Kairos: Transactional Security Enforcement at the System Boundary for LLM Agents
cs.CR, cs.LG
Submitted: 2026-10-13
Updated: 2026-10-13
Code: https://github.com/UzL-ITS/Kairos
License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/
Sources
- Agentic Coding in the Wild: Characterizing GitHub Copilot Traces at Production Scale
- Can LLMs Separate Instructions From Data? And What Do We Even Mean By That?
- The Promptware Kill Chain: How Prompt Injections Gradually Evolved Into a Multistep Malware Delivery Mechanism
- Invitation Is All You Need! Promptware Attacks Against LLM-Powered Assistants in Production Are Practical and Dangerous
- LLMail-Inject: A Dataset from a Realistic Adaptive Prompt Injection Challenge
- The Instruction Hierarchy: Training LLMs to Prioritize Privileged Instructions
- Defeating Prompt Injections by Design
- Securing AI Agents with Information-Flow Control
- Progent: Securing AI Agents with Privilege Control
- LlamaFirewall: An open source guardrail system for building secure AI agents
- Agent Security is a Systems Problem
- ActPlane: Programmable OS-Level Policy Enforcement for Agent Harnesses
- AgentDojo: A Dynamic Environment to Evaluate Prompt Injection Attacks and Defenses for LLM Agents
- AgentDyn: Are Your Agent Security Defenses Deployable in Real-World Dynamic Environments?
- AutoDojo: A Generative Benchmark for Evaluating Prompt Injection Defenses in LLM Agents
- The Attacker Moves Second: Stronger Adaptive Attacks Bypass Defenses Against Llm Jailbreaks and Prompt Injections
- Preventing Prompt Injection with Type-Directed Privilege Separation
- SAFEFLOW: A Principled Protocol for Trustworthy and Transactional Autonomous Agent Systems
- Formal Policy Enforcement for Agentic Systems
- Efficient and Sound Probabilistic Verification for AI Agents
Related papers
- SoK: AI-Augmented Binary Reversing
- Relaxed Sender Anonymity for CBDC Interbank Settlement: A Zero-Knowledge Approach on Permissioned EVM
- Calibration-Family Overfit: Why Trusted Sabotage Monitors Don't Transfer Across Lineages
- Efficient Fuzzy PSI under One-Sided Assumptions
- Sealing the Audit-Runtime Gap for LLM Skills
- Token Composition: A Graph Based on EVM Logs