resemble-detect

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

• Third-party content exposure detected (high risk: 0.80). The skill explicitly requires ingesting media from arbitrary public HTTPS URLs (see "Media Requirement: All media must be at a publicly accessible HTTPS URL" and the POST /detect "url" parameter in SKILL.md), so untrusted third-party audio/image/video content can be analyzed and its extracted/transcribed text or intelligence can influence the agent's decisions and follow-up actions.