IPI-exposure-signal

IPI Exposure Signal This is the code repository for our paper: Your Agentic LLMs Secretly Encode Latent Signals of Indirect Prompt-Injection Exposure. ArXiv version and paper link: https://arxiv.org/abs/2608.02657 This repository implements the probing pipeline for latent IPI-exposure signals:...

ActGuard

ActGuard ActGuard is a pre-execution action-auditing defense against indirect prompt injection in tool-using LLM agents. This repository contains the final ActGuard implementation and the AgentDojo-based runtime needed to evaluate it. The release contains only the final, full ActGuard...

benign-instruction-bench

Passing the Test You Trained On Re-evaluating prompt-injection detectors where LLM agents actually use them: on the tool outputs an agent reads. Teams pick injection detectors by their benchmark scores. We check whether those scores predict behavior inside an agent, and find that they mostly...

Ignored, Not Refused: What Low Prompt-Injection Success Rates Hide in Small Tool-Using Agents

Tool-using agents read untrusted external content, which exposes them to indirect prompt injection IPI, where hidden instructions hijack the agent's tool calls. Benchmarks measure IPI security by the attack success rate ASR, which treats every failed attack as a successful defense. To test this...

Correct Answers, Unsupported Findings: Evidence Binding in Forensic Reconstruction of LLM Agent Logs

Forensic reconstruction of LLM-agent actions requires not only recovering the correct value, but establishing which preserved record supports that finding. Tool logs, generated explanations, and local citation identifiers capture different parts of this evidence, yet a citation identifier does no...

The Same Zero: Why Identical ASR Can Imply Different Guarantees in LLM-Agent Security

LLM-agent security has produced a dense landscape of defenses - prompt hardening, content filters, permission gates, sandboxes - yet no framework tells a deployer what a defense actually guarantees, or where that guarantee comes from. We apply Verification Autonomy Levels VAL - L0: LLM...

Ajar: Measuring Open Privilege in Agent Defenses

A language model agent acts through the tools it is given. The data it reads while working on a task can redirect what it does with those tools. A growing set of techniques for safe and secure agent execution therefore sits between the agent and its tools, aiming to enforce access control,...

ToolHazard: Scaling Adversarial Environments for Security Evaluation and Alignment of LLM-Based Agents

Large language model LLM agents integrated with external tools are vulnerable to indirect prompt injections embedded in environmental states. However, existing studies largely rely on manually implemented or reused environments, stochastic LLM-based tool simulation, and predefined injection...

Optimizing Agent Planning for Security and Autonomy

Indirect prompt injection attacks threaten AI agents that execute consequential actions, motivating deterministic system-level defenses. Such defenses can provably block unsafe actions by enforcing confidentiality and integrity policies, but currently appear costly: they reduce task completion...

Progent: Programmable Privilege Control for LLM Agents

LLM agents are an emerging form of AI systems where large language models LLMs serve as the central component, utilizing a diverse set of tools to complete user-assigned tasks. Despite their great potential, LLM agents pose significant security risks. When interacting with the external world, the...