creatok-analyze-video

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

• Third-party content exposure detected (high risk: 1.00). The skill takes an arbitrary public TikTok URL (SKILL.md), calls creatok-client.analyze(tiktokUrl) (lib/analyze-video.js → lib/creatok-client.js) to fetch transcript/vision/response data from the open web, and the model is required to read that outputs/result.json and use the content to drive analysis and next actions, so untrusted third‑party content can influence behavior.

MEDIUM W012: Unverifiable external dependency detected (runtime URL that controls agent).

• Potentially malicious external URL detected (high risk: 0.70). The skill calls the CreatOK API (https://www.creatok.ai, e.g. /api/open/skills/analyze) at runtime and ingests returned response.content and suggestions into outputs/result.json which are then used to drive the agent's analysis, so remote content can directly control agent instructions.