Use case

Summarize the latest YouTube video of any channel with your AI agent

Your agent lists the channel's recent uploads, takes the newest one, reads its captions and summarizes them. You get the point of a 10-minute video in five lines. It takes two ClawLink calls and costs $0.00482; you need no YouTube account.

by Jay4 min read

Contents

TL;DR

QuestionAnswer
What does it do?Your agent lists the channel's recent uploads, takes the newest one, reads its captions and summarizes them.
Who is it for?People who follow many channels, researchers, and teams that track what reviewers say about a product.
What do I connect?Nothing. It reads public data with ClawLink's own provider.
What does it cost?About $0.00482 of ClawLink credit per run. New accounts get $0.50 free.
Does it work?Yes. We ran it on 2026-10-03; the results are below.
Which agents?OpenClaw, Hermes, Claude, Codex, ChatGPT and any MCP client connected to ClawLink.

How do you try it?

  1. Set up ClawLink once. Add it to your agent from the install page. This job needs no connected app.
  2. Send this prompt to your agent. Click the card to copy it.

It works the same in OpenClaw, Hermes, Claude and Codex.

How does the agent do it?

The agent makes 2 calls through ClawLink. You do not pick the tools: the agent finds them by searching ClawLink, reads their input schema and runs them.

Your agentClawLinkyoutubedata_channel_videosyoutubedata_video_captionsYouTube public data
The calls your agent makes, in order. Tool names are the real ClawLink tool names.
StepToolWhat it doesPrice
1youtubedata_channel_videosLists a channel's recent uploads by handle, with views, length and publish date.$0.002
2youtubedata_video_captionsReturns the video's captions as text, with timestamps.$0.00282

What happened when we ran it?

We sent this job through ClawLink on 2026-10-03. These are the real calls and what came back, not an example.

CallInput and resultTimeCost
youtubedata_channel_videos@mkbhd, 30 recent videos returned6.6 s$0.002
youtubedata_video_captions"Xiaomi 18 Fold: How Does This Happen?" (10:47), 7 caption languages2.3 s$0.00282

What the data showed:

  • The newest upload was 15 hours old and already had 3.93M views.
  • The video compares the Xiaomi 18 Fold with the iPhone Duo: same size and aspect ratio, unveiled two days apart.
  • The Xiaomi is 30 g lighter than the 254 g iPhone Duo and has a 6,000 mAh silicon-carbon battery with 67 W charging.
  • It has three rear cameras, including a 50 MP 3.5x zoom; the iPhone Duo has two and no zoom.
  • The last minute is a merch segment (a desk pad), which a summary should drop.

How do you write the prompt?

  • Use the handle. "@mkbhd" is enough. The agent does not need the channel id.
  • Say how long. "5 points" or "one paragraph" keeps the summary short.
  • Ask for timestamps. Captions come with times, so "with timestamps" lets you jump to the part you want.

What can go wrong?

  • Sponsor segments end up in the summary. In our run the last minute of the video promoted a desk pad. Add "skip ads and sponsors" to the prompt.
  • Words glued together. The timed caption lines sometimes join words across a line break ("afolding"). The same response also carries a clean full-text version; a good agent uses that one.
  • No captions, no summary. A video with captions turned off returns nothing to read. Ask the agent to say so instead of guessing from the title.

How does this compare with other ways to do it?

OptionHow it worksCostBest for
ClawLink + your agentOne prompt, the agent picks the video and summarizes the captions$0.00482 per runFollowing channels from the agent you already use
Paste the link into a chatbotCopy the URL by hand each timeFree to paidOne video now and then
YouTube Data APIGoogle Cloud project and key; captions need the owner's permissionFree quotaYour own channel

Frequently asked questions

Do I need a YouTube account?

No. This job reads public YouTube data through ClawLink's own data provider, so there is nothing to connect. It runs on your ClawLink credit; every new account gets $0.50 of free credit on its first data call.

Does it work in other languages?

Yes, when the video has captions in that language. Our test video had 7 caption tracks; the tool takes a language code.

Can it summarize a video by URL instead?

Yes. Give the agent the link and it skips the channel step, so the run costs $0.00282.

Can it read the comments too?

Yes, with a third call. See the use case on what viewers ask for.

Does it work for Shorts?

Yes, captions work for Shorts URLs too.

Keep reading

Related articles

All use cases →