An intelligent task-planning application inspired by modern AI-assisted workflows and Copilot-style UX.
Reads your goal → breaks it into steps → streams the plan into a sleek UI card.
Tested. Deployed. Fake streaming included. Built to showcase frontend ✕ AI ✕ UX skills.
Type your goal → Click Generate Plan → Watch the plan stream in real time.
| Area | Tech Used |
|---|---|
| Framework | Next.js 14 (App Router, TypeScript) |
| Styling | Tailwind CSS + Shadcn UI + Lucide Icons |
| Animation | Custom fade-in via Tailwind config |
| API Layer | Simulated OpenAI with ReadableStream |
| Hooks | Custom useStream() for real-time UI |
| Testing | Playwright (End-to-End tests) |
| Deploy | Vercel with environment config |
- Clean React component architecture with reusable UI blocks
- Predictable client-side state handling using hooks
- Simulated streaming responses to mirror real API behavior
- Emphasis on frontend performance, UX flow, and maintainability
chain_of_mind/
├── app/
│ ├── api/
│ │ └── plan/
│ │ └── route.ts # Fake OpenAI streaming logic
│ ├── dashboard/
│ │ └── page.tsx # UI: textarea, button, result card
│ ├── layout.tsx # Layout shell with header
│ └── globals.css # Tailwind + base styles
│
├── components/
│ ├── ui/
│ │ ├── button.tsx # Shadcn button
│ │ ├── card.tsx # Shadcn card
│ │ └── input.tsx # Shadcn input
│ └── AgentCard.tsx # Custom animated result card
│
├── lib/
│ └── useStream.ts # Hook for streaming data
│
├── public/ # Default assets
├── tests/
│ └── agent.spec.ts # Playwright E2E test
│
├── .env.local # OPENAI_API_KEY (optional/fake)
├── tailwind.config.ts # Extended with animation
├── package.json
└── README.md
## ⏱️ Hour-by-Hour Build Log
| Hour | Milestone |
|------|-----------|
| **Hour 1** | Created project with `create-next-app` + Tailwind + TypeScript + Shadcn |
| **Hour 2** | Built layout shell with header and global styles |
| **Hour 3** | Integrated fake OpenAI streaming API and useStream hook |
| **Hour 4** | Built AgentCard component with fade-in animation |
| **Hour 5** | Wrote Playwright test and deployed to Vercel |
---
## 🧪 Test Result
✅ All end-to-end tests pass locally using Playwright:
```bash
> npm run test:e2e
✓ 1 passed (Playwright: streamed agent response)
👩💻 Author
Anvi Yadav
Final-year CS student | Building production-ready, AI-powered frontend systems.
Focused on building clean, maintainable React applications with reliable UI behavior and thoughtful user experience.
🔗 GitHub: github.com/Yanvi09
🔗 Project URL: chain-of-mind.vercel.app