A developer secretly embedded a prompt injection vulnerability into the jqwik testing framework. This vulnerability targets AI coding agents, instructing them to delete application output, highlighting novel security risks in AI-assisted development workflows.
Opening Kapyn…