本文へ移動
cccskills
無料GitHub で公開

nw-test-refactoring-catalog

Detailed refactoring mechanics with step-by-step procedures, and test code smell catalog with detection patterns and before/after examples

インストール方法を見る

含まれるファイル(1)

  • SKILL.md3.5 KB

SKILL.md(原文)

インストールする前に、エージェントに与えられる指示の中身を確認できます。

Test Refactoring Catalog

Test Code Smells (Full Detail)

L1 Readability Smells

Obscure Test

  • Problem: test name does not reveal business scenario being tested
  • Detection: generic names like Test1(), ProcessOrderTest(), or names requiring reading test body to understand
  • Solution: rename to Given_When_Then or should_do_expected_thing_when_condition format
Before: public void Test1() { /* ... */ }
After:  public void ProcessOrder_PremiumCustomer_AppliesCorrectDiscount() { /* ... */ }

Hard-Coded Test Data

  • Problem: magic numbers and strings obscure business rules being tested
  • Detection: numbers like 1000, 0.15, strings without explanation
  • Solution: extract to named constants that reveal business meaning
Before: Assert.Equal(850, result.Total);  // What discount?
After:  const decimal EXPECTED_TOTAL = 1000 * (1 - 0.15m);
        Assert.Equal(EXPECTED_TOTAL, result.Total);

Assertion Roulette

  • Problem: multiple assertions without messages make failures unclear
  • Detection: multiple Assert.* calls without message parameter
  • Solution: add descriptive message to each assertion explaining expected business outcome

L2 Complexity Smells

Eager Test

  • Problem: single test verifies multiple unrelated behaviors
  • Detection: multiple arrange/act/assert cycles or assertions testing different concerns
  • Solution: split into focused tests, one per business scenario
Before: ProcessOrderTest() { /* tests discount AND shipping AND tax */ }
After:  ProcessOrder_AppliesDiscount()
        ProcessOrder_CalculatesShipping()
        ProcessOrder_CalculatesTax()

Prefer parameterized tests for variations of the same behavior.

Test Code Duplication

  • Problem: repeated test setup logic across multiple tests
  • Detection: same object creation, mock setup, or data builders copied in 3+ tests
  • Solution: extract helper methods
Extract: CreatePremiumCustomer(), CreateHighValueOrder()

Conditional Test Logic

  • Problem: if/switch statements in test code make tests non-deterministic
  • Detection: if, switch, for loops in test methods
  • Solution: replace with parameterized tests
# Before: if/else in test
# After:
@pytest.mark.parametrize("input,expected", [...])
def test_behavior(input, expected): ...

L3 Organization Smells

Mystery Guest

  • Problem: test depends on external files or hidden dependencies
  • Detection: File.ReadAllText, database queries, external config in tests
  • Solution: inline test data or make dependency explicit in test setup

Test Class Bloat

  • Problem: single test class contains tests for multiple unrelated concerns
  • Detection: test class with 15+ tests covering different features
  • Solution: split by feature
Before: UserServiceTests (31 tests)
After:  UserAuthTests, UserProfileTests, UserNotificationTests

General Fixture

  • Problem: shared fixture used by tests with different needs
  • Detection: SetUp method creates data used by only some tests
  • Solution: move to per-test setup methods or test-specific fixtures

For production code refactoring techniques and mechanics, load the progressive-refactoring skill.

レビュー

まだレビューはありません。使ってみた感想をお寄せください。

同じリポジトリのスキル

概要と使いどころ

Review dimensions for validating agent quality - template compliance, safety, testing, and priority validation

日本語の概要は準備中です。原文の説明を表示しています。

nWave-ai/nWave6162026年9月16日 更新

Review dimensions for validating agent quality - template compliance, safety, testing, and priority validation

日本語の概要は準備中です。原文の説明を表示しています。

nWave-ai/nWave6162026年9月16日 更新

Review dimensions for acceptance test quality - happy path bias, GWT compliance, business language purity, coverage completeness, walking skeleton user-centricity, priority validation, observable behavior assertions, traceability coverage, and walking skeleton boundary proof

日本語の概要は準備中です。原文の説明を表示しています。

nWave-ai/nWave6162026年9月16日 更新

Detailed 5-phase workflow for creating agents - from requirements analysis through validation and iterative refinement

日本語の概要は準備中です。原文の説明を表示しています。

nWave-ai/nWave6162026年9月16日 更新

5-layer testing approach for agent validation including adversarial testing, security validation, and prompt injection resistance

日本語の概要は準備中です。原文の説明を表示しています。

nWave-ai/nWave6162026年9月16日 更新

Architectural style selection decision matrices, trade-off analysis, structural enforcement rules, and combination patterns. Load when choosing or evaluating architecture styles.

日本語の概要は準備中です。原文の説明を表示しています。

nWave-ai/nWave6162026年9月16日 更新

nWave-ai のスキルをすべて見る

このスキルの問題を報告する