faceless-channel — Codex skill: narrator-led video generation
faceless-channel produces finished multi-scene videos with narrator voiceover for YouTube automation, explainer content, documentaries, storybook retellings, and kids songs.
no license 3 installs ★ 8,052 updated 2026-10-08
npx skills add openai/plugins --skill faceless-channel -a codexWhat faceless-channel does
The skill assembles a final file from five channel types on a unified pipeline: generates frames via seedream_v5_pro, edits clips as 10-second blocks of five hard cuts (~2 seconds each) through gemini_omni, adds narrator voice (seed_audio), and burns subtitles timed by Whisper. Every scene references approved assets — location, characters, props.
Kids blocks use a four-cut interplay pattern (2.5 seconds per frame). Characters react with gestures but never speak on screen; speech stays with the narrator. Music beds are generated through sonilo_music for kids videos and fairy tales.
Runs only in the Higgsfield sandbox. Requires sandbox_exec, ffmpeg, ffprobe, python3, faster_whisper, and the HF_WORKFLOWS environment variable. Scripts live in ${HF_WORKFLOWS}/faceless-channel-video/scripts/. Aspect ratio must be passed explicitly on every call, defaults to 16:9.
What it helps with
- assemble a video for a faceless YouTube channel with narrator voiceover
- edit an explainer video from scripted scenes with burned subtitles
- produce a documentary with sequential hard-cut blocks
- generate a kids song with animated character reactions
- burn Whisper-timed captions into the final video file
Example prompts for the agent
- Make a faceless channel video about Ancient Rome
- Build an explainer video on space with narrator voiceover
- Create a kids song about animals with animation
How to install faceless-channel
1Skills CLI (Vercel)
npx skills add openai/plugins --skill faceless-channel -a codex2Manually: put the files in place
.agents/skills/faceless-channel/SKILL.mdFor all projects use ~/.agents/skills/
1Skills CLI (Vercel)
npx skills add openai/plugins --skill faceless-channel -a claude-code2Manually: put the files in place
~/.claude/skills/faceless-channel/SKILL.mdCopy the whole skill folder; for one project use .claude/skills/faceless-channel/
1Skills CLI (Vercel)
npx skills add openai/plugins --skill faceless-channel -a cursor2Manually: put the files in place
.cursor/skills/faceless-channel/SKILL.mdCursor also reads .agents/skills/ and .claude/skills/
1GitHub CLI
gh skill install openai/plugins plugins/higgsfield/skills/faceless-channel2Manually: put the files in place
.github/skills/faceless-channel/SKILL.md1Skills CLI (Vercel)
npx skills add openai/plugins --skill faceless-channel -a gemini-cli2Platform CLI
gemini skills install https://github.com/openai/plugins --path plugins/higgsfield/skills/faceless-channelManually: put the files in place
.devin/skills/faceless-channel/SKILL.mdThe legacy .windsurf/skills/ path still works
Claude.ai
Zip the folder and upload it in Settings → Capabilities → Skills
Step by step, with commands for each platform: Codex skills: how to install, where they live, how to call them
Questions about faceless-channel
How to install faceless-channel in Codex?
Run npx skills add openai/plugins --skill faceless-channel. The skill will appear in .agents/skills/faceless-channel/SKILL.md.
What does faceless-channel do?
It generates finished multi-scene videos with narrator voiceover, burned subtitles, and music beds for YouTube automation, explainer channels, documentaries, and kids songs.
What dependencies does faceless-channel need?
The Higgsfield sandbox with sandbox_exec, ffmpeg, ffprobe, python3, the faster_whisper library, and the HF_WORKFLOWS variable pointing to the scripts directory.