LatentTriad: Pre-Action Risk Representations for Malicious Agent Skill Detection
cs.AI
Submitted: 2026-11-03
Updated: 2026-11-03
Code: https://github.com/cisco-ai-defense/skill-scanner
Terminology
Sources
- Refusal in Language Models Is Mediated by a Single Direction
- Formal Analysis and Supply Chain Security for Agentic AI Skills
- Mind the GAP: Text Safety Does Not Transfer to Tool-Call Safety in LLM Agents
- Defeating Prompt Injections by Design
- MalSkillBench: A Runtime-Verified Benchmark of Malicious Agent Skills
- One Skill, Two Masters: Serving Users and Attackers through Skill Injection
- SkillSieve: A Hierarchical Triage Framework for Detecting Malicious AI Agent Skills
- Cloak and Detonate: Scanner Evasion and Dynamic Detection of Agent Skill Malware
- ShareLock: A Stealthy Multi-Tool Threshold Poisoning Attack Against MCP
- Agent MechSuits: Mechanistic Subspace Safety Steering for Multi-Turn CLI Agents
- PromptArmor: Simple yet Effective Prompt Injection Defenses
- When Safe Skills Collide: Measuring Compositional Risk in Agent Skill Ecosystems
- RouteGuard: Internal-Signal Detection of Skill Poisoning in LLM Agents
- Benign in Isolation, Harmful in Composition: Security Risks in Agent Skill Ecosystems
- Attention is All You Need to Defend Against Indirect Prompt Injection Attacks in LLMs
- Improving Alignment and Robustness with Circuit Breakers
Related papers
- Multi-Objective Structured Pruning of LLMs for Latency and Model Size Optimization
- MAVEN-T: Reinforced Heterogeneous Distillation for Real-Time Multi-Agent Trajectory Prediction
- Model Discovery Agent: LLM-assisted Bayesian experiment design for data-efficient discovery of mechanistic world models
- Embeddings for Preferences, Not Semantics
- Can We Trust Item Response Theory for AI Evaluation?
- Diagnosing Task Insensitivity in Language Agents