mrbeast-perspective

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 Agentic Protocol explicitly requires using WebSearch and other tools to fetch live web data ("⚠️ 必须使用工具（WebSearch等）获取真实信息") and the repository includes runtime scripts that download public YouTube content (scripts/fetch_youtube_subtitles.sh), so the agent will ingest untrusted, user-generated third‑party webpages/subtitles as part of its decision-making workflow.