The skill that listens and watches for you

Marty VaughnAI Agents on YouTube

Just watched the video on transcribing videos with Claude? This is the skill from it. It installs in about a minute and you use it by asking, in plain English, for a video to be watched.

What it actually does

There are a lot of times you want one specific thing out of a video and do not want to sit through the whole thing to find it. This hands that job to Claude. It pulls the transcript, and where the words alone leave a gap, a slide, a chart, a demo, an as you can see here, it grabs that moment off the screen and reads it. Then you get one write-up of what was said and what was shown, which you can question, summarize, or pull apart.

You: Watch this YouTube video for me: <link>

Claude: Pulled the transcript. Two moments needed the screen,
        the pricing table at 4:12 and the diagram at 9:40.
        Here is what was said and what was shown.

What you can do with it

• Find the one part you came for. Ask for the section you care about instead of scrubbing the timeline looking for it.
• Read the slides, not just the words. It only looks at the screen when the audio genuinely leaves a gap, which is what stops it wasting your time and tokens grabbing every frame.
• Study what makes a video work. Point it at a creator you admire, or at two of them, and compare what they are actually doing side by side.

What it costs to run

The transcript is free, because captions are just text. Screenshots use a small amount of Claude's own vision, and the skill is deliberately sparing about taking them, so a normal talking-head video may need none at all. There is one paid fallback for the rare video with captions switched off, and it asks before it spends anything. You do not need to buy anything to use this.

Already subscribed? Enter your email anyway. You land on the same page with the same file, and you will not be added twice.