SuperAPI AI Directory

test-engineer

Testing & quality

View repository

Updated 10-15 17:31

About this skill

Test Engineer

Iron rule: Tests must prove behavior, not mechanically replicate implementation details.

Workflow

  • Step 1: Understand the object under test ⚠️ REQUIRED
    • 1.1 First read the source code and existing tests to identify key branches and dependencies.
    • 1.2 Determine whether it is more appropriate to write unit, integration, or higher-level verification.
  • Step 2: Design the test set ⚠️ REQUIRED
    • 2.1 Cover at least the main flow, failure paths, and boundary conditions.
    • 2.2 Confirm which dependencies need to be mocked and which are more suitable for real calls.
  • Step 3: Implement tests
    • 3.1 Test descriptions should focus on business behavior and expected results.
    • 3.2 Avoid excessive coupling to internal implementation details.
  • Step 4: Execute and explain results
    • 4.1 Run the most relevant test command.
    • 4.2 When it fails, first explain the root cause, then decide whether to change the code or change the test.

Focus points

  • Whether the test naming explains the behavior.
  • Whether mocks mask real integration risks.
  • Whether boundary conditions cover null values, exceptions, and ordering issues.
  • Whether regression protection was added for new features.

Project-specific tips

  • If the repository already has tests/testSetup.ts or a global mock entry point, reuse it first; do not reinvent the wheel in every test.
  • When frontend logic is involved, prioritize controllable mocks for useI18n, routing, and external requests.
  • Before writing tests, first verify the actual test commands and run methods that exist in package.json.

Anti-patterns

  • Testing only the happy path.
  • Replacing key business assertions with snapshots or internal implementation assertions.
  • Directly changing assertions to make them green when tests fail.

Pre-delivery checklist

  • The main flow, failure paths, and boundary scenarios are covered.
  • Test assertions reflect business behavior rather than internal details.
  • Relevant tests have been run, or the reason they were not run has been clearly stated.
  • If test gaps remain, they have been clearly pointed out.