发现和使用优秀的技能扩展
生成精确的、带时间码的Seedance 2.0提示词,整合多模态输入与资产映射,用于可控的4-15秒视频创建和编辑。
Generate precise, timecoded Seedance 2.0 prompts integrating multimodal inputs with asset mapping for controlled 4-15s video creation and editing.