Family Task Board

Example exercise built during the Saranen / ModernPath AI-assisted coding course.

This repository is a teaching artifact: it shows how a local-first family chore board can be designed, specified, and implemented with AI-assisted workflows (research → PRD → specs → TDD → review). It is not a production product.

What it is

A wall-tablet chore app for a shared kitchen:

• Today — who does what today (rotation + open pool tasks)
• Week — rotation plan for the next 7 days
• Coach — AI homework coach (who should do what, who has done the most)
• Setup — members, tasks, rewards, backup/import
• Local-first — SQLite in the browser, no accounts, PWA-friendly

Tech stack

• Vite 6, React 18, TypeScript, React Router 6
• Vitest + Testing Library + Playwright smoke tests
• Spec-driven development (specs/, AGENTS.md)

Quick start

npm install
npm run dev       # http://localhost:5180
npm run coach:api # homework coach FastAPI on http://127.0.0.1:8001
npm test
npm run test:agent
npm run build

Load sample data: Setup → Backup → Load Vuorio family sample.

Week 3: ready-made hosting examples

Docker and Vercel walkthrough (Finnish) — includes the working Coach API, AI prompts, verification, troubleshooting and storage limits.

docker compose up --build -d
# Open http://localhost:8080

For Vercel, import this repository with the Vite preset; vercel.json configures the frontend and Python function. Neither example needs an API key. These are synthetic-data demos: household data remains browser-local, the hosted coach is stateless, and accounts/cloud sync are not implemented by hosting.

Course context

Built as a hands-on example for learning:

• Product research and PRD writing with AI
• Feature specs with Given/When/Then acceptance criteria
• Test-driven development against specs
• Iterative UX polish and audit cycles

See specs/PRD.md, specs/features/, and prompts.md for the documentation trail.

Python agents live in agents/ — copyable kit in agents/AGENTS.md, worked example agents/homework-coach-agent/.

License

MIT — use freely for learning and teaching.