Dan Vibe Codes

Vibe coding apps and automating distribution

[Dan Vibe Codes @DanVibeCode]

@katarinaore Keep going! I asked Claude cowork to check RevenueCat and ASC using the browser extension and it worked its magic!

[Dan Vibe Codes @DanVibeCode]

@katarinaore Figma MCP + Claude worked great for me. Good luck!

[Dan Vibe Codes @DanVibeCode]

If I had to give one piece of advice to a founder about to start their app's visual content: build the pipeline before you generate the first asset. It feels slow for 3 hours. It compounds for 3 years.

[Dan Vibe Codes @DanVibeCode]

The indie app marketers who win the next 2 years won't be the ones with the flashiest prompts. They'll be the ones who schematise content, build re-runnable pipelines, and ship libraries instead of galleries.

[Dan Vibe Codes @DanVibeCode]

AI lets a solo founder ship a studio's worth of brand content. But only if you stop using AI like a chat app and start treating it like a microservice. Different mental model, different outcome.

[Dan Vibe Codes @DanVibeCode]

Reference images for your character. Templates for your prompts. JSON for your scene list. A retrying async pipeline. A CSV log. That's the whole toolkit. Every indie app should have this in their repo. Most don't.

[Dan Vibe Codes @DanVibeCode]

The unglamorous truth of AI for app marketing: the wins come from plumbing, not models. Retry logic, resume, idempotent writes, structured prompts. Boring engineering is what separates a demo from a production content machine.

[Dan Vibe Codes @DanVibeCode]

Useful heuristic: if you're generating 5+ of something, schematise it. Scene prompts, post drafts, product descriptions, ad variations. 5+ means JSON. 5+ means pipeline. Anything less is still a one-off.

[Dan Vibe Codes @DanVibeCode]

Stop thinking about AI image tools as "make a picture" buttons. Start thinking about them as APIs you wrap in a pipeline. First framing caps you at one-offs. Second framing turns content into infrastructure.

[Dan Vibe Codes @DanVibeCode]

Every indie app I see struggling with marketing has the same bottleneck: not ideas, not budget, not ads. Every visual asset is hand-assembled. The minute that's true, volume is capped by your energy.

[Dan Vibe Codes @DanVibeCode]

The highest-leverage asset in indie app marketing right now isn't a growth tactic. It's a repeatable content pipeline. Tactics change weekly. A well-built pipeline is still shipping a year later with minor edits.

[Dan Vibe Codes @DanVibeCode]

App marketing at solo-founder volume is an exercise in compounding leverage. Either you build systems that turn one hour of work into a month of content, or you don't, and you stall out at image #30 like everyone else.

[Dan Vibe Codes @DanVibeCode]

Every product I've shipped in the last year uses a variant of the same pipeline. The JSON changes, the references change, the Python barely changes. The tool is general. The content is specific. That's the point.

[Dan Vibe Codes @DanVibeCode]

The mascot I generated for DailyDog has shipped in 400+ images across the app, marketing site, App Store, and social. That single PNG has done more brand work than any other file in the project. Back up your references.

[Dan Vibe Codes @DanVibeCode]

Lesson from Snaglist: a weekly content pipeline isn't a marketing tactic. It's infrastructure. Once the JSON is schematised, the team's conversation shifts from "what should we post" to "what's the next topic list?"

[Dan Vibe Codes @DanVibeCode]

Lesson from SafeBowl: the moment the same product hero shipped across Instagram + App Store + onboarding, recognition compounded. Users could tell it was the same app at a glance. That's what brand actually buys you.

[Dan Vibe Codes @DanVibeCode]

The honest numbers on DailyDog: 134 images, $15 API spend, 89 min runtime, 4 transient failures auto-retried to success, 0 images thrown away for drift. The consistency came from the reference PNGs, not from cleverness.

[Dan Vibe Codes @DanVibeCode]

Common pattern across DailyDog, SafeBowl, Snaglist: references + style anchor + JSON + ~200 lines of Python. The surface changes. The architecture doesn't. Build it once, port it across every product you ever ship.

[Dan Vibe Codes @DanVibeCode]

Snaglist, a TikTok-heavy product, runs a weekly content pipeline. One topic list in, 15 on-brand vertical images per week out. Every week. The content machine stopped being a human bottleneck 6 weeks after I shipped it.

[Dan Vibe Codes @DanVibeCode]

SafeBowl (a safety-focused kitchen app) needed 30 carousel slides for Instagram with a consistent product hero. Same pipeline, different JSON. Ran in under 20 minutes. The ops cost of batch two was basically zero.