Playwright Coffee Cart

This simple project was made to experiment with the Playwright Test Agents: planner, generator, and healer and assess their output, capabilities, and limitations for future projects.

The system under test (SUT) for this project is Coffee Cart by jecfish.

Workflow

1. Use Playwright Planner to generate a test plan for each basic feature in the Coffee Cart app. For the sake of simplicity, test plans were made for only three basic features.

@playwright_test_planner Explore the Coffee Cart app and create test plans for 3 basic features. Each feature should have its own test plan file. Name the file using ##-feature.plan.md (e.g., 01-add-coffee.plan.md).

2. Use Playwright Generator to generate tests for each test plan. I instructed the agent to reference Andrew Knight's test-automation-patterns.md for best practices and rules when generating the tests (e.g., use of POMs and fixtures, generating atomic tests). In the future, I would include test automation guidelines in a spec file that the agents can refer to, alongside any other product or engineering standards (in separate spec files).

@playwright_test_generator Generate Playwright tests for the test plan at @01-add-coffee-plan.md. Follow the conventions in the attached Test Automation Patterns markdown.

3. Use Playwright Healer to fix tests that fail. Playwright Generator automatically runs and tweaks tests as needed after generation. However, Playwright Healer was useful in fixing the first feature tests since I accidentally deleted the node_modules folder required to run the tests (oops!). It seems that Playwright Healer is more useful for repairing tests that fail later on.