produce-original-faceless-video

Produces approval-gated, source-cited faceless motion-explainer video packages with layered animated assets.

Updated Jul 22, 2026
One-click install
npx skills add https://github.com/Chau165/local_skill --skill produce-original-faceless-video-chau165
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: produce-original-faceless-video
Source: https://github.com/Chau165/local_skill/tree/main/codex/skills/produce-original-faceless-video
Command: npx skills add https://github.com/Chau165/local_skill --skill produce-original-faceless-video-chau165

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes scripts (resource) and references (resource) components.

What problem does it solve? Creating original faceless explainer videos requires coordinating scripts, citations, generated image assets, animation, and rights tracking; this Skill turns that into a gated, auditable production workflow that prevents copying a reference channel's identity. ## Core Features & Use Cases - Gated production workflow: Initializes a workspace from an analyzer channel-profile.json and enforces concept, script, budget, rough, and final approval checkpoints with a validating Python script. - Animation-forward asset pipeline: Builds style bibles, asset manifests, and layered image-based cartoon assets, then animates them through a local HyperFrames render adapter with motion-quality QA thresholds. - Evidence and rights enforcement: Registers every factual claim with sources, records provenance, license, and SHA-256 for every asset, and prohibits voice cloning and automatic publishing. - Use Case: Given a completed channel analysis and a topic like "how solar inverters work", produce a 60-second 9:16 image_only cartoon explainer with cited claims, generated layered assets, a HyperFrames rough render, and a final package ready for manual publishing. ## Quick Start Use the produce-original-faceless-video skill to create an original 60-second image_only explainer on my topic using my completed channel-profile.json, stopping for my approval at the concept stage.

Frequently Asked Questions about produce-original-faceless-video

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

FAQPage Schema
How do I create an original faceless explainer video from a channel profile?

Run scripts/production_workspace.py init with a completed channel-profile.json, topic, duration, language, aspect ratio, and image_only or image_video mode. Then follow the gated workflow: draft the concept, script, and storyboard, obtain approvals at each checkpoint, produce layered assets, and render rough and final cuts.

What inputs does the faceless video production workflow require?

It requires a completed analyzer channel-profile.json with schema_version 1.0, analysis_status complete, observed source coverage, evidence-linked patterns, and pattern-transfer-only originality guardrails, plus a topic, duration in seconds, language, aspect ratio, and mode. Draft or copying-unsafe profiles are rejected before workspace creation.

Can I use HyperFrames to render the animated video locally?

Yes, when the local HyperFrames CLI and skills are installed, register the hyperframes-local render adapter and create the project under hyperframes/ in the workspace. You must write world-motion-proof.json mapping every storyboard shot before rendering, and cloud, auth, or publish commands remain prohibited.

Does image_only mode allow a static slideshow with zoom effects?

No. image_only restricts source media types but still requires at least two independently animated layers per shot, visible cause-and-effect interactions, and camera_only set to false. Ken Burns zooms or pans over a flattened image fail the motion standard and the motion_quality QA thresholds.

Why does workspace validation fail at the script stage?

The script stage requires approved timed segments, verified claim citations with source evidence, an approved script checkpoint, and a complete motion_design for every storyboard shot. Missing claim IDs on factual segments, unverified claims, or shots lacking subjects, interactions, and animated layers all produce validation errors.

Can the workflow clone a narrator's voice or publish automatically?

No. The manifest enforces voice.real_person_clone as false, allowing only verified synthetic TTS voices, consented original recordings, or no narration. Publishing stays manual with publish.automatic false, and the skill never uploads, schedules, or distributes renders.