scrape-codegen-analyze

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

• Third-party content exposure detected (high risk: 0.85). The skill ingests outsider-authored free text by reading a saved HTML page (page_html_path) that originates from a runtime-fetched source URL recorded in meta.json, and then it parses the cleaned HTML (clean_html.py output) to analyze and extract text for the LLM context.