Researchers reportedly found the first vulnerability in Claude Code shortly after an accidental source leak. That makes this a timely AI security story, especially as coding assistants become more integrated into developer workflows and enterprise environments.