content-extract

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

• Third-party content exposure detected (high risk: 0.80). The skill explicitly fetches and ingests arbitrary public webpages (via the web_fetch probe in SKILL.md and the MinerU fallback called by scripts/content_extract.py -> mineru_parse_documents.py) and returns extracted markdown that downstream agents are expected to read and act on, so untrusted page content could materially influence agent behavior.