"""
SkillJack: Memory-to-Skill Poisoning Attack Framework

A research framework for studying security vulnerabilities in the
memory-to-skill automatic conversion paradigm used by LLM agent systems.

When agent systems automatically convert accumulated experience (trajectories,
memory) into reusable, persistent skills via LLM-driven pipelines, attackers
can poison the memory to automatically generate persistent, cross-task
propagating malicious skills.

Paper: "SkillJack: Exploiting the Memory-to-Skill Conversion Paradigm
        in LLM Agent Systems"
"""

__version__ = "1.0.0"
__author__ = "SkillJack Authors"