byt-workflow

What This Skill Does

The byt-workflow skill is an automated pipeline designed to bridge the gap between YouTube video content and localized text translations. It orchestrates a series of technical steps—audio extraction, window management, media playback, and optical character recognition (OCR) or text capture—to convert spoken content from any YouTube URL into a structured transcript. By automating the interaction with the Doubao translation interface, it removes the manual drudgery of transcribing long-form video content, making it an essential utility for researchers, content creators, and language learners.

Installation

To integrate this skill into your OpenClaw environment, execute the following command in your terminal: clawhub install openclaw/skills/skills/banner90/byt-workflow Ensure your system has the necessary dependencies for WSL (Windows Subsystem for Linux) and Python, as the workflow relies on a cross-platform execution model to manage Windows-based GUI automation tasks from a Linux-based environment.

Use Cases

• Academic Research: Rapidly transcribe and translate long lecture videos for analysis.
• Content Localization: Generate preliminary subtitles for international audiences by automating the Doubao translation window.
• Media Archiving: Automatically extract audio from technical tutorials and generate searchable text transcripts for your personal database.
• Language Learning: Use the 'dual' mode to compare audio playback with side-by-side translated text, helping to improve listening comprehension and vocabulary acquisition.

Example Prompts

1. "Translate the YouTube video at https://www.youtube.com/watch?v=example123 into Chinese using the bilingual mode."
2. "Run the byt-workflow for https://www.youtube.com/watch?v=video456 and save the translation to my works directory."
3. "Process this video https://www.youtube.com/watch?v=tutorial789 using the default dual translation settings."

Tips & Limitations

• Desktop Visibility: Because this skill automates GUI interactions with the Doubao application, it requires an active, visible desktop session. If you are using RDP (Remote Desktop Protocol), ensure the session remains active, as disconnecting or minimizing the window may break the capture process.
• File System Paths: Be mindful of the path mappings. The skill expects files to be stored in the Windows works/ directory, accessed via WSL as /mnt/h/.... Ensure your WSL configuration has proper read/write permissions to these Windows partitions.
• Error Monitoring: Always verify the success field in the returned JSON. If success is false, inspect the error_message provided by the skill to diagnose potential issues like network failures or GUI timeouts.