Drop in 5 to 10 carousels you admire. The skill extracts the style fingerprint once, then every future carousel runs in that style.
Works on Claude.ai, Cowork, Claude Code · $14 lifetime
Every carousel generator I tried built carousels that looked like carousels. Not like the carousels I actually saved to my camera roll.
So I built one that works the other way around. You give it 5 to 10 carousels you genuinely admire. It reads them with vision, extracts the structure, hooks, layout, colors, fonts, voice, decorations. Then every future carousel runs from that fingerprint.
I run my own carousels on it. So do my clients.
Kata, The Igniting Studio
You bring 5 to 10 carousels you admire. The skill reads them with vision, extracts the style fingerprint, and writes every future carousel against that fingerprint. No generic shells. Your output sounds like what you wanted, not like everyone else.
A built-in CSS and SVG library of 12 decoration recipes: crosshair marks, corner brackets, dot grids, brush-stroke underlines, number badges, pills, frames, ticks. The decorations that make editorial carousels editorial. The skill picks them based on what your inspiration uses.
I dropped in a set of carousels I actually liked. The skill pulled the fingerprint and wrote this one against it. Swipe through to see how the style carries from cover to CTA.
Swipe horizontally →
First run only. PNGs, screenshots, or Canva JSON. The skill reads them with vision and writes a full style fingerprint: structure, hooks, layout, palette, fonts, voice, decorations. Saved once. Never asked again.
One line. Plus format (educational, listicle, story, framework, comparison, or "you decide") and platform (Instagram, LinkedIn, or both). Four questions total.
Picks a hook pattern from your inspiration. Writes every slide to match the anatomy, layout, voice, and decoration usage of the set. Every slide is judged against the fingerprint before output.
Mode A is a branded HTML file with fonts embedded as base64. Open, click Save PNG per slide. Mode B is one full ChatGPT prompt per slide, tuned brutally specific. Each handles a different kind of inspiration. The skill recommends one based on the fingerprint.
Upload the PNGs to Instagram or LinkedIn in slide order. Caption ready to copy, hashtags researched in lowercase, posting time suggested. 5 to 10 minutes per Mode A carousel, 30 to 60 minutes per Mode B carousel.
A self-contained .html file with display and body fonts baked in as base64. No Google Fonts CDN, no internet needed. Works offline. The file renders identically on any machine. Click Save PNG per slide. Upload to Instagram or LinkedIn.
Best for: type-driven carousels that use decorative flourishes the library can draw (crosshair, dot grid, brush strokes, number badges, pills, frames).
For inspirations that lean on illustrations, photos, character renders, or screenshot mockups. The skill emits one full ChatGPT prompt for every slide in the carousel, brutally specific about aesthetic, palette, fonts, and decorations. ChatGPT generates each slide as a complete image. Canva is optional, only for tweaks.
Best for: inspirations with rich visuals (illustrated characters, real photos, 3D renders, UI mockups). Budget 30 to 60 minutes per carousel.
Carousel generators that ship with templates make everyone's output look the same. The skill works the other way around. You bring the carousels you admire. The skill matches them.
The fingerprint is the spec. Change the inspiration set, the output changes. Add new inspiration later, rebuild the fingerprint, the output updates. Your style stays yours.
Sold through Polar (Merchant of Record). Instant download after checkout. 14-day refund.
brand-overrides.md anytime to pin specific hex codes or font names that override what the fingerprint pulled.$14 for the skill. $197 for the full pack (this skill plus seven more). Either way, lifetime.
Questions? Email info@theignitingstudio.com