firmware-analyst

SUSPICIOUS: The skill is internally consistent and does not show credential theft, exfiltration, or installer abuse, but it equips an AI agent with offensive-adjacent reverse-engineering capability. Risk is driven by the nature of the capability, not by malicious data flows or supply-chain behavior in the skill itself.

Purpose and capabilities are internally consistent for a firmware security research skill, with no clear credential harvesting or exfiltration path. However, it equips an AI agent with meaningful offensive security and exploit-analysis workflows, so it is not malicious but is high-risk as an exploit-oriented skill.