All projects
C

Chapterly

AI-powered reading companion with gamification, social features, and smart insights.

Next.js 14TypeScriptSupabaseAnthropic ClaudeStripeTailwind CSSVercel

Overview

Chapterly is a full-stack reading tracker that uses Claude AI to generate personalized book insights, predict reading moods, and surface hidden themes. It features an XP/leveling system, reading streaks, social follows, buddy reads, a live reading timer, Goodreads CSV import, and Stripe-powered premium subscriptions.

Live
Status
116+
Tests
30+
Features

Key Features

  • AI-generated reading insights, mood predictions, and genre analysis powered by Claude Haiku
  • Full gamification engine: XP rewards, level-up system, reading streaks, and leaderboards
  • Social layer: follows, buddy reads, shared challenges, and reading activity feeds
  • Stripe subscriptions (monthly + annual), referral system, and push notifications via VAPID
  • GitHub Actions CI with 70+ Vitest unit tests and 46 Playwright API contract tests
  • Goodreads CSV import supporting up to 500 books with Google Books metadata enrichment

What I Learned

Supabase RLS policies require careful design — a single missing policy silently hides data from the UI
AI features need retry logic (429/529) and greedy JSON extraction to handle Claude preamble in production
Gamification loops (XP, streaks, leaderboards) dramatically increase session depth — build them early