watching-videos

Download videos and analyze frames and transcripts with yt-dlp and ffmpeg.

2|Updated Mar 2, 2026

Search Tags:#video processing #video analysis #transcript extraction #video summarization #frame analysis

Ready to run

npx skills add https://github.com/joegoldin/agent-skills --skill watching-videos

💬 Prompt for ChatGPT / Claude Chat▼

Please help me install this Agent Skill.
Skill: watching-videos
Source: https://github.com/joegoldin/agent-skills/tree/main/skills/watching-videos
Command: npx skills add https://github.com/joegoldin/agent-skills --skill watching-videos

SYSTEM DOCUMENTATION & REQUIREMENTS

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

What problem does it solve?

This Skill addresses the challenge of analyzing video content, providing a framework for frame-by-frame examination and synchronized transcript analysis, suitable for extracting specific information or summarizing video content.

Core Features & Use Cases

Video Frame Analysis: Offers a detailed analysis of video frames, enabling the identification of key visual elements and events.
Transcript Extraction: Automatically generates transcripts for video content, supporting both native captions and transcriptions via third-party services.
Use Case: When a user wants to quickly determine the language spoken in a video or needs a summary of the video content, this Skill can provide both the visual frame-by-frame analysis and a text-based summary of the spoken content.

Quick Start

To analyze a video and get a transcript, use the command: /watch [video URL or path]

Frequently Asked Questions about watching-videos

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

FAQPage Schema

How do I extract transcripts and analyze video frames from an online video?▼

To extract transcripts and analyze video frames, you process the video by downloading it, extracting frames, and transcribing audio. This Skill performs all three steps automatically to provide synchronized visual and text-based content analysis.

How does frame-by-frame video analysis work for summarizing content?▼

Frame-by-frame video analysis works by extracting individual images from the video using ffmpeg. These frames are examined to identify key visual elements, while the audio is transcribed to create a comprehensive content summary.

Do I need yt-dlp and ffmpeg installed to extract video transcripts?▼

Yes, you need yt-dlp and ffmpeg installed to extract video transcripts. yt-dlp is required for downloading the video content, and ffmpeg is necessary for the frame extraction process.

Can I use audiomemo for video transcription if a video has no native captions?▼

Yes, you can use audiomemo for video transcription when native captions are unavailable. It serves as an optional third-party service to automatically transcribe the audio track for text-based analysis.

What is the best way to determine the language spoken in a video file?▼

The best way to determine the language spoken in a video file is through automated transcript extraction. By transcribing the audio track, the text output reveals the spoken language for quick identification and summarization.

What are the limitations of automated video frame analysis and transcription?▼

Limitations of automated video frame analysis and transcription include dependency on external tools like yt-dlp and ffmpeg. Additionally, transcription accuracy relies on audio quality and the capabilities of optional services like audiomemo for processing spoken content.