web-to-markdown

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

• Third-party content exposure detected (high risk: 0.90). The skill fetches and ingests arbitrary public web pages (via WebFetch in "Step 4: 웹페이지 가져오기 및 변환" and optional Playwright fallbacks) and then has the agent read/structure that untrusted, user-provided content (blogs, docs, forums, arbitrary URLs) as part of its AI-optimization workflow, creating a clear avenue for indirect prompt injection.