kelly-digital-humanlisted

# Kelly Digital Human ## Overview Use this skill as Kelly's digital-human solution desk. It helps pick the right path, draft the integration plan, prepare scripts and QA gates, and open a local multimodal demo showing voice/text input, avatar rendering, lip-sync/video stream status, latency, and launch readiness. Default to the local demo app when the user asks for a demo, prototype, vendor comparison, or "快上线". Use chat-only mode only when the user explicitly says "chat only", "纯聊天", "不要打开 UI", or similar. ## Two Implementation Paths ### 1. 2D photoreal digital human, low cost and fast launch Use this when the goal is an online demo, customer-service agent, product explainer, training host, livestream assistant, or any scene where speed and cost matter more than fully custom body motion. Operating model: - Connect an existing digital-human service, for example Silicon Intelligence, Tencent Zhiying, ZEGO-style real-time avatar/RTC providers, or another vendor the user already has access to. - Send text or a speech/audio stream into the service. - Receive a rendered video stream or clip with lip-sync and facial motion. - Keep all business logic, script review, safety wording, telemetry, and QA gates in Kelly's local workflow. - Best first milestone: one approved persona, one scene, one voice, one Chinese and one English demo script, one web demo page, and a latency/quality dashboard. ### 2. 3D custom digital human, high freedom Use this when the brand needs a custom