Talking Head Editor

Edit raw talking-head footage into a finished video with Claude. Tight cuts that never clip a word, studio-clean voice, reframing and punch-ins, word-timed captions, and cinematic motion graphics written as code, all rendered frame-exact with ffmpeg and HyperFrames. Use when someone hands you a raw recording (camera file, optional separate mic) and asks to "edit this video", "cut my talking head", "add motion graphics", "clean up the audio" or "make this look edited".

9 stars 1 forksImages, Video & Audio

Install

skills CLI (Claude Code, Cursor, Codex and more)

npx skills add griffinwooldridge/talking-head-editor --skill talking-head-editor

Manual install

git clone --depth 1 https://github.com/griffinwooldridge/talking-head-editor.git /tmp/talking-head-editor
mkdir -p ~/.claude/skills
cp -r /tmp/talking-head-editor/. ~/.claude/skills/talking-head-editor

Personal skills live in ~/.claude/skills. Use .claude/skills in a project folder to share a skill with your team.

From the SKILL.md

# Talking-head editor

Turns a raw talking-head recording into an edited video: cut, cleaned voice, framing, captions and motion graphics (MG). Every step is a script you can re-run, and every MG run is code, so notes like "speed that up" or "rethink this scene" are quick edits.

Skill dir = the folder holding this file (`$SK` below). Work in a project folder next to the footage:
```
project/
  source.mp4          raw camera file (or any name; pass --src)
  edit/               words.json, transcript.txt, keep.json, voice.wav, edl.json, words_cut.json, cut.mov
  mg/                 copied from $SK/mg: plan.json, runs/rNN.js, renders/
  out/final.mp4
```

## 0. Setup (once)
`$SK/scripts/setup.sh` creates `$SK/.venv` (whisper, DeepFilterNet, OpenCV), fetches GSAP and checks ffmpeg, node and HyperFrames. Scripts re-exec themselves inside that venv.

## 1. Voice
`python3 $SK/scripts/voice.py source.mp4` writes `edit/voice.wav`:
- DeepFilterNet removes noise and room echo, capped at 24 dB of attenuation (stronger sounds robotic).
- A gentle match-EQ toward a natural speech spectrum, then -14 LUFS.

If the user recorded a separate mic, pass that file instead, but only once it's synced to the camera, starting on the same frame.
Listen to 10 s of it; if it sounds processed, re-run with `--atten 18`.

## 2. Transcribe and choose the take
`python3 $SK/scripts/transcribe.py edit/voice.wav` writes:
- `edit/words.json`, word timings in source seconds

Read the full skill on GitHub

Details

Skill name: talking-head-editor
Repository: griffinwooldridge/talking-head-editor
Repo updated: 27 October 2026

FAQ

What does the Talking Head Editor skill do?

How do I install the Talking Head Editor skill?

skills CLI (Claude Code, Cursor, Codex and more): npx skills add griffinwooldridge/talking-head-editor --skill talking-head-editor. Skills go in ~/.claude/skills (personal) or .claude/skills (project), and Claude loads them automatically when a task matches the description.

Is Talking Head Editor popular?

It lives in griffinwooldridge/talking-head-editor, which has 9 GitHub stars and 1 forks.

Related skills

3d Brain Animations

SamedVossberg/3d-brain-animations

Render 3D still images and rotating turntable animations of brain regions from ANY BrainGlobe atlas (mouse, rat, human, zebrafish, etc.), with optional glow/bloom highlighting and a semi-transparent whole-brain shell for context. Use this WHENEVER the user wants to visualize, render, picture, or animate brain anatomy: any species, any region (hippocampus, CA1, dentate gyrus, cortex, thalamus, amygdala, ...), any color, a glowing/highlighted region, a rotating brain video, or a neuroanatomy figure for a paper/thesis/slide — even if they don't mention Blender or BrainGlobe. Handles requests like

Images, Video & Audio

Reeljet

Xabilimon1/reeljet

Use when the user wants to create a short promo or demo video (Reels/TikTok/Shorts/landing hero) for an app, web, or SaaS from their own screen recordings and screenshots. Orchestrates creative direction, Higgsfield-generated motion/b-roll, local music, and deterministic ffmpeg assembly into master 16:9 + 9:16. Trigger: /reeljet.

Images, Video & Audio

Notebook Video

hyt315/notebook-video

Turn a topic into a Chinese 16:9 teaching video (flat / cel component themes, or the journal / papercut / archive hand-made skins driven by a metaphor-first director layer) with React, TypeScript and Remotion. Every shot is a code-drawn component scene — consoles, diffs, checklists, stat rows, funnels, progress rings, kinetic numbers, stamp seals, verdict bars, or a real screenshot / screen recording in a MediaFrame — moved by an engine-level virtual camera that pushes in on whatever the narration names. Weak-model friendly: five teaching blueprints with per-shot component picks, golden exampl

Images, Video & Audio

Media Analysis

kajisho5/media-analysis-skill

Measure deterministic facts about a media file (container, streams, video/audio format, silence, loudness, integrity, scene cuts, timing) with local ffprobe/ffmpeg and return a structured Observation. Use when a task needs to *know* something about a video or audio file before deciding anything: duration, resolution, fps, CFR/VFR, codecs, channel layout, where the silence is, how loud it is (LUFS / true peak), whether it decodes cleanly, where hard cuts are. Do not use it to edit, convert, normalise or export media (that is ffmpeg-skill) or to interpret the numbers (that is the agent). Python

Images, Video & Audio

Vsdx Trace

ZJUZhiyuCai/vsdx-trace

Reconstruct raster references as high-fidelity, independently editable Microsoft Visio VSDX files. Use when a user asks to 复刻、临摹、重绘 a screenshot, paper figure, flowchart, system diagram, or model architecture as editable Visio, especially when iterative visual matching matters. Do not use for ordinary image generation or flat image embedding.

Images, Video & Audio

Linear Style Illustration

yanliudesign/linear-style-illustration

Dark "Linear/AI-OS" style technical diagrams — near-black background, cyan × purple × lilac duotone, JetBrains Mono uppercase labels, pill-shaped SVG nodes, glowing hub/kernel layer. Use when the user wants a "Linear 风格 diagram", a dark tech/architecture diagram, a product flow / state machine / layered-stack / hub-and-spoke diagram for a deck or landing page, or references screenshots that look like a dark AI product OS deck. Outputs a self-contained HTML file with inline SVG, ready to drop into a slide deck or export as PNG. Keywords: Linear style diagram, dark tech diagram, kernel diagram,

Images, Video & Audio

Browse the most popular Claude skills, all images, video & audio skills, or write your own with the skills generator.