Self-assess at the end of each week. Bring unchecked items to your 1-on-1 with your mentor.

Weeks 1–2 | React 19 & App Router

• I can create a new Next.js 16 project from scratch using pnpm create next-app
• I understand the difference between Server and Client Components
• I can use all 4 new React 19 hooks: use(), useActionState(), useFormStatus(), useOptimistic()
• I can write a loading.tsx skeleton and explain what Suspense does
• I can create dynamic routes and await params correctly
• I understand Tailwind v4 CSS-first config and can create custom tokens
• My portfolio is deployed to Vercel and publicly accessible
• I have submitted my portfolio for mentor review and addressed the feedback received

Weeks 3–4 | Full-Stack

• I can define a Drizzle schema and run migrations with drizzle-kit
• I can perform CRUD operations using Drizzle ORM v2 and Neon
• I can write a Server Action with Zod validation that mutates the database
• I can wire a form to a Server Action using useActionState()
• I understand revalidatePath and revalidateTag and can use them correctly
• My blog is deployed with comments persisting in Neon Postgres
• I have submitted my blog for mentor review and addressed the feedback received

Weeks 5–6 | Capstone Setup + Auth

• I have forked the capstone repo and set it up locally
• I have read all 5 key documents in the capstone repo
• Authentication is working — users can sign up, log in, log out
• Protected routes redirect unauthenticated users to /sign-in
• New user sign-ups create a record in Neon via Clerk webhook
• I have implemented forbidden.tsx and unauthorized.tsx

Weeks 7–8 | Application Core

• Users can create, read, update, and delete projects
• Users can create, read, update, and delete lists within a project
• Users can create, read, update, and delete tasks within a list
• I have used useOptimistic() for at least one mutation
• PPR is enabled on the board page

Weeks 9–10 | Polish

• Drag-and-drop works between Kanban columns using dnd-kit
• Drag-and-drop uses useOptimistic() for instant UI feedback
• I have used the after() API for at least one deferred operation

Weeks 11–12 | Testing + Ship

• At least 10 Vitest unit/integration tests covering core features
• At least 2 Playwright E2E tests for critical user flows
• All tests pass locally: pnpm check && pnpm tsc --noEmit && pnpm test --run
• pnpm build completes with zero errors before final submission
• Application is deployed to Vercel production with custom domain or live URL
• instrumentation.ts initialises OpenTelemetry
• I presented my feature in the final project showcase