ai-video-to-talking

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

• Potentially malicious external URL detected (medium risk: 0.30). The skill instructs fetching and running a remote npm package (@sync.so/mcp-server@latest) and loading a remote MCP server and REST API from sync.so, establishing a common third-party operational dependency.

MEDIUM W009: Direct money access capability detected (payment gateways, crypto, banking).

• Direct money access detected (high risk: 0.70). The skill uses API calls to generate lip-sync videos and speech previews via the sync.so service (e.g., generate_estimate-cost, create-lipsync, create-speech-preview, create-timeline-lipsync), which represents direct financial operations (purchases/charges). These calls rely on a SYNC_API_KEY or MCP server for authentication, but no qualifying independent access control is documented for these payment/billing routes. Therefore, it is a direct operation without established qualifying control.