Cover Image

BadSkill: Backdoor Attacks on Agent Skills via Model-in-Skill Poisoning

Fuente: arXiv

Saved in:

Bibliographic Details
Main Authors:	Tie, Guiyao, Shi, Jiawen, Zhou, Pan, Sun, Lichao
Format:	Preprint
Published:	2026
Subjects:	Cryptography and Security Artificial Intelligence
Online Access:	Acceder al recurso
Tags:	Add Tag No Tags, Be the first to tag this record!

Similar Items

BadVLA: Towards Backdoor Attacks on Vision-Language-Action Models via Objective-Decoupled Optimization
by: Zhou, Xueyang, et al.
Published: (2025)

SkillTrojan: Backdoor Attacks on Skill-Based Agent Systems
by: Feng, Yunhao, et al.
Published: (2026)

Poisoned-MRAG: Knowledge Poisoning Attacks to Multimodal Retrieval Augmented Generation
by: Liu, Yinuo, et al.
Published: (2025)

Prompt Injection Attack to Tool Selection in LLM Agents
by: Shi, Jiawen, et al.
Published: (2025)

RouteGuard: Internal-Signal Detection of Skill Poisoning in LLM Agents
by: Xiao, Wenjie, et al.
Published: (2026)

BadToken: Token-level Backdoor Attacks to Multi-modal Large Language Models
by: Yuan, Zenghui, et al.
Published: (2025)

Supply-Chain Poisoning Attacks Against LLM Coding Agent Skill Ecosystems
by: Qu, Yubin, et al.
Published: (2026)

SkillJect: Effectively Automating Skill-Based Prompt Injection for Skill-Enabled Agents
by: Jia, Xiaojun, et al.
Published: (2026)

SkillTester: Benchmarking Utility and Security of Agent Skills
by: Wang, Leye, et al.
Published: (2026)

No Attack Required: Semantic Fuzzing for Specification Violations in Agent Skills
by: Li, Ying, et al.
Published: (2026)

HarmfulSkillBench: How Do Harmful Skills Weaponize Your Agents?
by: Jiang, Yukun, et al.
Published: (2026)

SkillSieve: A Hierarchical Triage Framework for Detecting Malicious AI Agent Skills
by: Hou, Yinghan, et al.
Published: (2026)

Merge Hijacking: Backdoor Attacks to Model Merging of Large Language Models
by: Yuan, Zenghui, et al.
Published: (2025)

Optimization-based Prompt Injection Attack to LLM-as-a-Judge
by: Shi, Jiawen, et al.
Published: (2024)

Proteus: A Self-Evolving Red Team for Agent Skill Ecosystems
by: Zhou, Zhaojiacheng
Published: (2026)

AutoBackdoor: Automating Backdoor Attacks via LLM Agents
by: Li, Yige, et al.
Published: (2025)

Defenses & Enablers For Skill Injection Attacks on Terminal Based Agents
by: Fujinuma, Yoshinari, et al.
Published: (2026)

Under the Hood of SKILL.md: Semantic Supply-chain Attacks on AI Agent Skill Registry
by: Saha, Shoumik, et al.
Published: (2026)

AgentTrap: Measuring Runtime Trust Failures in Third-Party Agent Skills
by: Zhuang, Haomin, et al.
Published: (2026)

When Safe Skills Collide: Measuring Compositional Risk in Agent Skill Ecosystems
by: Wang, Su, et al.
Published: (2026)

BadSNN: Backdoor Attacks on Spiking Neural Networks via Adversarial Spiking Neuron
by: Miah, Abdullah Arafat, et al.
Published: (2026)

CBPF: Filtering Poisoned Data Based on Composite Backdoor Attack
by: Xia, Hanfeng, et al.
Published: (2024)

Technical Report: Exploring the Emerging Threats of the Agent Skill Ecosystem
by: Beurer-Kellner, Luca, et al.
Published: (2026)

Structured Security Auditing and Robustness Enhancement for Untrusted Agent Skills
by: Lv, Lijia, et al.
Published: (2026)

BadAgent: Inserting and Activating Backdoor Attacks in LLM Agents
by: Wang, Yifei, et al.
Published: (2024)

SkillSafetyBench: Evaluating Agent Safety under Skill-Facing Attack Surfaces
by: Jin, Chang, et al.
Published: (2026)

Towards Secure Agent Skills: Architecture, Threat Taxonomy, and Security Analysis
by: Li, Zhiyuan, et al.
Published: (2026)

Context Matters: Repository-Aware Security Analysis of the Agent Skill Ecosystem
by: Holzbauer, Florian, et al.
Published: (2026)

Behavioral Integrity Verification for AI Agent Skills
by: Wu, Yuhao, et al.
Published: (2026)

A Set of Generalized Components to Achieve Effective Poison-only Clean-label Backdoor Attacks with Collaborative Sample Selection and Triggers
by: Wu, Zhixiao, et al.
Published: (2025)

SkCC: Portable and Secure Skill Compilation for Cross-Framework LLM Agents
by: Ouyang, Yipeng, et al.
Published: (2026)

Credential Leakage in LLM Agent Skills: A Large-Scale Empirical Study
by: Chen, Zhihao, et al.
Published: (2026)

Semia: Auditing Agent Skills via Constraint-Guided Representation Synthesis
by: Wen, Hongbo, et al.
Published: (2026)

Toward Polymorphic Backdoor against Semantic Communication via Intensity-Based Poisoning
by: Yang, Xiao, et al.
Published: (2026)

When the Manual Lies: A Realistic Benchmark to Evaluate MCP Poisoning Attacks for LLM Agents
by: Liu, Shi, et al.
Published: (2026)

BELT: Old-School Backdoor Attacks can Evade the State-of-the-Art Defense with Backdoor Exclusivity Lifting
by: Qiu, Huming, et al.
Published: (2023)

ClawKeeper: Comprehensive Safety Protection for OpenClaw Agents Through Skills, Plugins, and Watchers
by: Liu, Songyang, et al.
Published: (2026)

Poison Once, Exploit Forever: Environment-Injected Memory Poisoning Attacks on Web Agents
by: Zou, Wei, et al.
Published: (2026)

BadImplant: Injection-based Multi-Targeted Graph Backdoor Attack
by: Khan, Md Nabi Newaz, et al.
Published: (2026)

LoopTrap: Termination Poisoning Attacks on LLM Agents
by: Xu, Huiyu, et al.
Published: (2026)