jianying-editor

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

• Third-party content exposure detected (high risk: 1.00). This skill ingests public third-party content: docs/web-vfx.md explicitly instructs creating local HTML that loads external CDN scripts (e.g., GSAP/Three.js) which the recorder will fetch and execute, and rules/media.md / rules/audio-voice.md document add_cloud_media/add_cloud_music (which "automatically searches, downloads and imports from the cloud database"); those external assets are untrusted, are read/ingested as part of the workflow, and can materially influence execution (e.g., signaling recorder completion or changing what is imported), so they could enable indirect prompt injection.

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

• Potentially malicious external URL detected (high risk: 0.80). The Web-to-Video runtime can load CDN-hosted JavaScript into generated HTML that the recorder renders (example script tag shown: https://cdn.../gsap.min.js), which will be fetched and executed at skill runtime and thus constitutes executing remote code from an external URL.