jayce-agent-video-pipeline skill
A Codex skill that builds a video from a voice-over script: narration, subtitles, visuals, animation, checks and a presenter.
- 108 ★ on the repository
- Last commit
- MIT
Broad access
Why this level
- A voice sample and presenter portrait are placed into the video, so rights to them are required
- The skill runs ffmpeg, Node.js and local models and writes files to the workspace
Install
- Works in
In your terminal, with SkillFoxx CLI
npx skillfoxx add skills/jayce-agent-video-pipelineDetects the agents on your machine, shows the access level and pins the version.
Other ways to install
Access and checks
How we reviewBroad accessThe tool can change data, run commands or handle secrets.
- A voice sample and presenter portrait are placed into the video, so rights to them are required
- The skill runs ffmpeg, Node.js and local models and writes files to the workspace
This is third-party code. Review the repository files before installing.
What it does
The pipeline takes a confirmed voice-over script and turns it into a finished video. Stages: narration, subtitles, picking and rendering visuals, animation, quality checks and, if needed, overlaying a digital presenter with the `compose-avatar-video` command. The base video must pass checks before the presenter is added. All settings live outside the repository: the `scripts/init_config_root.py` script creates an `.agent-video/` directory in the workspace with profiles, local configuration, projects and assets (voice sample, presenter portrait or video, character, logo, music). Python 3.10 or newer, `ffmpeg` and `ffprobe` are required; animation needs Node.js and HyperFrames, and local narration and avatar need their own tools. `references/` holds profile contracts, frame layout schemas, a motion catalogue and a voice stability check.
Who it is for. Authors of short educational and expert videos who write the text themselves and want to assemble a video with narration and animation through a pipeline.
Good fit when
- A confirmed script must quickly become a video with narration and subtitles.
- Frame animation is made in HyperFrames and checked before delivery.
- A video series should stay on one voice and style profile.
- A digital presenter needs adding to a finished base video.
Not a fit when
- The text is not yet approved: the pipeline expects a finished script.
- You lack rights to the voice or portrait that goes into the video.
- A fully cloud service is needed without installing Python, `ffmpeg` and Node.js.
Example request
Build a video from the approved text in script.md: narration, subtitles, animation, then a quality check.Limitations
The repository documents installation only for Codex: the `~/.codex/skills/` path is hard-coded in the instructions. A full cycle needs several external tools, and local narration and avatar require separately built models. The voice sample and presenter portrait are placed into the video, so rights to them stay with you. The README and instructions are in Chinese and the video tutorial is hosted on Feishu. The repository was last updated on 26 October 2026 and uses the MIT licence.
How to disable. Delete the `~/.codex/skills/jayce-agent-video-pipeline` folder and, if you like, the `.agent-video/` working directory with projects and assets.
FAQ
Where to start?
With installation into `~/.codex/skills/`, a virtual environment with `requirements.txt` and the `scripts/init_config_root.py` command that creates `.agent-video/`.
What must be installed beforehand?
Python 3.10 or newer, `ffmpeg` and `ffprobe`. Animation also needs Node.js and HyperFrames, and local narration and avatar need their own tools.
Where are voice samples and the presenter portrait kept?
In `.agent-video/assets/voice/` and `.agent-video/assets/avatar/` of the workspace.
When is the digital presenter added?
After the base video passes checks: then `compose-avatar-video` is called.
Does it work with Claude Code?
The documentation describes installation for Codex; other agents require moving the skill to their directory yourself.
Discussion
No questions or reviews yet.
Related
SkillsA skill that turns an idea, a link or a screen recording into finished video: HeyGen avatar, b-roll, captions and quality checks
A video editing skill for Claude Code and Codex: the agent cuts pauses and filler words, grades color, burns subtitles and adds animated overlays
A skill that takes Claude Code from a song and lyrics to a finished Vocaloid-style lyric video, rendered in a browser.
The Video by Reference skill learns style from a sample clip and builds an animated short from code with an original soundtrack
A Codex and Claude Code plugin with a local storyboard table: the agent writes shots and makes images, motion and voice-over
Skills, a CLI and an MCP server that let Claude Code or Codex make video from a brief: editing, graphics, voiceover, subtitles
Comparison with similar tools
Cards that share the main search query, and how they differ.
| Tool | Official | Authentication | Repository stars | Main difference |
|---|---|---|---|---|
| jayce-agent-video-pipeline (this card) | no | 108 | ||
| HyperFrames | no | 83,283 | HeyGen skills and CLI where the agent writes HTML compositions and renders them to MP4; here a full pipeline from text to videos with narration and checks | |
| short-video-maker | no | not required | 1,491 | an MCP server and REST API for short vertical videos; here a Codex skill with profiles and a digital presenter |
README in short
The README walks through steps: preparing the environment, installing into Codex, creating external configuration and starting a project, with the full video course linked on Feishu.