
yang0's comic storyboard gallery showing grid, four-panel, large-panel and diagonal storytelling layouts.
Project screenshotsAbout this Skill
Configure your own service account through the bundled voice client, then synthesize requested text or transcribe supplied audio. Episode narration uses the video controller gates. Audio durations are measured and subtitles use returned speech boundaries rather than guessed timing; listen to the result and check the transcript.
Why we picked it
This workflow focuses on volcano voice & subtitle tools within the Handraw Style package. Its own entrypoint defines the task; retain the shared package assets and linked sibling workflows.
What it helps with
- Use Volcano or Doubao voice services for narration and speech recognition with actual subtitle timestamps.
What the workflow does
Configure your own service account through the bundled voice client, then synthesize requested text or transcribe supplied audio. Episode narration uses the video controller gates. Audio durations are measured and subtitles use returned speech boundaries rather than guessed timing; listen to the result and check the transcript.
Install and invoke the workflow
Keep the full package
Ask Codex to install the official repository, keeping the root SKILL.md, images, shared scripts, references and sibling Skills together. The command below clones the source for inspection; configure it in the Agent Skills root as described by the author.
git clone https://github.com/yang0/handraw-style.gitChoose this entrypoint
Ask the agent to use skills/handraw-volc-tts/SKILL.md (declared name: handraw-volc-tts). Follow its linked dependencies and, for route-specific video producers, the video controller's route and plan requirements.
Upstream Skill entrypointOfficial package and installation guidePackage dependencies
A useful first request
Generate calm narration from this text with my configured voice account and prepare SRT captions using actual speech timestamps.
Tools, costs and shared previews
Python 3.10+, FFmpeg, network access and your own Volcano voice service permissions are required. Credentials and service charges are not supplied by the repository.
The repository is free under its attribution license. Agent subscriptions and external generation or voice services have separate costs. This detail page reuses the collection screenshots supplied by the user; they do not demonstrate a new run of this particular workflow.
Before you use it
- Python 3.10+, FFmpeg, network access and your own Volcano voice service permissions are required. Credentials and service charges are not supplied by the repository.
- Install the complete Handraw Style repository and read the root entrypoint before invoking this nested workflow. Catalog entries are separate; their shared files remain one package.
Supported agents
Sources & references
Original sources are linked below. Practical starting points are editorial suggestions from SkillHey.


