Search papers, labs, and topics across Lattice.
This paper introduces SkillJack, a novel attack that exploits the experience-to-skill transformation process in self-evolving agents to implant persistent malicious behaviors. Unlike traditional memory poisoning techniques, SkillJack allows adversaries to embed harmful skills into an agent's repertoire, effectively bypassing detection mechanisms and maintaining functionality even after the original poisoned experiences are removed. Evaluations on SkillX and Anything2Skill demonstrate that skill extraction significantly reduces attack detectability, with success rates of 56.2% and 89.2%, respectively, highlighting a critical vulnerability in the skill evolution process of autonomous agents.
SkillJack reveals that self-evolving agents can unknowingly incorporate malicious skills, making traditional safety measures ineffective against persistent threats.
Self-evolving agents increasingly convert interaction histories into reusable skills that persist beyond individual tasks. While prior work studies memory and retrieval poisoning, such attacks only affect agents when poisoned records are retrieved as context. We uncover a new and more fundamental risk: poisoned experiences can be transformed by the agent itself into durable behavioral artifacts. We present \textbf{SkillJack}, the first attack that exploits the experience-to-skill pipeline of self-evolving agents. Instead of directly manipulating runtime context, SkillJack hijacks the agent's own learning process to implant malicious behaviors into its reusable skill repertoire. We identify three key properties of this transformation: \emph{sanitization whitewashing}, where malicious intent is obscured during skill extraction; \emph{cross-layer promotion}, where transient experiences become persistent capabilities; and \emph{persistence isolation}, where the attack survives removal of its original source records. We evaluate SkillJack on two representative systems, SkillX and Anything2Skill, using a shared dataset of 150 trajectories across four policy-risk categories. Results show that skill extraction substantially reduces attack detectability: in SkillX, safety detection drops from 98.5\% for poisoned trajectories to 11.4\% for extracted skills, while Anything2Skill shows a similar effect. Meanwhile, the implanted skills remain effective, achieving attack success rates of 56.2\% and 89.2\% on the two systems, respectively. Furthermore, 80.0\% of skill-mediated attacks persist after deleting the original poisoned records, and some skills unintentionally activate on benign queries. Our findings reveal skill evolution as a new attack surface and motivate provenance-aware skill lifecycle protection. Our code is available at https://github.com/Tencent/AI-Infra-Guard/research/skilljack.
