The Hacker News

Agentjacking Attack Tricks AI Coding Agents Into Running Malicious Code

Researchers warn Agentjacking can abuse Sentry errors to make AI coding agents run malicious code on developer machines.