Rebuilding the Viral 8‑Yuan iOS App in 30 Minutes with AI Coding and Supabase

Why an 8‑Yuan App Became a Hit

The "Dead Yet?" app, priced at 8 CNY, topped the App Store paid chart in early 2026 by solving a simple yet painful problem for people living alone: a daily check‑in that triggers an emergency email if missed.

From Idea to Launch: AI‑Powered Full‑Stack Development

Using AI coding tools to generate the SwiftUI frontend and Supabase (referred to as ADB Supabase) for all backend services eliminates the need for a traditional server stack, databases, authentication, and cron jobs. The entire workflow—from UI to data storage and email notifications—can be built in under half an hour.

Tooling and Methodology

We selected Qoder as the AI coding IDE because it supports custom Rules and Skills that let the model break down requirements, plan architecture, and generate code incrementally. The BMAD‑Method (Rule‑based, Skills‑driven, Agent‑oriented Development) guides the AI through a "plan‑then‑implement" cycle, reducing context overload and improving code stability.

Technology Stack

Frontend: SwiftUI (declarative UI, type‑safe, AI‑friendly)

Backend: Supabase (hosted PostgreSQL, auth, Row‑Level Security, Edge Functions)

Edge Functions Runtime: Deno with native TypeScript

Supabase’s SDK exposes simple methods like .insert() and .select(), allowing the AI to generate concise data‑access code without writing REST endpoints.

Core Feature Breakdown

The app consists of a few modules:

Anonymous Authentication: On first launch the client calls Supabase Auth to create an anonymous user_id linked to the device.

Daily Check‑In: A button records the check‑in in the check_ins table; UI updates to show "Checked in today".

Missed Check‑In Detection: An Edge Function scans all users daily (cron expression 0 1 * * *) to find those who missed two consecutive days.

Email Notification: A second Edge Function sends an alert email to the emergency contact using Supabase’s built‑in email service.

Implementation Highlights

When the user signs in, the app stores the generated user_id locally and uses it for all subsequent queries. The frontend reads the latest check‑in status from Supabase, displaying either a success animation or prompting the user to check in.

Edge Function check‑missed‑check‑ins runs each night, queries the check_ins table, and triggers send‑notification‑email for any qualifying user. Both functions run in a serverless environment, require only the service role key (kept in cloud‑side environment variables), and are invoked via HTTP POST at /functions/v1/check-missed-check-ins and /functions/v1/send-notification-email.

Cron Job Configuration

In the Supabase dashboard a cron job named check‑missed‑check‑ins is created with the UTC schedule 0 1 * * * (9 AM Beijing time). The job calls the corresponding Edge Function using the service role key, bypassing Row‑Level Security for full‑table scans.

Key Principles for AI‑Assisted Development

Prefer Declarative Technologies: SwiftUI, SQL schemas, and RLS policies are easier for LLMs to generate correctly.

Choose Documentation‑Rich Tools: Supabase’s comprehensive docs and SwiftUI’s sample code guide the model.

Leverage High‑Automation Services: BaaS eliminates manual server setup, reducing context size.

Use Type‑Safe Languages: Swift and TypeScript catch errors at compile time, improving AI output quality.

Conclusion

The "Dead Yet?" case study demonstrates that with the right stack—SwiftUI for the client, Supabase for backend‑as‑a‑service, and AI coding assistants—developers can prototype and launch a fully functional mobile app in minutes, focusing effort on product logic rather than infrastructure.