cdp

CRITICAL E006: Malicious code pattern detected in skill scripts.

• Malicious code pattern detected (high risk: 1.00). The skill intentionally includes multiple capabilities that can capture credentials, cookies, network requests, keystrokes, auto-attach to user tabs, bypass user prompts, and evaluate arbitrary JS via a local REPL daemon — features that enable credential theft, silent user-activity recording, and remote-control/exfiltration when abused.

MEDIUM W013: Attempt to modify system services in skill instructions.

• Attempt to modify system services in skill instructions detected (high risk: 0.70). The skill includes behavior that programmatically bypasses a macOS "Allow debugging connection?" prompt via osascript (auto-dismissing a security dialog) and declares installation of a CLI under /usr/local/bin, which touches privileged system paths — both are actions that can alter or bypass local security controls.