problem_statement.txt

Standard Exercise: Reading List Feature
Build a small reading list tracker. This is intentionally scoped — we care about your process and how you work with AI agents, not a polished product.
What to build
API — add a book (title, author, status: to-read / reading / done), list books filterable by status, and one simple aggregation (e.g., count by status). Basic validation required (title is required; status must be one of the three values).
UI — a single page to add a book, view the list, and change a book's status. Basic success/error feedback.
Tests — cover core API behavior (add, list, filter). At least one test must fail initially and then be fixed on screen.
Deploy — ship it somewhere publicly accessible. Free tiers on Railway, Render, Vercel, or Hugging Face Spaces are all fine. A Dockerfile is acceptable if live deployment isn't possible. A local run script alone is not sufficient. The live URL must be testable by our team directly from a browser without any local setup on our end.
In-memory or file-based storage is completely fine.