Qa Automation

---
name: Qa Automation
description: Use when building test frameworks, designing test architecture, setting up CI integration, and managing test data for regression detection.
category: QA & Testing
version: 1.0.0
tools: []
---

# QA Automation

## Test Architecture Design

1. Structure tests in the testing pyramid: 70% fast unit tests, 20% fewer integration tests, 10% minimal end-to-end tests.
2. Organize tests by feature, not by type. Each feature directory contains its unit, integration, and e2e tests together.
3. Implement the Page Object Model for UI tests. Each page or component gets a class that encapsulates selectors and interactions.
4. Create a shared test utilities library: custom assertions, data builders, mock factories, and wait helpers.
5. Use test tags (smoke, regression, critical-path) to enable selective test execution per context.

## Test Framework Configuration

Use Playwright for browser-based e2e tests. Configure multiple browser projects (Chromium, Firefox, WebKit) with shared setup. Use Vitest or Jest for unit and integration tests. Configure code coverage thresholds: 80% line coverage minimum for critical modules. Configure test parallelization: Playwright runs tests in parallel workers. Implement test retries with limits: retry flaky tests up to 2 times in CI, but flag them for investigation.

## Test Data Management

Use factories (factory-bot pattern) to generate test data. Each factory produces a valid entity with sensible defaults that can be overridden. Isolate test data per test. Each test creates its own data, runs assertions, and cleans up. Tests must not depend on shared state. Use database transactions for integration tests: start a transaction before the test, roll back after. This is faster than truncating tables. Seed reference data (countries, currencies, permission types) once in a fixture that all tests share.

## CI/CD Integration

Run unit tests on every commit. Run integration tests on every pull request. Run full regression suites nightly. Cache test dependencies (node_modules, browser binaries) to reduce CI setup time. Fail the build immediately when tests fail. Do not allow merging PRs with test failures. Upload test artifacts on failure: screenshots, video recordings, trace files, and HTML reports. Report test results as PR checks with inline annotations.

## Flaky Test Management

Track flaky test occurrences. A test that fails more than 5% of runs without code changes is flaky. Quarantine flaky tests: move them to a separate test suite that runs but does not block deployments. Fix flaky tests by root cause: timing issues (add explicit waits), test isolation (remove shared state), environment differences (use containers). Review the flaky test dashboard weekly. Set a team target: zero flaky tests in the critical-path suite.

## Assertion Best Practices

Write assertions that describe the expected behavior, not the implementation: `expect(order.status).toBe('confirmed')` not `expect(db.query).toHaveBeenCalled()`. Use custom matchers for domain-specific assertions. Assert on visible behavior in UI tests: text content, element visibility, URL changes. Avoid asserting on CSS classes or DOM structure. Use snapshot testing sparingly for serialized output (API responses, rendered components).