When to Apply This Skill
- Writing unit tests for any function, method, or class
- Generating tests from a requirement or user story before code exists
- Producing tests for an API endpoint from an OpenAPI spec
Usage
Apply this skill when:
- Writing unit tests for any function, method, or class
- Generating tests from a requirement or user story before code exists
- Producing tests for an API endpoint from an OpenAPI spec
Read references/index.md before executing any step.
Step 1 — Understand the Unit
Read the provided source code or requirement carefully:
- What does this unit DO? (single responsibility check)
- What are its inputs and their types?
- What does it return or mutate?
- What can go wrong? (exceptions, error states)
- Does it call external systems? (DB, API, file, time, random)
Step 2 — Build the Test Case Matrix
Before writing a single line of test code, map out every case in a table:
Unit: calculate_discount(price: float, tier: str) -> float
| # | Condition | Input | Expected | Category |
|---|---|---|---|---|
| 1 | Standard tier, valid | 100.0, "standard" | 90.0 | Happy |
| 2 | Premium tier | 100.0, "premium" | 80.0 | Happy |
| 3 | Unknown tier | 100.0, "vip" | ValueError | Error |
| 4 | Price is zero | 0.0, "standard" | 0.0 | Boundary |
| 5 | Price is negative | -10.0, "standard" | ValueError | Boundary |
| 6 | Price is None | None, "standard" | TypeError | Boundary |
| 7 | Tier is None | 100.0, None | TypeError | Boundary |
Step 3 — Write Each Test (AAA Pattern)
Python (pytest):
def test_calculate_discount_when_standard_tier_expects_ten_percent_off():
# ARRANGE
price = 100.0
tier = "standard"
# ACT
result = calculate_discount(price, tier)
# ASSERT
assert result == 90.0JavaScript (Jest):
test("returns 10% discount for standard tier", () => {
// ARRANGE
const price = 100.0;
const tier = "standard";
// ACT
const result = calculateDiscount(price, tier);
// ASSERT
expect(result).toBe(90.0);
});Java (JUnit 5):
@Test
void calculateDiscount_whenStandardTier_returnsTenPercentOff() {
double result = calculator.calculateDiscount(100.0, "standard");
assertEquals(90.0, result, 0.001);
}Step 4 — Mock External Dependencies
Python:
@patch("module.external_api.call")
def test_fn_when_api_success_expects_result(mock_api):
mock_api.return_value = {"status": "ok", "data": 42}
result = my_function()
assert result == 42
mock_api.assert_called_once()JavaScript (Jest):
jest.mock("./externalService");
test("should return result when service succeeds", async () => {
externalService.fetch.mockResolvedValue({ data: 42 });
const result = await myFunction();
expect(result).toBe(42);
expect(externalService.fetch).toHaveBeenCalledTimes(1);
});Step 5 — Parametrize for Multiple Cases
pytest:
@pytest.mark.parametrize("price,tier,expected", [
(100.0, "standard", 90.0),
(100.0, "premium", 80.0),
(200.0, "standard", 180.0),
])
def test_calculate_discount_parametrized(price, tier, expected):
assert calculate_discount(price, tier) == expectedJest:
test.each([
[100.0, "standard", 90.0],
[100.0, "premium", 80.0],
])("discount(%f, %s) should be %f", (price, tier, expected) => {
expect(calculateDiscount(price, tier)).toBe(expected);
});Step 6 — Exception Testing
pytest:
def test_calculate_discount_when_negative_price_expects_value_error():
with pytest.raises(ValueError, match="Price cannot be negative"):
calculate_discount(-10.0, "standard")JUnit 5:
@Test
void calculateDiscount_whenNegativePrice_throwsIllegalArgumentException() {
assertThrows(IllegalArgumentException.class,
() -> calculator.calculateDiscount(-10.0, "standard"));
}Quality Checklist
Before emitting the final test file:
- Test matrix was built first (not ad-hoc)
- Every row in the matrix has a corresponding test
- All mocks have assertions (not just setup)
- Tests are fully isolated (no shared mutable state)
- Naming follows: test_unit_when_condition_expects_outcome
- File runs with a single command, zero additional config