From Idea to Tool in 48 Hours: The Neorgon Build Process

Test if the idea is worth building at all.

Questions to answer:

• ❓ Does this solve a real problem I've experienced?
• ❓ Would I use this tool weekly?
• ❓ Can it be built in 48 hours?
• ❓ Does it fit Neorgon's "no paywall, no signup" philosophy?

Decision: If 3+ answers are "yes", proceed. Otherwise, abandon.

cp -r _template/app new-tool-site
cd new-tool-site
python3 -m http.server 8800  # Start dev server

Build the minimum viable product. Ignore:

• ❌ Animations
• ❌ Settings/preferences
• ❌ Advanced features
• ❌ Perfect styling

Focus: The single feature that makes the tool useful.

• Apply glassmorphism styles
• Add basic hover effects
• Ensure mobile works
• Test keyboard navigation

Generate OG image in OG Studio

Create favicon.ico

Design tool icon

# Add card to neorgon-site/index.html
# Update category in js/search.js
# Add preview to js/previews.js
git add .
git commit -m "Add new-tool card"

• OG meta tags (title, description, image)
• JSON-LD schema markup
• Canonical URL
• Twitter Card tags

Accessibility Test:

• Enable prefers-reduced-motion (macOS Settings → Accessibility)
• Navigate with keyboard only
• Test with VoiceOver
• Check color contrast with WebAIM

Performance Test:

• Lighthouse audit (target: 95+ performance)
• 3G network simulation
• Mobile viewport testing

Create minimal README.md:

• What the tool does (one sentence)
• How to run locally
• Live URL
• Screenshot

Push to GitHub Pages:

git add .
git commit -m "Initial release: new-tool-site"
git push origin main
gh-pages -d .  # Deploy to GitHub Pages

Add to neorgon-site:

• Create HTML card with correct accent color
• Add to search.js categories
• Test search/filter

✅ Tool deployed and accessible
✅ Hub card added and working
✅ Search finds the tool
✅ OG image generated
✅ Analytics tracking (Plausible)
✅ Lighthouse score ≥ 95
✅ Mobile responsive
✅ README created
✅ Social post drafted

❌ Skip These (They Don't Matter at Launch)

• E2E Tests: Manual testing is faster for 48-hour tools
• Unit Tests: Not worth it for 500-line apps
• CI/CD: git push is our CI
• TypeScript: Vanilla JS is faster to write
• Feature Flags: Ship finished features only
• Analytics Dashboard: Plausible is configured automatically
• User Accounts: Goes against "no signup" philosophy

❌ Pitfall 1: Perfectionism

Symptom: Spending 3 hours on a button animation.

Fix: Set a timer. 30 minutes max per feature.

❌ Pitfall 2: Scope Creep

Symptom: "Wouldn't it be cool if it also had AI features?"

Fix: Write the scope on a sticky note. If it's not on the note, it doesn't exist.

❌ Pitfall 3: Testing Everything

Symptom: 6 hours writing tests for 400 lines of code.

Fix: Manual testing is sufficient for 48-hour tools.

❌ Pitfall 4: Deployment Drama

Symptom: "It works locally but breaks in production."

Fix: Test production deploys on staging branches first.

# _template/app Structure
├── index.html          # Title, meta, script tags
├── css/
│   └── style.css       # Glassmorphism styles
└── js/
    ├── app.js          # Main entry (import everything)
    ├── state.js        # Mutable state object
    ├── render.js       # DOM rendering
    ├── events.js       # Click handlers, etc.
    └── utils.js        # Helper functions

# Launch checklist
✅ Local dev server working
✅ Core feature functional
✅ Mobile responsive
✅ OG image created
✅ Added to hub
✅ README written
✅ Deployed to GitHub Pages
✅ Shared on social media