Skip to content
TakoAPI
Back to skills

ytwatchervideo

by stveenli

Coding Agents & IDEs
—Not yet rated
Your rating

Brief

Fetch and read transcripts from YouTube videos.

Description

What This Skill Does

Fetches text transcripts from YouTube videos using yt-dlp. Lets an agent summarize, answer questions, and extract information from video content without watching it.

Pulling the transcript is faster than watching and gives the agent exact text to search and quote from rather than relying on memory or description.

When to Use It

  • Summarizing a long tutorial you don't have time to watch
  • Extracting key points from a conference talk or lecture
  • Searching a video for a specific quote or piece of information
  • Getting a written record of a video for notes or reference
  • Answering questions about a video's content during a research session

Example Workflow

INPUT

User asks: Summarize this YouTube video: https://www.youtube.com/watch?v=dQw4w9WgXcQ

AGENT
  1. 1

    Run get_transcript.py with the provided YouTube URL

  2. 2

    Read the returned transcript text

  3. 3

    Identify the main topics and key points in the text

  4. 4

    Write a summary based on the extracted content

OUTPUT

A written summary of the video's content derived from its transcript

Requirements

  • yt-dlp installed and available in PATH

Installation

clawhub install stveenli/ytwatchervideo

Run the CLI command, or paste the prompt into Claude Code, Cursor, Codex or whichever coding agent you use.

Links