🧠 Dopamind

You started reading. Then you zoned out. Then you forgot where you were. Sound familiar?

Dopamind is an open-source, ADHD-friendly learning tool that transforms long-form content — PDFs, ebooks, articles — into focused, digestible learning experiences.

Built for anyone who struggles to focus on long content — which, in a world of endless feeds and notifications, is most of us.

---

Try it (no API key needed)

git clone https://github.com/jayjoolee/dopamind.git
cd dopamind
npm install
npm start

Open http://localhost:3000 and click ✨ Try a demo (no key) — you'll see exactly what Dopamind does to a piece of text, instantly, without signing up for anything.

When you're ready to run it on your own content, add a Claude API key in ⚙️ Settings (it stays in your browser, never touches any server).

---

The Problem

Long content is brutal for ADHD brains:

• You start a PDF, lose the thread by page 3
• You read the same paragraph four times
• You finish a chapter and remember nothing
• You zone out mid-article and have no idea what you missed

Existing tools summarize content. That's not the same as helping you learn it.

---

What Dopamind Does

Dopamind breaks content into short, focused chunks — each with:

• A one-line hook before you read ("Here's why this part matters")
• Key points after each chunk (so zoning out isn't fatal)
• A "catch me up" button — instantly see what you missed without scrolling back
• Progress that feels real — small wins trigger the dopamine hit your brain needs

---

Roadmap

• v0.1 — PDF / plain text input → ADHD-friendly reading view
• v0.2 — EPUB support
• v0.3 — YouTube video (with captions) support
• v0.4 — "Catch me up" interactive mode
• v1.0 — Web app with BYOK (Bring Your Own API Key)

---

How It Works

Dopamind uses an LLM (Claude / OpenAI — your choice, your API key) to:

1. Parse your content
2. Break it into attention-sized chunks
3. Generate hooks, key points, and catch-up summaries
4. Present it in a clean, distraction-free interface

Your content never leaves your machine unless you choose to send it to an LLM API.

---

Status

🚧 Early stage — building in public. Star this repo if the idea resonates.

---

Contributing

Ideas, feedback, and PRs welcome. Especially from people who know firsthand what it's like to read the same sentence six times.

Open an issue or start a discussion.

---

License

MIT