vibe-matcher

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

• Third-party content exposure detected (high risk: 0.70). The skill's VisualDNA explicitly includes moodBoard.references as "URLs to inspirational examples" and expects moodBoard.takeaways ("What to learn from each reference"), which indicates the agent will accept and interpret arbitrary external URLs/user-provided web content as part of its workflow (see the VisualDNA moodBoard fields and the validation script that checks references), exposing it to untrusted third‑party content and potential indirect prompt injection.