Vox Director skill
A skill that turns one topic into a finished Vox-style paper-collage video: script, keyframes, motion, voice-over, music and captions
- 2,810 ★ on the repository
- Last commit
- MIT
Extended access
Why this level
- The skill calls the paid Atlas Cloud API and spends money on generation
- It runs local scripts and ffmpeg and needs a key in the environment
Install
- Works in
Manual install
git clone https://github.com/Alisa0808/vox-director.git ~/.claude/skills/vox-directorOther agents read AGENTS.md and SKILL.md. There is also a packaged vox-director.skill file to install through the Claude UI.
Access and checks
How we reviewExtended accessThe tool runs code or makes network calls and only reads external data.
- The skill calls the paid Atlas Cloud API and spends money on generation
- It runs local scripts and ffmpeg and needs a key in the environment
This is third-party code. Review the repository files before installing.
What it does
The skill drives a video from one topic to a finished clip in the torn-paper Vox collage style. It writes a script and beat map, generates a collage poster per beat, then adds motion: by default it animates the whole poster with a video model, and for piece-by-piece assembly it cuts the poster into parts and drives them with a local keyframe engine. Then come voice-over, music and captions, assembled into the final file with ffmpeg. There are three input modes: a topic, a talking head and a single photo of a person or product placed into the collage.
Who it is for. For creators, marketers and designers who need a short explainer or ad video in a collage style.
Good fit when
- You need a short explainer or ad in a paper-collage style
- You want to turn a topic or product into a narrated collage video
- You need a clip with voice-over, music and captions without manual editing
Not a fit when
- You have no Atlas Cloud key or cannot pay for generation
- You need a different visual style, not a paper collage
- You cannot install ffmpeg and python locally
Example request
Make a collage video about Mexican street food, English, 16:9, 15 secondsLimitations
Generation runs through the paid Atlas Cloud API with an ATLASCLOUD_API_KEY, which spends money on image, video and voice models. Locally it needs ffmpeg, ffprobe and python with Pillow. Atlas Cloud availability from Russia without a VPN is not verified.
How to disable. Remove the vox-director folder from the agent's skills directory and the ATLASCLOUD_API_KEY variable from your environment.
FAQ
What else do I need besides an agent?
An Atlas Cloud API key, plus local ffmpeg with ffprobe and python with Pillow for caption and watermark overlays.
How is motion added?
By default a video model animates the whole poster; for piece-by-piece assembly the poster is cut into parts and driven by a local engine.
Discussion
No questions or reviews yet.
Related
SkillsA skill that builds a narrated paper-collage explainer video from a topic, a talking-head clip or a single photo
Claude Code skill that turns one topic into a 60-second Vox-style explainer: script, keyframes, animation, voiceover and music.
Skill for Codex and Gemini: turns a phrase, text or SRT into a Vox-style storyboard with prompts and JSON for Google Flow
A skill that turns a topic into a narrated 4K video: research, script, local TTS, Remotion composition and an FFmpeg mix
A skill that turns your narration and real assets into a 20–60 second video in the style you pick, rendered locally as MP4.
Director skill that drives a video from idea and script to scenes, characters, voices and finished clips
README in short
The README explains that the collage look is born in the image step and motion is added separately: either a video model for the whole poster or a piece-by-piece local engine. It shows the three input modes, install by cloning into the skills directory or from a packaged .skill file, the Atlas Cloud key and the ffmpeg and python requirements. The skill fetches the live model list through the API before running. The repo holds references on prompts, beats and voices, examples and assembly scripts. MIT licensed.
README badge
Original SKILL.md
--- name: vox-director description: > Turn ONE topic into a finished Vox-style paper-collage explainer / ad video, end to end on the Atlas Cloud API + local ffmpeg — script, collage keyframes, motion, voice-over, music, captions, all automated. Use this whenever the user wants a "Vox style" video, a paper/torn-paper collage animation, a "motion collage", a narrated explainer or short ad built from AI-generated collage posters, a scrapbook-style tribute, or wants to turn a topic / product / person into a punchy narrated collage video — even if they don't say the word "Vox". Also use when reproducing Stav Zilber / rom1trs / Higgsfield-style collage ad workflows. Three input modalities: a topic (B-roll), a talking-head video (A-roll mode), or a single photo of a person/product anchored into the collage (C-roll mode). Triggers: "vox video", "collage video", "motion collage", "paper collage explainer", "make a collage ad", "turn this topic into a collage video", "turn my photo/this product shot into a collage video". --- # Vox Director Turn a one-line topic into a finished **Vox-style paper-collage video**: a bold, punchy, narrated explainer/ad where each beat is a torn-paper collage poster that comes alive, with voice-over, music and captions. Runs on **one Atlas Cloud API key** + local **ffmpeg**. The look is the modern editorial paper-collage popularized by Vox explainers and creators like Stav Zilber / rom1trs: hand-cut paper cut-outs, torn edges, tape, halftone dots, newspaper clippings, bold flat color per beat, big cut-out headlines. ## The core idea (read this first) The Vox collage look and the collage motion are **two different steps**: 1. **The look is born in the IMAGE step.** Each beat is a finished collage *poster* made by a text-to-image model. All the collage DNA (torn paper, cut-outs, halftone, bold color, headline text) lives in that image. If the image isn't a rich collage, nothing downstream will save it. 2. **The motion is added after.** By default an AI video model animates the whole poster (the "living poster" path — simple, automated). For dramatic *piece-by-piece* assembly you cut the poster into parts and drive them with the local keyframe engine (advanced path).