video-summary

Extract subtitles from video URLs and generate structured summaries for LLM processing.

1|1|Updated Mar 18, 2026
One-click install
npx skills add https://github.com/xianmingyao/openclaw-CaySon --skill video-summary
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: video-summary
Source: https://github.com/xianmingyao/openclaw-CaySon/tree/main/skills/video-summary
Command: npx skills add https://github.com/xianmingyao/openclaw-CaySon --skill video-summary

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill requires yt-dlp, jq, ffmpeg, ffprobe, bc, whisper.

What problem does it solve?

Video creators and analysts spend hours watching long videos to pull key insights. This skill automates subtitle extraction and structured summarization across YouTube, Bilibili, Xiaohongshu, Douyin, and local files, saving time and ensuring consistent outputs.

Core Features & Use Cases

  • Auto-detect platform from URL and pull transcripts or subtitles
  • Generate structured summaries with key insights, timestamps, and actionable takeaways in text, JSON, or Markdown
  • Support restricted-content access via cookies and optional Whisper transcription when subtitles are unavailable
  • Outputs are prepared for downstream LLM processing as structured prompts, not direct API calls

Quick Start

Pass a video URL to video-summary to produce a structured, platform-aware summary of the content.

Frequently Asked Questions about video-summary

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

FAQPage Schema
How do I extract subtitles and generate video summaries from YouTube or Bilibili?

Video summaries are generated by auto-detecting the platform from a URL, extracting subtitles or transcripts, and producing structured outputs with key insights and timestamps. It supports YouTube, Bilibili, Xiaohongshu, Douyin, and local files.

Do I need ffmpeg and yt-dlp to summarize videos with subtitles?

Yes, ffmpeg, ffprobe, yt-dlp, jq, and bc are required dependencies for video summarization, while whisper is optional for transcription when subtitles are unavailable on the target platform.

Can I transcribe videos that have no subtitles available?

Yes, when subtitles are unavailable, the skill optionally uses whisper for transcription to generate the text needed for structured video summaries across supported platforms or local files.

How do I get structured JSON or Markdown summaries from video transcripts?

Structured video summaries in text, JSON, or Markdown format are produced by extracting transcript data and emitting a machine-readable prompt containing key insights, timestamps, and actionable takeaways for downstream LLM processing.

Does video summarization work with restricted content on platforms like YouTube?

Yes, restricted content access is supported via cookies, allowing the tool to extract subtitles and generate summaries from videos that require authentication on platforms like YouTube or Bilibili.

What's the best way to automate video summarization across multiple platforms?

Automating video summarization across platforms involves passing video URLs to the skill, which auto-detects the source, extracts subtitles, and outputs structured prompts for downstream LLM processing without making direct API calls.