gentic-knowledge

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

• Third-party content exposure detected (high risk: 1.00). The skill's SKILL.md explicitly instructs using vectorize_web_content to scrape arbitrary web pages and vectorize_content to ingest emails, Slack messages, reviews, and social/social-post text, so the agent fetches and reads untrusted, user-generated third-party content (web pages and social/forum-like content) that can influence its actions and tool use.