Verification Skill

Purpose

Provide structured methodology for comprehensive software verification including testing, static analysis, coverage measurement, and quality metrics.

Testing Pyramid

        ╱╲
       ╱  ╲       E2E Tests (few, slow, high confidence)
      ╱────╲
     ╱      ╲     Integration Tests (moderate count)
    ╱────────╲
   ╱          ╲   Unit Tests (many, fast, focused)
  ╱────────────╲

• •Unit tests: Test individual functions/methods in isolation. Mock dependencies.
• •Integration tests: Test module interactions. Use real (or test) databases.
• •E2E tests: Test full user flows. Slowest but highest confidence.

Target ratio: ~70% unit, ~20% integration, ~10% E2E.

Test Structure: Arrange-Act-Assert

Every test should follow this pattern:

def test_user_creation_with_valid_email_returns_user():
    # Arrange — set up preconditions
    email = "alice@example.com"
    repo = FakeUserRepository()
    service = UserService(repo)

    # Act — perform the action under test
    user = service.create_user(email)

    # Assert — verify the result
    assert user.email == email
    assert user.id is not None
    assert repo.find_by_id(user.id) is not None

Test Naming Convention

Format: test_<unit>_<scenario>_<expected_result>

Good examples:

• •test_parse_config_with_missing_key_raises_validation_error
• •test_calculate_tax_with_zero_amount_returns_zero
• •test_authenticate_with_expired_token_returns_unauthorized

What to Test

Always test:

• •Happy path (normal operation)
• •Edge cases: empty input, null/None, zero, negative, max values
• •Error paths: invalid input, network failures, timeout, permission denied
• •Boundary values: off-by-one, min/max of ranges, empty collections
• •State transitions: initial → active → suspended → terminated

Test smells to avoid:

• •No assertions: Test runs code but doesn't verify anything
• •Tautological tests: assert True, assert x == x
• •Implementation coupling: Tests break when internals change but behavior doesn't
• •Test interdependence: Test B only passes if Test A runs first
• •Excessive mocking: Mocking the thing you're trying to test

Framework Quick Reference

Python (pytest)

# Run all tests
pytest -v

# Run with coverage
pytest --cov=src --cov-report=term-missing --cov-fail-under=80

# Run specific test
pytest tests/test_auth.py::test_login_success -v

# Run tests matching pattern
pytest -k "test_payment" -v

# Type checking
mypy src/ --strict

TypeScript (Vitest / Jest)

# Run all tests
npx vitest run

# Run with coverage
npx vitest run --coverage

# Run specific file
npx vitest run src/auth.test.ts

# Type check
npx tsc --noEmit

Rust

# Run all tests
cargo test

# Run with output
cargo test -- --nocapture

# Run specific test
cargo test test_payment_processing

# Coverage
cargo tarpaulin --out Html

C/C++ (GoogleTest / CTest)

# Build and run
cmake --build build --target test
ctest --test-dir build -V

# Run specific test
./build/tests/test_binary --gtest_filter="SuiteName.TestName"

Coverage Targets

Static Analysis Tools

Test Templates

See the templates/ directory for ready-to-use test file templates:

• •test_template.py.tmpl — Python pytest template
• •jest.test.ts.tmpl — TypeScript Jest/Vitest template