Law Major Ships iPhone App in 13 Days With Claude Code
By @sharedot · · 6 pages
- Programming
- AI Coding Tools
- Vibe Coding
- Claude Code
- Ios Development
A law graduate with no native-app experience shipped Tsuranari Calendar to the App Store in 13 days using Claude Code, submitting in about 73 hours.
A first native app, on the App Store in 13 days
shattered.io reports that a developer whose background is in law, not computer science, shipped a native iPhone app called Tsuranari Calendar to Apple's App Store in 13 days, with Anthropic's Claude Code doing much of the work.
According to reporting by BigGo Finance and TechRadar, development began on October 23, the reported launch date of the iPhone Duo, and the path from concept to first submission took roughly 73 hours.
The developer's prior programming experience came from studying for Japan's Fundamental Information Technology Engineer Examination, and they had published web apps before but had never built a native app. The app cleared Apple's review and went live on the 13th day after the device's launch.
Vibe coding: judge the app, not the code
The reported method was vibe coding, defined in the coverage as having the human avoid reading the generated code.
Instead of debugging line by line, the developer acted as product owner and tester: writing instructions, installing builds on a device, tapping through screens, and reporting what looked wrong while the AI revised its own work.
shattered.io notes this is why a device-running version existed by the night after the concept began. The human still made every product decision — what to build, what counted as done, and when to submit — and handled the Apple review response.
The honest reading, the site argues, is not that AI replaced the developer, but that it shrank the skill barrier between an idea and a working native app.
One review hiccup and an honest fact check
The most instructive detail was the smallest: before resubmitting, the developer changed a button label from "Allow Access" to "Continue," and approval came shortly after midnight on the 13th day.
shattered.io notes that Apple's review scrutinizes how apps ask for permissions, and that most of the 13-day calendar belonged to Apple's approval process, not coding.
The site also separates confirmed facts from viral claims: reports support extensive use of Claude Code, not the claim that every line was AI-generated, and no on-record quote from the developer or Anthropic exists.
Claude Code is included with Anthropic's paid plans, listed at $20 per month for Pro and $100 per month for Max at the time of reporting.
The stakes: fast shipping, exposed agents
The story lands the same week security teams learned how much damage an agent-runtime can do when it is reachable.
Help Net Security reports that Darktrace researchers turned Claude Code into a remote-controlled implant in a lab test, steering the coding agent through the memory stores feature of Anthropic's Claude platform and pulling a 1 GB file out through the Claude Files API using stolen API keys.
Help Net Security also reports a campaign in which a threat actor using stolen GitHub credentials pushed a credential-stealing payload into tens of thousands of repositories since November 10, with malware running in more than 1,800 of them.
Speed-to-ship is now real; so is the attack surface that comes with it.