testing

name	testing
description	Test writing protocol with structured coverage analysis

Testing Skill

Write tests for the specified code following a structured protocol. Produce thorough, maintainable tests that cover the important behavior.

Steps

1. Identify what to test: Analyze the code to determine test targets.

   • Public API functions and methods
   • Edge cases: empty inputs, nil/null values, boundary conditions
   • Error paths: invalid input, missing dependencies, timeout conditions
   • State transitions: before/after side effects

2. Choose test structure: Select the appropriate pattern.

   • Table-driven tests: When testing the same function with multiple inputs.
   • Arrange-Act-Assert (AAA): For single-scenario tests.
   • Given-When-Then: For behavior-focused tests.

3. Name tests descriptively: Test names should explain the scenario.

   • Format: TestFunctionName_Scenario_ExpectedResult
   • Example: TestValidate_EmptyInput_ReturnsError
   • Failed test names should make the failure obvious without reading the test body.

4. Set up test fixtures: Prepare the test environment.

   • Use temp directories for filesystem tests.
   • Create minimal fixtures — only what the test needs.
   • Clean up after tests (use deferred cleanup or framework helpers).

5. Write assertions: Verify behavior clearly.

   • Assert on specific values, not just "no error."
   • Check both positive cases (correct output) and negative cases (expected errors).
   • Verify side effects (file created, state changed) when relevant.

6. Mock boundaries, not internals: Keep tests focused.

   • Mock external dependencies: HTTP, databases, filesystems, clocks.
   • Do not mock internal functions or private methods.
   • Use interfaces at boundaries to enable test substitution.

7. Check coverage gaps: Review what is not yet tested.

   • Concurrent access (if applicable).
   • Large inputs or performance-sensitive paths.
   • Integration between components (consider integration tests).

Output Format

Produce test code that:

• Follows the project's existing test conventions.
• Includes a brief comment explaining each test's purpose.
• Groups related tests together.
• Is runnable without modification.