video-watcher

Fetch cleaned subtitle transcripts from YouTube or Bilibili videos.

1|Updated May 20, 2026
One-click install
npx skills add https://github.com/Wanli-Lee/CUA-Claw-Harness --skill video-watcher
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: video-watcher
Source: https://github.com/Wanli-Lee/CUA-Claw-Harness/tree/main/skills/5/bilibili-youtube-watcher
Command: npx skills add https://github.com/Wanli-Lee/CUA-Claw-Harness --skill video-watcher

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill requires yt-dlp, and includes scripts (resource) components.

What problem does it solve?

It helps you quickly obtain clean subtitle transcripts from YouTube or Bilibili videos so you can summarize or extract accurate information without manually watching the video.

Core Features & Use Cases

  • Transcript fetching for two platforms: Automatically detects whether a URL is from YouTube or Bilibili and retrieves subtitles for it.
  • Language control with sensible defaults: Uses en for YouTube and zh-CN for Bilibili unless you override with a language code.
  • Readable transcript output: Cleans VTT/SRT content into plain text with timestamps and formatting removed.
  • Common use cases: Summarize a talk, answer questions about a video’s content, or extract specific details (e.g., steps, claims, or dates) for downstream analysis.

Quick Start

Run the transcript fetcher with an auto-detected platform by giving your video URL as the first argument to the script.

Frequently Asked Questions about video-watcher

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

FAQPage Schema
How do I extract YouTube transcripts without downloading the video?

Extract YouTube transcripts without downloading by using yt-dlp to fetch only subtitle files, then parse VTT or SRT formats into cleaned plain text. The script automatically detects YouTube URLs and defaults to English subtitles for immediate text output.

Can I fetch Bilibili video subtitles in Chinese?

Fetch Bilibili video subtitles in Chinese by providing the video URL; the script defaults to zh-CN language code. It parses VTT or SRT subtitle formats and removes timestamps to produce readable plain text transcripts.

Do I need yt-dlp to get video transcripts for summarization?

Yes, yt-dlp is required to get video transcripts for summarization. It handles fetching subtitle files from YouTube or Bilibili without downloading the video, providing the VTT or SRT inputs for plain text conversion.

How do I summarize a video using subtitle parsing?

Summarize a video using subtitle parsing by fetching the cleaned plain text transcript first. Once the VTT or SRT file is converted into readable text, it can be passed to downstream tasks for content summarization or Q&A.

Can I override the default subtitle language for video transcripts?

Override the default subtitle language for video transcripts by specifying a desired language code. Defaults are en for YouTube and zh-CN for Bilibili, but the script supports configurable subtitle languages for broader video comprehension.

What is the best way to extract information from a video without watching it?

The best way to extract information from a video without watching is to fetch readable subtitle transcripts. The script cleans VTT or SRT content into plain text, allowing accurate information extraction and content summarization across supported domains.