Project · Build a fully-tested module from scratch

project

hard

module project

Ship something real. Submit your work when you're done.

Brief

Write a `payments.py` (or `payments.ts`) module that handles: coupon validation, price calculation (base price + tax + discount), order creation, and refund logic. Write unit tests for every function using proper AAA, fakes instead of mocks wherever possible, and boundary value tests for all validation rules. Achieve ≥80% mutation score (mutmut/Stryker). Write an integration test for the `create_order` function using an in-memory database. Document AI-assisted tests in the README (which tests were AI-generated, which were human-written).

Deliverables

A `payments.py` / `payments.ts` module with at least 5 exported functions.
A `tests/unit/` directory with unit tests achieving 100% line coverage and ≥80% mutation score.
A `tests/integration/` directory with at least 3 integration tests using an in-memory database.
A mutation testing report (mutmut/Stryker output) showing mutation score.
A `README.md` documenting: which tests were AI-generated, the review checklist applied, and mutation score.

How we grade it

All unit tests follow AAA structure with clear comments.
No `Mock()` / `jest.fn()` calls where a fake could be used instead.
Line coverage ≥ 95%.
Mutation score ≥ 80% (mutmut or Stryker).
At least one integration test that verifies DB persistence (not a mock).

Hints

Start with the validate_coupon function — it has clear rules (expired, wrong plan, already used) that generate good test cases.

For mutation testing: run mutmut run --paths-to-mutate payments.py after your tests pass. Survival of arithmetic mutants (+ → -, > → >=) means you're missing boundary tests.

Use better-sqlite3 (JS) or sqlite3 (Python) for integration tests — no Docker needed, spins up in milliseconds.

AI prompt: 'I have this coupon validation function: [paste code]. Generate pytest unit tests covering all error conditions and all boundary values. Use AAA structure and assert exact return values.'

Expected output

$ pytest tests/ --cov=payments --cov-report=term-missing ========================= test session starts ========================= tests/unit/test_coupon.py ....... 7 passed tests/unit/test_pricing.py ........ 8 passed tests/unit/test_order.py ...... 6 passed tests/integration/test_order_db.py ... 3 passed ========================= 24 passed in 0.42s ========================= Name Stmts Miss Cover -------------------------------------- payments.py 78 2 97% $ mutmut results 🎉 24 out of 27 mutants killed (89% mutation score)

Stretch goals

Add property-based tests using Hypothesis (Python) or fast-check (TypeScript). For calculate_price(base, tax_pct, discount_pct), generate random valid inputs and assert that the result is always between 0 and base * (1 + tax_pct/100). What invariants can you assert?

Write a test that runs the full payment flow: create coupon → create order with coupon → verify discount applied → refund → verify balance restored. This is an integration test that spans multiple functions.

Integrate mutation testing into CI: add a GitHub Actions step that runs mutmut and fails if mutation score drops below 75%.

Add a parametrized test that covers all 15 combinations of (coupon type: percent/flat/none) × (plan: monthly/annual/lifetime) × (user status: new/existing). Use @pytest.mark.parametrize or test.each.