derek-zhuolin/interflow-video-cut — 把本地口播视频自动剪成卡片式成片的 Agent Skill：抽音轨 → ElevenLabs 转录 → AI 逐张写 HTML 卡片 → 渲染 MP4。10 种视觉风格 × 4 布局，转录只走 Ele…

Description and use

**derek-zhuolin/interflow-video-cut** automates the repurposing of talking-head videos into vertical, card-style shorts by leveraging ElevenLabs for transcription and AI-generated HTML layouts, eliminating the need for local model inference. The practical adoption path suggests starting with a small proof-of-concept to validate the ElevenLabs dependency and rendering pipeline against specific content workflows. With medium production readiness, it is currently best suited for prototypes or internal automation, requiring a thorough review of licensing, security, and long-term maintenance before deploying in a production environment.

Problem

derek-zhuolin/interflow-video-cut helps automate short-form video production and repurposing for vertical channels.

Outcome

Medium: useful for prototypes or internal workflows, with dependency and maintenance checks before production.

Workflow

• produce Shorts/Reels-style clips
• repurpose long videos into 9:16 assets
• automate captions and clipping

Evidence

把本地口播视频自动剪成卡片式成片的 Agent Skill：抽音轨 → ElevenLabs 转录 → AI 逐张写 HTML 卡片 → 渲染 MP4。10 种视觉风格 × 4 布局，转录只走 ElevenLabs 永不下本地模型。Turn a talking-head video into an AI-composed card-based video.

Source