skills.rest
CodeWithSally
CodeWithSallyOFFICIAL
Explore all skills by this author →
GitHub

talking-head-recut

Overlay timed, designed graphic cards onto talking-head videos synced to the transcript.

13|17|Updated Jan 30, 2025
Search Tags:#transcription#whisper#video editing#gsap#talking head#mp4 rendering#graphic overlays
One-click install
npx skills add https://github.com/CodeWithSally/CodeWithSally-Apex-English --skill talking-head-recut-codewithsally
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: talking-head-recut
Source: https://github.com/CodeWithSally/CodeWithSally-Apex-English/tree/main/Sessions/ClaudeCode/Session007/.agents/skills/talking-head-recut
Command: npx skills add https://github.com/CodeWithSally/CodeWithSally-Apex-English --skill talking-head-recut-codewithsally

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill requires hyperframes, gsap, and includes references (resource) and assets (resource) components.

What problem does it solve? Raw talking-head, interview, and podcast footage is visually flat. This Skill packages an existing clip with timed, designed graphic overlay cards — kinetic titles, lower-thirds, data callouts, quotes, side panels, and picture-in-picture — synced to what is actually being said, while the original video plays untouched underneath. ## Core Features & Use Cases - Transcript-driven card design: Transcribes audio locally with Whisper via hyperframes transcribe, then plans card timing and content from the word-level transcript with duration- and density-based pacing rules. - Composable visual system: Mixes 10 card styles, 4 layouts (split, stack, pip, overlay), and 3 video frames across 16:9, 9:16, and 4:5 canvases, rendered to MP4 with GSAP-driven animation. - Use Case: Take a 2-minute founder interview clip, confirm a 9:16 portrait canvas with a stack layout, and produce a social-ready MP4 where quote cards and data callouts appear in rhythm with the speaker's words. ## Quick Start Package my interview clip videos/demo.mp4 with designed graphic overlay cards and render it as a 9:16 portrait video.

Frequently Asked Questions about talking-head-recut

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I add graphic overlays to a talking-head video?▼

Provide the video file and confirm the aspect ratio, layout, style group, and card count. The skill transcribes the audio with local Whisper, designs timed HTML cards from the transcript, assembles a GSAP composition, and renders the result to MP4 with hyperframes.

What is the difference between graphic overlays and embedded captions?▼

Embedded captions render the spoken words as subtitle text on the video. Graphic overlays add designed cards — titles, lower-thirds, data callouts, quotes — on top of the playing footage. Use the embedded-captions skill for plain subtitles and this skill for designed graphics.

Does video transcription require an API key or external service?▼

No. Transcription runs locally through Whisper via the hyperframes transcribe command, producing a word-level transcript.json with start and end timestamps. No API key, proxy, or rate-limited third-party service is involved.

What aspect ratios and layouts does the video output support?▼

Output supports 16:9 (1920x1080), 9:16 (1080x1920), and 4:5 (1080x1350) canvases. Layouts include split, stack, picture-in-picture, and full-screen overlay, combined with 10 card styles and 3 video frame treatments.

Why does my rendered video show a black tail at the end?▼

Whisper can return a final word end timestamp slightly past the actual clip duration. Clamp every card endSec and the composition durationSeconds to the duration reported in metadata.json so the render does not extend past the source video.

Related Skills

Explore More Content & Communication Skills→
heygen-comheygen-com
43.5k

talking-head-recut

Overlay timed graphic cards onto talking-head videos synced to transcripts.

Official
Advanced
agentpoetagentpoet

Publisher Skill (Blotato MCP)

Publish and schedule video content to TikTok, Instagram Reels, and YouTube Shorts.

Community
Intermediate
bachdyonbachdyon

subtitle-screen-splitter

Convert spoken subtitle text into paginated screen-sized pages for karaoke and video rendering.

Community
Intermediate
Marshal-NguyenMarshal-Nguyen
1

hyperframes

Routes video and animation requests into HyperFrames HTML composition workflows.

Community
Advanced