markdown-url

CRITICAL E006: Malicious code pattern detected in skill scripts.

• Malicious code pattern detected (high risk: 1.00). This skill explicitly requires routing visited URLs through an external proxy (https://markdown.new/) and pulling content from that proxy, which creates a clear channel for exfiltrating full URLs (including paths and query parameters/tokens) to an external service — a high-risk data-exfiltration behavior.

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

• Third-party content exposure detected (high risk: 0.85). SKILL.md:19-28 requires visiting a runtime-rewritten URL via https://markdown.new/... and extracting the page’s rendered/raw Markdown, which can include outsider-authored free text from arbitrary web pages.

MEDIUM W012: Unverifiable external dependency detected (runtime URL that controls agent).

• Potentially malicious external URL detected (high risk: 0.90). The skill mandates rewriting and fetching pages through the external proxy URL https://markdown.new/ (e.g., https://markdown.new/https://example.com) at runtime and using the fetched Markdown as source material that will directly influence agent prompts and outputs.