A critical attack vector dubbed “Agentjacking” that turns popular AI coding assistants into execution engines for malicious code. Tenet Security’s Threat Labs discovered this vulnerability, revealing the implicit trust AI coding agents, such as Claude Code and Cursor, place in external tool outputs. By exploiting the intersection of Sentry’s event ingestion and the Model Context […]
The post New Agentjacking Attack Compromises AI Coding Agents for Code Execution appeared first on Cyber Security News.