offensive-fuzzing

SUSPICIOUS: the skill is internally coherent as an offensive fuzzing guide, but its purpose is to equip an AI agent with offensive security techniques against arbitrary targets, including kernels and EDR components. There is little evidence of malware or credential theft, yet the capability itself is high risk for AI-agent deployment.