Testing Patterns

What it does

This skill, "Testing Patterns," has been split into five specialized sub-skills to provide more focused testing support. It offers guidance and patterns for various testing types including unit tests, end-to-end (E2E) browser tests, integration tests involving APIs and databases, LLM evaluation, and performance/load testing. Each sub-skill addresses a specific area of the software quality assurance workflow.

When to use it

• Unit Testing: When you need to isolate and test individual components or functions of your business logic.
• End-to-End (E2E) Testing: For browser-based tests, particularly when using Playwright and incorporating visual regression or accessibility checks.
• Integration Testing: When verifying interactions between different parts of a system, such as API endpoints or database connections.
• LLM Evaluation: When evaluating the performance and accuracy of Large Language Models (LLMs) and RAG systems.
• Performance Testing: To measure and improve the speed and stability of your application under load using tools like k6 or Locust.

Key capabilities

• Provides patterns for unit testing, including AAA pattern, fixtures, mocking, and factories.
• Supports end-to-end testing with Playwright, page objects, visual regression, and accessibility checks (a11y).
• Offers guidance on integration tests involving API endpoints, databases, and contract testing.
• Includes patterns for evaluating LLMs using techniques like mock responses and tools such as DeepEval/RAGAS.
• Supports performance and load testing with tools like k6, Locust, pytest-xdist, and benchmarking.

Example prompts

• "/ork:testing-unit How can I use the AAA pattern in my unit test?"
• "/ork:testing-e2e What are best practices for using Playwright page objects?"
• "/ork:testing-llm Can you show me how to mock an LLM response for testing purposes?"

Tips & gotchas

• This skill has been replaced by five sub-skills. Use /ork:testing-unit, /ork:testing-e2e, /ork:testing-integration, /ork:testing-llm, or /ork:testing-perf to access the appropriate testing patterns.
• The specific tools mentioned (Playwright, k6, Locust, DeepEval/RAGAS) are examples and may require separate installation and configuration.