hunt-xxe

SUSPICIOUS: the skill is internally consistent as an XXE hunting guide, but its actual footprint is an offensive security playbook for an AI agent, including target probing, file-read payloads, OOB callback validation, and SSRF techniques. It is not malware and does not show stealthy exfiltration from the user's machine, but it materially increases the agent's ability to conduct exploitation against external systems, so overall security risk is high.