SkillSecurer: Detecting and Patching Prompt-Injection Vulnerabilities in AI Agent Skills | AIChainDay