Tier 01 · Chrome Extension · Source Code Included

One tab.
Every format.
Every platform.

Not a repurposing tool. A content engine that turns any link or idea into a published video that captures leads in your DMs. Paste a YouTube video, Instagram Reel, TikTok, or article, or type an idea. The Content Engine writes it in your voice, clones your audio, builds your talking head, cleans up a take you filmed yourself, or goes faceless, draws the B-roll in code, publishes to 11 accounts across 9 platforms, and arms your IG and FB DM automation.

Get the Engine $297 $147 once See what it does ↓
Claude Code or OpenRouter ElevenLabs V3 HeyGen Avatar IV or V Remotion Blotato or Zernio Airtable Claude Code
5
Source types
11
Accounts posted
9
Platforms
4
Avatar modes
2
Approval gates

Publishes to 11 accounts across 9 platforms

Pinterest
Instagram
Threads
YouTube
TikTok
Bluesky
X (Twitter)
LinkedIn
Facebook
Watch It Work

One link in.
Watch the whole run.

About three minutes, start to finish. It goes from a post open in your browser tab to a finished short, one step at a time.

The Real Problem

You already made
the content.
Then it died.

You watched a video that said exactly what your audience needs to hear. You had a better take. And it stayed in your head, because turning one idea into a script, a voiceover, a video, nine captions, and a follow-up takes a full afternoon you do not have.

So you post to two platforms, skip the other seven, and nobody who comments ever hears from you again. That is the gap the Content Engine closes.

Doing it by hand
3h+
Script, record, edit, caption nine times, post nine times, reply to comments. Per piece.
With the Engine
1 click
Then two quick approvals: the script, then the video with its captions. The rest runs in the background.
Leads captured per post
Auto
Every video ships with a comment keyword and the DM that answers it, armed automatically.
Real Output

One script in.
Here is what comes out.

A few samples of what the system makes, straight from the render folder. Same pipeline, different looks. Tap any one to watch the full video.

Talking avatar + B-roll
A HeyGen avatar of me · B-roll drawn in code
AI avatar · B-rollBasic
AI avatar · B-rollSplit Card + Web
AI avatar · B-rollLetterpress Split
AI avatar · B-rollChalkboard Split
Faceless, no camera
No face on screen · the whole frame comes from the script
+ Your voiceover on real runs
FacelessPapercraftShown silent here. On your runs it adds your voiceover, the same cloned voice as every other mode.
+ Your voiceover on real runs
FacelessLetterpressShown silent here. On your runs it adds your voiceover, the same cloned voice as every other mode.
FacelessStick Figure
+ Your voiceover on real runs
FacelessChalkboard

More B-roll styles are on the way. Every order gets free Engine updates, so each new style lands in your list to pick from.

My Own Numbers

My Facebook page
sat dead for 3+ years.

Then I started posting educational reels, and kept posting them. These are my own dashboards, not a mock-up. The video walks through them.

Facebook · 28 days
680,218
Views, up 473% on the period before.
Facebook · 28 days
6,021
Follows, up 836% on the period before.
Instagram · 30 days
142,894
Views, 99.8% of them from Reels.
TikTok · 60 days
33.4K
Video views, up 106.7% on the period before.
Instagram account insights, last 30 days: 142,894 views, 99.8% from Reels
Instagram · last 30 days
TikTok analytics overview, last 60 days: 33.4K video views, up 106.7%
TikTok · last 60 days
Get the Engine · $147 →
Any Link In

Whatever tab you have open
is the starting point.

The extension detects the page you are on. No copying transcripts, no downloading files. If the source has no transcript, it transcribes the audio itself. For Reels and TikToks it keeps the hook that already proved it works and rewrites the rest as you. No page open? Type the idea and it starts from that.

▶
YouTube
Long videos, Shorts, youtu.be links. Transcript pulled directly.
📸
Instagram Reel
Reels and posts. Audio transcribed if no captions exist. Remix mode keeps the proven hook.
🎵
TikTok
Any TikTok URL, including short links. Same remix treatment.
📰
Any article
Blog posts, docs, newsletters. Reads the page text straight from the tab.
💡
An idea
Type a sentence or two into the popup. No source page needed. The script is written from your brief.
What Happens After One Click

Eight steps.
You approve two.

Nothing that costs money runs before you have read the script. The pipeline lives in the extension's background worker, so you can close the popup, keep browsing, and come back when the badge says it is ready. Manual and Record modes pause once more in between, to take the file you filmed. If a step fails, you resume from that step. It never re-posts something it already posted.

  • Script length45 to 55 sec
  • Caption sets10 platforms
  • Output1080×1920 · 30fps
  • Paid steps before approvalZero
  • Double postsBlocked
1
Read the source
Transcript, scraped caption, page text, or its own audio transcription as a fallback
2
Write the script in your voice brand.config
Hook, value, one-word CTA keyword. Your brand voice and audience live in one config file you edit once. Claude writes it on your Claude Code subscription, or through OpenRouter if you would rather pay per call
✋
Gate 1 · You read and edit the script
Inline editor with live word count. Your edits flow into the voice, the captions, and the on-screen text. Close the popup and it keeps working in the background. Press Stop at any point and resume later
3
Ten caption sets + Airtable row
TikTok, Instagram, YouTube, Facebook, LinkedIn, X, Threads, Pinterest, Bluesky, Substack Note. Character limits enforced per platform
4
Your cloned voice ElevenLabs V3
Locked anti-drift settings so it sounds like you on take 40, not just take 1. Or let HeyGen speak the script with its own voice. Skipped entirely in Record it myself: the audio off your own take is the voiceover
5
Your talking head HeyGen Avatar IV or V
Your own HeyGen avatar, rendered on your HeyGen plan credits. Pick Avatar IV for standard credits or Avatar V for the most lifelike look at about twice the credits, and set the shape you filmed it in. Lip-synced to your cloned voice, not a stock TTS. Or be the talking head: Record it myself pauses the run, takes the video you filmed, and cleans it. Manual mode takes a HeyGen Studio render. Faceless skips the head entirely
6
B-roll drawn in code Remotion + canvas
Basic infographics, Split Card + Web, Papercraft Split, Letterpress Split, Chalkboard Split, Voxel Diorama Split, or a style you add yourself. Word-highlighted captions in your brand colors. Title baked into frame 0 so the thumbnail is never black. Faceless mode builds the whole visual instead
✋
Gate 2 · You watch the video, tweak captions
One editable caption box per platform. Do not like the B-roll or the avatar? Re-render just that part and the script and voice stay as they are. What you approve is what posts
7
Publish to 11 accounts Blotato or Zernio
Post now, drop into the next free slot per platform, or save only. Two Instagram and two Threads accounts supported
8
Arm the DM automation
Comment keyword → freebie DM on Instagram and your Facebook page, created automatically per video on whichever scheduler posted it
Where Repurposing Tools Stop

A view is not a lead.
So every video comes with a way in.

Most tools hand you a video and wish you luck. This one writes the CTA keyword into the script and sets up the DM that answers it, on the same run that posts the video. You get the whole loop from one link.

01 · In the script
A keyword, not a "link in bio"
Every script ends on a one-word comment CTA unique to that video. Unique keywords mean the automation can safely apply across your whole account.
02 · In the DMs
The reply that runs while you sleep
Instagram (both accounts) and your Facebook page. The DM half lives in Blotato's cloud, 24/7. A local poller also posts a public "sent it to your DMs" reply so the comment thread keeps working for you.
Auto DMHey, thanks for commenting REPURPOSE! Here's your freebie: hub.digicuratoragency.com/freebie
03 · Your scheduler
Blotato or Zernio, your pick
Choose the scheduler per run from the popup. Both post to your accounts and both create the keyword DM automation, so the loop works whichever one you pay for. Add the key for either and the option lights up.
6 B-roll Styles + Faceless Mode

Same script.
Six looks, or no face at all.

No stock footage. Every B-roll style is drawn in code from a scene plan Claude writes for that script, so the visuals match what you are saying. Split Card + Web records real websites scrolling in a headless browser and times each site's appearance to the exact word you say it. Papercraft, Letterpress, Chalkboard and Voxel Diorama fill the top half with paper models, wood-type proofs, a chalk lesson or a toy island built from falling blocks and give the whole frame to your head on the key lines. Faceless mode drops the talking head and builds the entire video from the script, in one of seven tracks. Any style you add yourself shows up in the same list. New styles I build are added to that list in the free updates. Every B-roll style sits over whichever avatar you picked, including a video you filmed yourself.

01 · B-roll
Basic
Animated infographics over the talking head. Fastest render.
02 · B-roll
Split Card + Web
Clean UI cards and full-screen tours of up to five sites. Cream or dark glass theme. Drop in your own clip.
03 · B-roll
Papercraft Split
Printed paper models fold themselves up on a green cutting mat above your head. Paper cards ask for the keyword at the end.
04 · B-roll
Letterpress Split
A fresh wood-type proof pulled on a press for every beat, with rubber stamps and press sounds.
05 · B-roll
Chalkboard Split
A chalk lesson written line by line on a green slate above your head, with the camera sliding along the board.
06 · B-roll
Voxel Diorama Split
A toy island built from falling blocks above your head.
07 · Faceless
TSX Animation
Code-drawn motion graphics in Dark editorial or Chess. Your cloned voice is the only cost.
08 · Faceless
Vox Collage
Paper-collage documentary look. AI image layers, a traveling camera, and a music bed if you want one.
09 · Faceless
Papercraft
The Papercraft look filling the whole 9:16 frame. Builds in about 5 to 10 minutes.
10 · Faceless
Letterpress
The Letterpress look filling the whole frame. Same 5 to 10 minute build.
11 · Faceless
Chalkboard
The chalk lesson filling the whole 9:16 frame, written on as you speak.
12 · Faceless
Voxel Diorama
The toy island built from falling blocks, filling the whole 9:16 frame.
13 · Faceless
Stick Figure
A stick-figure drawing on beige paper for every line, each with a slow camera move. Images from Kie.ai or Google Flow.
+
Yours
Add your own style and it shows up in the B-roll list. You own the source, so Claude Code can build one with you. Every new style I add comes to you free in an update.
Runs On Your Machine

Your keys.
Your accounts.
Your code.

Nothing routes through my servers. The backend runs locally in Node, the extension talks to it on localhost, and every API is billed to your own account at cost. There is no per-video fee and no monthly plan from me.

Airtable is the one key you need to start, plus a way to write scripts: your Claude Code subscription, or an OpenRouter key if you would rather pay per call. ElevenLabs is the next one only if you want the cloned voice, which is also what Faceless mode narrates with. Record it myself needs neither. HeyGen, Blotato or Zernio, Apify, Kie.ai, and Gemini unlock the rest as you connect them. A missing key skips that one step instead of breaking the run.

  • Software fee$147 once (regular price $297)
  • Monthly to me$0
  • Required keys to start2
  • Avatar cost, record modeNo voice or avatar spend
  • Avatar cost, manual modeStudio credits
  • Avatar cost, automatic mode~$1 / min
  • Faceless mode, TSX trackVoice cost only
  • Source codeYours
Everything Included

Your $147 gets
all of this.

📁
Full Source Code
The Chrome extension, the Node backend, and every Remotion composition, as a ZIP. Not a template. The same project I run daily.
🔑
License Key
Activates the extension on your install. Personal use, yours to keep.
📘
9-Section Setup Guide
Clone, one-command install, every API key explained, Airtable base, load the extension, license, first video, brand customization, troubleshooting table.
🎨
brand.config.js
Your voice, your audience, your caption colors. Edit one file and every script and video changes with it.
🗂️
Airtable Base Spec
Every video, script, caption set, and posted URL logged in one table you can build on.
Record It Myself

Read the script yourself.
No voice bill. No avatar bill.

The fourth avatar option, and the cheapest way to run the Engine. Approve the script, then read it into your phone or your webcam and upload the file. The Engine transcribes it on your own machine, cuts the filler words, and shortens every long pause. The cleaned take becomes your avatar and its audio becomes your voiceover, so captions, B-roll, Airtable, the review screen, DM automation, and posting all run exactly as they do in every other mode.

It calls no ElevenLabs and no HeyGen. Claude still writes the script, on your Claude Code subscription or through OpenRouter for cents, and that is the only per-video cost.

  • ElevenLabs billNone in this mode
  • HeyGen billNone in this mode
  • Kie.ai creditsNone in this mode
  • Cost per videoThe Claude script: your plan, or cents on OpenRouter
  • Keys to startAirtable + Claude Code or OpenRouter
  • To publishBlotato or Zernio
  • TranscriberLocal · free · installed for you
01 · You read it
The run waits for your take
After the script gate the pipeline pauses and shows you the clean version to read. Film it portrait on your phone or your webcam. If you stumble, stop and say the line again. MP4, MOV, or WEBM.
02 · It cleans the take
The um and the four-second pause, gone
A local transcription builds a cut list, filler words come out, and any pause over 600ms drops to 200ms. One pass cuts picture and sound together and re-encodes to h264 at a constant frame rate, so iPhone files behave. The review screen tells you exactly what it removed.
03 · Everything else is normal
Same B-roll, same DMs, same posting
Your cleaned video is the avatar and its audio is the voiceover, so every B-roll style, the ten caption sets, the Airtable row, the comment-to-DM automation, and all 11 accounts run unchanged. On-screen captions follow what you actually said, not what the script said.
Instant delivery · Source code included
AI Social Hub Content Engine

Any link to
9 platforms.

$297
$147
One-time · No subscription
Get the Engine · $147 →
  • SourcesYouTube · IG · TikTok · Articles · Ideas
  • Accounts11 across 9 platforms
  • SchedulerBlotato or Zernio
  • B-roll stylesBasic · Split Card · Papercraft · Letterpress · Chalkboard · Voxel Diorama · more in free updates
  • Avatar modesAuto · Manual · Record · Faceless
  • DM automationInstagram + Facebook
  • Source code + licenseYours
  • Engine updatesFree for life
  • Coding requiredZero to run
  • Runs onMac (macOS 15+) · Windows 10/11 (64-bit) · 8 GB RAM · 10 GB free
Before You Buy

Straight answers.

Do I have to be on camera?
No. Faceless mode builds the whole video from your approved script with no face and no camera. Automatic mode puts a HeyGen avatar of you on screen, lip-synced to your cloned voice. Manual mode renders that same avatar in HeyGen Studio on plan credits and you upload the MP4 back into the run. And if you would rather be on camera, Record it myself takes a video you filmed on your phone and cleans it up for you.
Will it run on my Mac or Windows laptop?
Yes, on both. Each one has its own one-command setup, and you get the same course and the same free updates. Mac: macOS 15 Sequoia or newer, on Apple Silicon or Intel. Windows: Windows 11, or Windows 10 version 22H2, 64-bit on an Intel or AMD processor. Windows on ARM (some Surface and Snapdragon laptops) is not supported yet. Either one: 8 GB of memory or more (16 GB renders smoother), about 10 GB of free space, an administrator account, and an internet connection. The setup script installs everything else for you, including Google Chrome, the Python and Playwright that Split Card + Web, Papercraft and Letterpress use, the Claude Code CLI for the TSX and Vox faceless tracks, and whisper.cpp for Record it myself. Nothing you record leaves your machine. Not sure about your machine? Email ness@digicuratoragency.com before you buy and I'll check it with you.
Do I need to code?
Not to run it. You fill in API keys, edit one brand config file, and load the extension. Changing how it works is where Claude Code comes in, and because you own the source, you can.
What does a video actually cost me?
Your own API usage. Claude writes the script on your Claude Code subscription, or through OpenRouter for cents, and that is the one cost every mode shares. ElevenLabs is cents on top whenever you use the cloned voice. HeyGen automatic mode runs on your own HeyGen plan credits, and Avatar V uses about twice the credits of Avatar IV. Record it myself calls neither of them, so there is no voice or avatar spend on that run at all. Faceless mode on the TSX track adds nothing beyond the voice. Blotato or Zernio and Airtable are your own plans. I charge nothing per video.
Will it post the same video twice if something fails?
No. Every successful post is written to a ledger the moment it lands. Resuming a failed run reports "already posted" for those targets and only fills the gaps.
What is your refund policy?
All Content Engine purchases are final and non-refundable. You get the software, the source code, and the downloadable files the moment you buy, and they can't be taken back. If you aren't sure Content Engine is right for you, email ness@digicuratoragency.com before you purchase and I'll answer your questions so you can buy with confidence.
What is Faceless mode?
Pick Faceless instead of an avatar and the Engine skips the voice-to-head steps and builds the full video from your approved script. Seven tracks: TSX Animation (code-drawn in Dark editorial or Chess, your voice is the only cost), Vox Collage (paper-collage documentary look), Papercraft, Letterpress, Chalkboard and Voxel Diorama (those four looks filling the whole frame), and Stick Figure (a drawing for every line). TSX and Vox are built by a Claude agent, so they need the Claude Code CLI installed. Vox and Stick Figure draw their images on Kie.ai, with a spend cap per build, or on Google Flow.
What is Record it myself?
The fourth avatar option. After you approve the script the run pauses and shows you the clean version to read. Film yourself portrait on your phone or your webcam, upload the MP4, MOV, or WEBM, and the Engine transcribes it on your own machine, cuts the filler words, and shortens any pause over 600ms to 200ms. One ffmpeg pass cuts picture and sound together and re-encodes to h264 at a constant frame rate, so iPhone files work. The cleaned take becomes your avatar and its audio becomes your voiceover, so captions, B-roll, Airtable, the review, DM automation, and posting all run the same, with every B-roll style. On-screen captions follow what you actually said. It calls no ElevenLabs and no HeyGen. It does need whisper.cpp, a free local transcriber the setup script installs, and the popup tells you if it is missing before you start a run.
Blotato or Zernio?
Either. Add the key for the one you use and pick it per run in the popup. Both post to your accounts and both create the comment-to-DM automation. Zernio comes with a free tier for a couple of accounts if you want to start without a paid scheduler.
Which platforms?
TikTok, Instagram (two accounts), YouTube Shorts, Facebook, LinkedIn, X, Threads (two accounts), Pinterest, and Bluesky. A Substack Note caption is written and saved to Airtable for you to post by hand.