clone-anywebsite

CRITICAL E006: Malicious code pattern detected in skill scripts.

• Malicious code pattern detected (high risk: 1.00). The document explicitly instructs automated scraping and "stealing" of visual assets (SVGs, fonts, data-URI logos, canvas/WebGL outputs) and provides scripts to extract and download those resources — a high-risk pattern enabling data exfiltration and intellectual-property theft.

MEDIUM W011: Third-party content exposure detected (indirect prompt injection risk).

• Third-party content exposure detected (high risk: 0.95). SKILL.md instructs the agent to fetch arbitrary OUTSIDER web content at runtime (e.g., navigate("https://target-site.com/") / cloning [TARGET_URL]), then run javascript_tool that reads outsider page DOM/text via document.querySelector(...)/innerText and returns it as JSON strings into the LLM context.