Give this skill to your agent

Copy the prompt and paste it into Claude Code, Codex or any agent that can read a web page. It installs the skill and tells you what it does.

Install this skill for me.

Read https://cwyz.com/skills/digital-human-video-production-en
The raw SKILL.md is at https://cwyz.com/api/skills/digital-human-video-production-en/raw

Save it as digital-human-video-production/SKILL.md in my skills folder (Claude Code: ~/.claude/skills/ for every project, or .claude/skills/ for this project only).

Then tell me in two sentences what it does and when it applies. Do not run anything it describes until I say so.

Or unzip the download into ~/.claude/skills/ (every project) or .claude/skills/ in your project. A skill tells an agent what to do. Read it before you install it, especially one you did not write.

The Digital Human Video Pipeline Goes From Brief to Final Cut

A general-purpose skill for Codex and local creative workflows that turns a product brief, reviewed narration, digital human or real-person on-camera speech, real product footage, and motion graphics into a deliverable video with review gates. It is not tied to any brand or script; product facts, copy, media, and vendor parameters live in the project directory. Multi-file repository, installed via git clone.

Use caseCombines a product brief, narration, and digital human on-camera speech into a deliverable product video with review gates.

Machine translation. Read the original

Digital Human Product Video Production Line

A general-purpose skill for Codex and local creative workflows that turns a product brief, reviewed narration, digital human or real-person on-camera speech, real product footage, and motion graphics into a deliverable video.

It is not tied to any particular brand, script, page, or visual system. Each project's product facts, copy, colors, fonts, media, and vendor parameters live in the project directory; the repository only keeps the reusable methods, rules, scaffolding, and quality gates.

What It Does

  • Turns a brief into a deliverable package with review gates: narration and on-camera speech preview → storyboard and timeline → local motion graphics and compositing → a verifiable final video.
  • The position, size, and hiding rules for the on-camera picture-in-picture (PIP) follow a quantified layout contract, not gut feel.
  • A "generation ledger" controls spending: every paid generation call produces a preview and waits for confirmation first, to avoid wasting money.
  • Quality checks before delivery (encoding, audio-video sync, subtitle safe area).

Reference Stack (Replaceable)

Voice cloning → digital human/on-camera rendering → motion graphics (seek-safe) → local deterministic compositing → FFmpeg encoding and QA. Each layer is a replaceable adapter; specific vendors, keys, and models live in the project, not in the repository.

Who It Is For

People who produce batches of on-camera product videos, explainers, demos, and launch clips; people who want the method to accumulate and be reused, while product facts, brand, and media are managed separately for each project.

Installation

This is a multi-file skill (including scripts, references, and agents), so clone the entire repository rather than a single SKILL.md. Put it in your skills directory:

# For all projects
git clone https://github.com/jaxxchen003/digital-human-video-production \
  ~/.claude/skills/digital-human-video-production
# Or only for the current project
git clone https://github.com/jaxxchen003/digital-human-video-production \
  .claude/skills/digital-human-video-production

Then have your agent (Claude Code, Codex, etc.) read SKILL.md and start. Follow its prompts to first create the project directory, and fill in product facts, brand, media, and vendor parameters.

Source and License

Done with this one? Mark it to track your path.Sign in
Save it for laterSign in
Questions about this piece?Ask AI about this

Comments (0)

No comments yet. Start the conversation.

Sign in to join the discussion.Sign in