website-to-video

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

• Third-party content exposure detected (high risk: 0.75). Outsider free text can enter the LLM context via Step 0’s website capture outputs (e.g., capture/extracted/visible-text.txt and capture/extracted/asset-descriptions.md) which are derived from runtime-fetched third-party webpages and then read in Steps 2–3 to draft the brief/script.

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

• Potentially malicious external URL detected (high risk: 0.90). The skill explicitly loads remote JS libraries at runtime (executing fetched code) from CDN URLs — e.g. GSAP at https://cdn.jsdelivr.net/npm/gsap@3.14.2/dist/gsap.min.js and Three.js at https://cdn.jsdelivr.net/npm/three@0.181.2/+esm — which the runtime depends on for composition rendering.