Moonbug: 우주 달력 및 이벤트 카탈로그

발행: (2026년 4월 20일 AM 05:39 GMT+9)
3 분 소요
원문: Dev.to

Source: Dev.to

How I Built It: The “Boss‑Agent” Workflow

My development process was a study in AI orchestration. I operated at a “Boss‑Agent” level, managing two distinct AI entities to take Moonbug from a mobile concept to a live deployment entirely from a smartphone.

Architectural Planning (Gemini as Architect)

I used Gemini 1.5 Pro to architect the project. We defined the MoonPlayer philosophy—aligning digital habits with lunar cycles rather than a 24‑hour clock. Gemini provided the logic for cosmic computations, structured the React‑based sleep screens, and designed the modularity between the web app and the browser extension.

The Command Center (Firebase Studio Agent)

Operating within Firebase Studio on my smartphone, I acted as the “Boss.” I injected high‑level prompts into the Studio’s AI agent to generate functional code components, set up the Firestore NoSQL schema, and configure the complex Firebase CLI environment.

The Feedback Loop & Debugging

I tested the app in real‑time on my mobile browser, reporting logs and errors back to the Firebase Studio agent. It acted as my engineering assistant, returning fixes for major hurdles like OAuth 2.0 --no‑localhost authentication and React routing conflicts. This iterative loop allowed me to resolve technical debt and major errors instantly, leading to a fully functional prototype without ever touching a desktop computer.

  • Live Demo (Functional Prototype)
    (Developed and hosted via Google Cloud Workstations)

  • GitHub Repository

Prize Category

Best Use of Google Gemini – Highlighting a sophisticated “Boss‑Agent” orchestration where Gemini acted as lead architect and project manager. I used Gemini 1.5 Pro’s reasoning capabilities to transform dry astronomical data into a personalized user experience and used the Firebase Studio agent to execute complex development tasks on mobile.

Specific Use of Google Gemini

The Intelligence Layer

Integrated the Gemini API as a scientific curator. It processes live NASA and NOAA data, summarizing it into “Nature Cards” so users stay informed about astronomical and environmental events without information overload.

The Context Engine

Gemini reasons through user‑specific data (like birth moon phases) to suggest nature‑themed hobbies, travel trends, and “Circadian Deep Work” blocks.

The “Awareness Guard”

In the companion browser extension, Gemini filters incoming weather and astronomical notifications, ensuring only peaceful, significant events trigger the custom “Solar Alarms.”

Environmental Impact

Moonbug bridges the gap between high‑tech living and the natural world. By replacing “doom‑scrolling” with “lunar‑scrolling,” the app fosters a mindset of environmental stewardship and astronomical awe. It encourages users to synchronize their digital lives with Earth’s natural rhythms, reducing digital fatigue and increasing planetary connection.

Tech Stack

  • AI: Gemini 1.5 Pro API (Orchestration & Data Curation)
  • Backend: Firebase (Firestore, Hosting, CLI)
  • Frontend: React, Vite, Tailwind CSS
  • IDE: Firebase Studio (Cloud Workstations)
  • Target: Cross‑platform Web + Chrome Extension
0 조회
Back to Blog

관련 글

더 보기 »

15분 도시 주변 데이터 파이프라인

개요 방금 두 번째 데이터 엔지니어링 파이프라인을 마무리했습니다. 리포지토리: GitHub – OSM 15 Minute City https://github.com/mushroomsandchai/osm 대시보드: S...