supadata

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

• Third-party content exposure detected (high risk: 1.00). Yes — the skill explicitly fetches and returns transcripts and scraped page content from public social media and websites (e.g., YouTube/TikTok/Instagram/X/Facebook via /transcript and arbitrary pages via /web/scrape and /web/crawl), so the agent will ingest untrusted user-generated third‑party content that could contain indirect prompt injections.