LLM semantic vetting to detect suspicious tool definitions, and lightweight guardrails that block anomalous tool behavior at runtime.<n>Our results show that the proposed framework reduces unsafe tool invocation rates without model fine-tuning or internal modification.
  arXiv  Detail & Related papers  (2025-12-06T20:07:58Z)
• Exploiting Web Search Tools of AI Agents for Data Exfiltration [0.46664938579243564]
  Large language models (LLMs) are now routinely used to execute complex tasks, from natural language processing to dynamic like web searches.<n>The usage of tool-calling and Retrieval Augmented Generation (RAG) allows LLMs to process and retrieve sensitive corporate data, amplifying both their functionality and vulnerability to abuse.<n>We analyze how susceptible current LLMs are to indirect prompt injection attacks, which parameters, including model size and manufacturer, shape their vulnerability, and which attack methods remain most effective.
  arXiv  Detail & Related papers  (2025-10-10T07:39:01Z)
• Alignment Tipping Process: How Self-Evolution Pushes LLM Agents Off the Rails [103.05296856071931]
  We identify the Alignment Tipping Process (ATP), a critical post-deployment risk unique to self-evolving Large Language Model (LLM) agents.<n>ATP arises when continual interaction drives agents to abandon alignment constraints established during training in favor of reinforced, self-interested strategies.<n>Our experiments show that alignment benefits erode rapidly under self-evolution, with initially aligned models converging toward unaligned states.
  arXiv  Detail & Related papers  (2025-10-06T14:48:39Z)
• MCP-Guard: A Defense Framework for Model Context Protocol Integrity in Large Language Model Applications [21.70488724213541]
  integration of Large Language Models with external tools introduces critical security vulnerabilities.<n>We propose MCP-Guard, a robust, layered defense architecture designed for LLM--tool interactions.<n>We also introduce MCP-AttackBench, a benchmark of over 70,000 samples.
  arXiv  Detail & Related papers  (2025-08-14T18:00:25Z)
• Detection Method for Prompt Injection by Integrating Pre-trained Model and Heuristic Feature Engineering [3.0823377252469144]
  prompt injection attacks have emerged as a significant security threat.<n>Existing defense mechanisms face trade-offs between effectiveness and generalizability.<n>We propose a dual-channel feature fusion detection framework.
  arXiv  Detail & Related papers  (2025-06-05T06:01:19Z)
• AgentVigil: Generic Black-Box Red-teaming for Indirect Prompt Injection against LLM Agents [54.29555239363013]
  We propose a generic black-box fuzzing framework, AgentVigil, to automatically discover and exploit indirect prompt injection vulnerabilities.<n>We evaluate AgentVigil on two public benchmarks, AgentDojo and VWA-adv, where it achieves 71% and 70% success rates against agents based on o3-mini and GPT-4o.<n>We apply our attacks in real-world environments, successfully misleading agents to navigate to arbitrary URLs, including malicious sites.
  arXiv  Detail & Related papers  (2025-05-09T07:40:17Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.