Obsession-I built an obsession tracker for developers...

This is a submission for Weekend Challenge: Passion Edition

What I Built

I built TOTAL_OBSESSION_DETECTOR (STRAT_OS), a Neo-Brutalist "Editorial Aesthetic" passion & neurosis analyzer designed specifically for developer subcultures.

Whether it's spending $600 tuning custom mechanical keyboard switches, building multi-node Kubernetes home labs in a closet, or obsessively optimizing database index selectivities, developers have high-severity fixations. TOTAL_OBSESSION_DETECTOR lets you log these behavioral symtoms, run dynamic, AI-powered "Passion Audits" to diagnose your obsession level (complete with a brutal roast and an intellectual validation manifesto), and instantly generate a Quick Defense Shield—a hyper-formal, jargon-rich legalistic defense letter to dismiss questioning from supervisors, landlords, or partners.

Demo

🔗 Live Deployed App: TOTAL_OBSESSION_DETECTOR

Code

🔗 GitHub Repository: [https://github.com/arfath62/obsession-]

How I Built It

I engineered a full-stack Node.js & React single-page application focused on high performance and extreme resilience:

1. Editorial Design System: Kept the layout true to print design and neo-brutalist aesthetics—featuring hard geometric lines, deep blacks, spacious negative layouts, raw code readouts, and responsive transitions using motion.
2. Resilient AI Pipeline: Implemented the official Google GenAI SDK (@google/genai) inside an Express.js backend. To safeguard the application from high-demand 503 UNAVAILABLE errors or rate-limiting:
   • Configured an active retry protocol using exponential backoffs.
   • Built an automated model-fallback waterfall. If gemini-3.5-flash experiences transient load, the server seamlessly cascades down to gemini-flash-latest and then gemini-3.1-flash-lite, assuring 100% backend uptime.
3. Structured Outputs: Forced Gemini to spit out structured, schema-validated JSON diagnostics so the client could instantly render progress gauges, witty deconstruction blocks, and formatted procedural steps.

Prize Categories

Best Use of Google AI

I integrated the official @google/genai SDK and established a multi-model fallback waterfall system (gemini-3.5-flash ➡️ gemini-flash-latest ➡️ gemini-3.1-flash-lite) to bypass high-demand unavailability spikes, assuring that developer audits and legal defense generation are always available.
thank you....