Big Idea

The argument in one line.

A video's visual style can be reverse-engineered with Claude into a reusable master prompt that turns any script into on-style AI-generated motion graphics, replacing manual design work with a repeatable text-to-video system.

Who This Is For

Read if. Skip if.

READ IF YOU ARE…

You make faceless or motivational content and want AI-generated motion graphics without hiring a designer or learning After Effects.
You already use Claude, ChatGPT, or similar tools and want a repeatable prompt system instead of one-off AI generations.
You run a content pipeline that needs the same visual look across dozens of videos (shorts, ads, educational explainers) without rebuilding the style every time.

SKIP IF…

You're looking for a free course on motion graphics fundamentals rather than an AI prompt workflow.
You need broadcast-quality, pixel-precise animation. AI-generated clips here are style-matched, not frame-perfect recreations.

TL;DR

The full version, fast.

ReachMora shows how to turn any existing motion graphic into a reusable AI production system. Find a reference clip in a style that fits your content, then feed it to Claude along with a prompt that breaks down its motion, color, typography, and camera language. Claude converts that analysis into a universal master prompt: feed it any script and it splits the script into visual beats, generating a short AI video prompt for each one. Those prompts get rendered in Google Flow (including a batch agent mode), narrated with an ElevenLabs voiceover, and assembled in CapCut by syncing visuals to the cleaned-up narration track. The result is a style you can reuse across unlimited future videos instead of rebuilding it each time.

Free for members

Chat with this breakdown — free.

Sign in and you get 23 free chat messages on us — ask for the hook, quote a framework, find the exact transcript moment, generate a markdown action plan. Bring your own key when you want unlimited.

Create a free account →

Chapters

Where the time goes.

00:00 – 00:58

01 · Hook: skip the template, skip the course

Cold open: a promise to create professional motion graphics with AI in minutes, explicitly not a template swap or a paid prompt pack, with the full method free in the description.

00:58 – 01:49

02 · Step 1: find a style to reverse-engineer

Search Pinterest (or any reference source) for a motion graphic style that matches the content type you want to make, then save or screen-record it.

01:49 – 03:40

03 · Step 2: reverse-engineer the style with Claude

Upload the reference clip to Claude along with a detailed visual-breakdown prompt. Claude analyzes motion, transitions, color, typography, composition, camera movement, and pacing, then outputs a generation prompt. A quick test in Google Flow's OmniFlash model confirms the style transfers, but a single clip isn't the goal.

03:40 – 04:39

04 · Step 3: turn it into a universal master prompt

Ask Claude to convert the one-off style prompt into a universal master prompt that accepts any script, breaks it into visual beats, and outputs a short AI video prompt per beat.

04:39 – 05:27

05 · Step 4: test the master prompt on a new script

Feed the master prompt a completely different, AI-generated test script. Claude splits it into visual sections and generates beat-by-beat prompts that still follow the original reference style.

05:27 – 06:27

06 · Step 5: generate the clips in Google Flow

Render the beat prompts in Google Flow one at a time, or use Flow's agent mode to submit the entire prompt list at once, with an explicit no-music, no-voiceover instruction.

06:27 – 06:55

07 · Step 6: generate the voiceover in ElevenLabs

Paste the original script into ElevenLabs and generate the voiceover separately from the visuals.

06:55 – 08:17

08 · Step 7: assemble and sync everything in CapCut

In CapCut: drop in the voiceover first, trim silent gaps, merge it into one track, then drag in the generated clips and match each one to its corresponding section of the narration before final export.

08:17 – 09:17

09 · Recap: the reusable system, and what's next

Recaps the full loop (find a style, reverse-engineer it, build a master prompt, feed it any script) and closes by asking viewers which AI workflow to break down next.

Atomic Insights

Lines worth screenshotting.

A video's visual style can be reverse-engineered into a text prompt that recreates its motion, color, typography, and camera language without copying the clip itself.
The deliverable of AI style analysis is a generation-ready prompt, not a written description of what the video depicts.
A one-off AI clip that nails a target style is a proof of concept, not a finished system.
Turning a single generated result into a universal master prompt is what makes an AI visual workflow reusable across future projects.
A master prompt can take any script, split it into visual beats, and generate a short AI video prompt for each beat automatically.
Testing a master prompt on an unrelated script is the real test of whether an AI style actually generalizes.
Batch or agent modes in AI video tools can generate an entire list of visual prompts in one pass instead of one clip at a time.
Explicitly telling an AI video agent to skip music and voiceover keeps those elements editable later instead of baked into the generated clips.
Trimming silence from a voiceover track before syncing visuals keeps pacing tied to narration instead of an arbitrary cut rhythm.
The entire workflow (find a style, reverse-engineer it, build a master prompt, feed it any script) turns one video analysis into a repeatable production system.

Takeaway

Turn one video's style into an endless prompt system

AI Workflow System

A single reference video's visual style can be decomposed and rebuilt into a master prompt that turns any script into on-style AI video clips, replacing per-project design work with a repeatable pipeline.

02Step 1: find a style to reverse-engineer

Start from a reference clip instead of a blank page: find one existing motion graphic whose look already matches your content type, then use it as the AI's style target.

03Step 2: reverse-engineer the style with Claude

An AI model can decompose a video's visual DNA (motion, transitions, color, typography, composition, camera movement, pacing, hierarchy) into a written prompt instead of just describing its plot.
The deliverable from analyzing a reference clip is a generation prompt built to recreate the look, not a summary of what happens on screen.
Generating one clip that matches a reference style proves the technique works, but a single clip is a demo, not a reusable system.

04Step 3: turn it into a universal master prompt

A style, once reverse-engineered, can be converted into a master prompt that accepts any script and still outputs visuals in that same style.
The reusable version of a workflow comes from explicitly asking the AI to turn a one-off result into a template, rather than reusing the original prompt.
A master prompt should instruct the model to break any input script into visual beats and generate a short, roughly 10-second prompt for each beat.

05Step 4: test the master prompt on a new script

Testing a master prompt against a completely unrelated script is what proves the system generalizes rather than memorizing one reference video.

06Step 5: generate the clips in Google Flow

Once visual prompts exist for every beat, clips can be generated one at a time or batched through an agent mode that processes a whole list automatically.
Explicitly telling a generation agent to skip music and voiceover on the clips keeps those elements under separate, controllable editing in a later step.

07Step 6: generate the voiceover in ElevenLabs

Voiceover generation is decoupled from visual generation entirely. The same script that seeded the visual prompts gets fed to a separate voice tool.

08Step 7: assemble and sync everything in CapCut

Editing starts with the voiceover track alone: trim the silent gaps first, then merge the cleaned clips into one continuous track before matching visuals to it beat by beat.

09Recap: the reusable system, and what's next

The reusable unit in this workflow is the master prompt itself, not any single generated video. Once built, it gets reused across unlimited future scripts and topics.

Glossary

Terms worth knowing.

Google Flow: Google's AI video generation tool where text prompts are turned into short video clips, including an agent mode that can batch-process multiple prompts at once.
Omni (OmniFlash) model: The specific AI video generation model inside Google Flow used in this workflow to turn style-matched prompts into clips.
Master prompt: A single reusable AI prompt, built from a reverse-engineered visual style, that converts any new script into a full set of on-style video generation prompts.
Visual DNA: The combination of motion, transitions, color, typography, composition, camera movement, and pacing that makes a specific animation style recognizable.
Agent mode: A mode inside an AI tool that can work through a list of prompts on its own instead of requiring one submission per clip.

Resources

Things they pointed at.

00:00toolClaude

01:20toolPinterest

03:02toolGoogle Flow (OmniFlash model)

04:39toolChatGPT

06:27toolElevenLabs

06:55toolCapCut

00:58linkFree master prompt document (Google Drive) ↗

Quotables

Lines you could clip.

00:00

“I'm not talking about downloading some random template and changing the text.”

anti-hype hook that sets up the real promise→ TikTok hook↗ Tweet quote

05:30

“We don't want to recreate one video. We want to create a system.”

tight thesis statement, works standalone→ IG reel cold open↗ Tweet quote

07:53

“Confidence doesn't come before action. It comes from action.”

motivational punchline from the demo script itself, quotable out of context→ newsletter pull-quote↗ Tweet quote

The Script

Word for word.

Read-along

Don't just watch it. Burn it in.

See every word as it's spoken — crank it to 2× and still catch all of it. The same dual-channel trick behind Amazon's Kindle + Audible.

metaphor

Today, I'm going to show you how to create professional -looking motion graphics in minutes using AI without spending a single dollar. And no, I'm not talking about downloading some random template and changing the text. I'm talking about taking a perfect visual style reference, reverse engineering the visual language behind it, and then turning that style into a reusable AI template you can use over and over again.

Because here's the crazy part. A motion graphic that could take a designer hours or even days to build manually can now be broken down, analyzed, and recreated with AI in just a few minutes.

And once you understand this workflow, you can use it for motivational videos, documentaries, educational content, YouTube shorts, social media ads, or even freelance projects for clients. And I'm going to give you the exact prompt I use to make this happen. For free, no course, no $97 prompt pack, no DM me on Instagram to get access.

Everything you need is in the free document linked in the description. my job is to give you the stuff internet gurus are trying to sell you as a course your job is to hit that like button subscribe if you're new and actually use it and make sure you watch all the way to the end because the final cap cut step is where we turn all these separate ai generated clips into an actual finished video so let's build this step one find a motion graphic style first we need to find the style we want to recreate for this example i'm going to search for motion graphics and as you can see there are tons of different styles here personally i like using pinterest whenever i'm looking for design inspiration or just trying to get my brain out of the same boring visual ideas the goal here is simple find a style that matches the type of content you want to create for this tutorial i'm going to use this one i really like this style because i think it would work extremely well

for motivational content. First, hit share and copy the link. Now go to Google.

You can use any of the tools shown here. They are free, so just pick one that works for you. And once you've got the file ready, let's move to the fun part.

Step two, reverse engineer the style. Now go to Claude. If you've never used Claude before, don't worry.

It's basically another AI assistant similar to ChatGPT. Just sign in with your Google account and I'm using the free plan here so you don't need to spend money. Now upload the motion graphic video you downloaded.

But don't hit submit yet, because there's one thing we need first, the magic prompt. You'll find the exact prompt I'm using in the free document linked in the description. Copy that prompt, paste it into Claude, and now submit everything together.

Give Claude a moment to analyze the video because we're not simply asking it to tell us what happens in the video. We're asking it to break down the visual DNA of the entire animation, the motion, the transitions, the colors, the typography, the composition, the camera movement, the pacing, the visual hierarchy. Basically, everything that makes this particular animation look the way it does.

And Claude will analyze all of that and generate a prompt designed to recreate the style using the Omni. And you can use that model inside Google Flow. So let's test it.

Copy the generated prompt. Go over to Google Flow. Select the video tab.

Choose the OmniFlash model. Paste the prompt and hit generate. And boom, here's the result.

Now, is it interesting? Absolutely. But this is not what we actually want because generating one cool clip is not the goal.

We don't want to recreate one video. We want to create a system. We want something we can use again tomorrow, next week, or on our next 50 videos.

And that's where the next step comes in. Step three, turn the style into a universal master prompt. We need a template, a master prompt that can take any script or text and automatically turn it into visual prompts while keeping the same visual style we just reverse engineered.

So let's go back to Claude, and now we're going to ask it to transform the prompt it created into a reusable system. I told Claude, transform the prompt above into a universal master prompt that I can use with any text or script. The master prompt should analyze the input, break it down into meaningful visual beats, and generate 10 second visual prompts for each beat.

And this is what we get. Now we have something completely different. Instead of giving Claude one specific video and asking it to recreate that video, we now have a universal visual generation template.

Give it a completely different script and it knows how to break that script into visual beats while maintaining the style we just analyzed. This is the part that makes the entire workflow reusable. So let's actually test it.

Step four, test the master prompt. You can use literally any script you want. For this example, I asked ChatGPT to generate a one -minute motivational script.

Nothing complicated, just a simple script we can use as a test. Now I'm going to take that script, go back to Claude, paste the script together with our master prompt, and submit it. And now Claude starts breaking the script down into individual visual sections.

Instead of giving us one giant prompt, it generates specific visual prompts for each part of the story. And because our master prompt contains the visual style we reverse engineered earlier, all of these prompts are designed to follow that same visual language.

This is where the workflow starts getting really useful. Because now we have our script, we have our visual direction, and we have individual prompts ready to generate. So let's make the actual clips.

Step five. Generate the video clips with Google Flow. Now we're going back to Google Flow.

If you don't have enough credits to generate everything at once, don't worry. You can simply copy the specific visual prompt you need, paste it into Flow, and hit Generate. And look at this.

You can see that the result follows the same visual style we reverse engineered earlier. That's exactly what we're looking for. Now, if you have enough credits and you want to speed things up, there's another option.

Copy all of the visual prompts Claude generated. Activate agent mode inside Google flow select the omni flash model and type something like generate all the clips Do not use any music or voiceover then paste all of your visual prompts underneath it and submit flow can then work through the prompts and Generate the clips for you.

So instead of sitting there generating every single scene manually You can let the workflow handle a huge part of the process and once all the clips are generated We've got our visuals, but we're still missing one very important Step 6. Generate the voiceover.

For the voiceover, I'm using 11 Labs. But you can use whatever voice generation tool you prefer. Take your original script, paste it into 11 Labs, choose your voice, and generate the voiceover.

And now we have basically everything we need. We've got the script. We've got the AI -generated visuals.

We've got the voiceover. Now we just need to put everything together. Step 7.

Edit everything in CapCut. Open CapCut. and import all of your assets.

First, drag your voiceover onto the timeline. Listen through it and remove all of those awkward silent gaps. Once you've finished cleaning up the voiceover, select all of the voiceover clips and combine them into one using Alt plus G.

Now we have one clean voiceover track. Next, drag all of your generated video clips onto the timeline. Start matching each visual with the corresponding part of the voiceover.

You don't need to overcomplicate this. Just make sure the visuals change naturally with the narration, and that nothing feels like it's sitting on screen for way too long.

Once everything is synced, give the entire timeline one final watch. Check your pacing. Check your transitions.

Check your audio. And if everything looks good, export the video in a good resolution. And that's it.

You keep waiting for the right moment, when you have more money, more confidence, more time, when you finally feel ready. But here's the truth. You're probably never going to feel ready.

Confidence doesn't come before action. from action so stop waiting for motivation it doesn't have to be perfect you just have to keep moving because six months from now you can either be looking back at the same excuses or looking back at the version of you who finally decided enough I'm starting today your future isn't built in one massive decision it's built in the small decisions you make every single day so don't wait for tomorrow start now So the entire workflow is basically find the style, reverse engineer it with AI, turn it into a master prompt.

Feed that prompt any script. Generate the visuals. Add the voiceover.

Edit everything together. And the best part? Once you've created that master prompt, you don't have to start from zero every time.

You can reuse the system for completely different scripts, topics, and videos. And if you want me to break down another AI workflow like this one, comment below and tell me what you want to see next. Maybe it's another viral content system.

A completely free AI video workflow. A faceless YouTube automation system. Or a way to turn these AI -generated visuals into actual shorts that can blow up?

Drop your idea in the comments. And if you enjoyed this tutorial, hit like, subscribe to ReachMora, and turn on notifications. Because I'm constantly testing these tools so you don't have to.

I've got more crazy AI workflows coming. And trust me, you don't want to miss the next one. I'll see you there.

The Hook

The bait, then the rug-pull.

ReachMora opens with a direct anti-hype promise: no downloaded template, no $97 prompt pack, no DM-me-on-Instagram upsell. Just a repeatable way to reverse-engineer any motion graphic's style and turn it into a master prompt that gets reused on every future script.

Frameworks

Named ideas worth stealing.

00:58list

The Reverse-Engineer-to-Master-Prompt Workflow

Find a style to reverse-engineer
Reverse-engineer the style with Claude
Turn the style into a universal master prompt
Test the master prompt on a new script
Generate the video clips with Google Flow
Generate the voiceover with ElevenLabs
Edit everything together in CapCut

A seven-step pipeline that converts any single reference video's visual style into a reusable AI prompt system for generating on-style motion graphics from any script.

Steal forany faceless or motivational content pipeline that needs a consistent AI-generated visual style across many videos

CTA Breakdown