Testing Standards

Guidance for implementing comprehensive testing that meets SDLC requirements.

Tooling

Available Tools: If using Claude Code, the pr-review-toolkit:pr-test-analyzer agent reviews test coverage and identifies gaps. Use after creating PRs to ensure adequate test coverage.

Test Organization

Required Test Categories (MUST)

Projects MUST maintain these test categories:

Category	Location	Purpose
Unit tests	Alongside source code	Test individual functions/methods
Integration tests	Dedicated test directory	Test component interactions
Documentation tests	Within doc comments	Verify examples work
End-to-end tests	Dedicated e2e directory	Test full system behavior

Naming Convention (MUST)

Test files MUST be clearly identifiable by naming convention:

Language	Unit Tests	Integration Tests
Rust	mod tests in source	tests/ directory
TypeScript	*.test.ts, *.spec.ts	tests/integration/
Python	test_*.py, *_test.py	tests/integration/
Java	*Test.java	*IntegrationTest.java
Go	*_test.go	*_integration_test.go

Test Coverage

Coverage Requirements

Requirement	Level
New functionality	MUST include tests
Bug fixes	MUST include regression tests
Coverage measurement	MUST be tracked
Minimum coverage	SHOULD target 80%
Critical paths	MUST have 95%+ coverage
Coverage reports	MUST be uploaded to tracking service

Regression Tests (MUST)

Bug fixes MUST include regression tests that:

1. Fail before the fix is applied
2. Pass after the fix is applied
3. Document the bug being fixed

#[test]
fn test_issue_123_null_pointer_on_empty_input() {
    // Regression test for issue #123
    // Previously caused null pointer when input was empty
    let result = process_input("");
    assert!(result.is_ok());
}

Critical Path Coverage (MUST)

Critical paths requiring 95%+ coverage:

• Authentication and authorization
• Data validation and sanitization
• Financial calculations
• Security-sensitive operations
• Data persistence operations

Test Execution

Pre-Merge Requirements (MUST)

1. All tests MUST pass before code can be merged
2. Tests MUST be deterministic (no flaky tests)
3. Tests MUST be isolated (no shared state between tests)
4. Tests MUST run on all supported platforms in CI

Performance Guidelines (SHOULD)

Test Type	Target Duration
Unit tests	< 1 second each
Integration tests	< 30 seconds each
Full test suite	< 10 minutes

Flaky Test Policy

Flaky tests (tests that sometimes pass, sometimes fail) MUST be:

1. Identified and tracked
2. Fixed or quarantined immediately
3. Never ignored or re-run until green

Test Best Practices

Arrange-Act-Assert Pattern (MUST)

Tests MUST follow the AAA pattern:

test("should calculate total with discount", () => {
  // Arrange
  const cart = new ShoppingCart();
  cart.addItem({ price: 100, quantity: 2 });
  const discount = 0.1;

  // Act
  const total = cart.calculateTotal(discount);

  // Assert
  expect(total).toBe(180);
});

Test Naming (MUST)

Test names MUST clearly describe what is being tested:

Good:

test_user_creation_with_valid_email_succeeds
test_login_with_invalid_password_returns_401
test_empty_cart_returns_zero_total

Bad:

test1
testUser
test_it_works

Coverage Requirements (MUST)

Tests MUST cover:

• Happy path: Expected inputs produce expected outputs
• Error cases: Invalid inputs produce appropriate errors
• Edge cases: Boundary conditions handled correctly

Edge Cases to Test

Category	Examples
Empty inputs	Empty strings, empty arrays, null/None
Boundary values	0, -1, MAX_INT, empty, single item
Invalid inputs	Wrong types, malformed data
Concurrent access	Race conditions, deadlocks
Resource limits	Large inputs, memory constraints

Implementation Checklist

• Organize tests by category (unit, integration, e2e)
• Follow naming conventions for test files
• Set up coverage measurement
• Configure coverage reporting in CI
• Ensure all tests are deterministic
• Verify test isolation
• Add regression tests for bug fixes
• Cover critical paths with 95%+ coverage

Compliance Verification

# Run all tests
make test

# Check coverage
make coverage

# Verify no flaky tests (run multiple times)
for i in {1..5}; do make test || exit 1; done

# Check critical path coverage (language-specific)
coverage report --include="src/auth/*,src/security/*" --fail-under=95

Additional Resources

Reference Files

• references/test-patterns.md - Common test patterns and anti-patterns
• references/coverage-config.md - Coverage tool configuration

Examples

• examples/rust-test-setup.md - Rust test organization
• examples/typescript-jest.config.js - Jest configuration
• examples/python-pytest.ini - Pytest configuration