AI/MCP-assisted Playwright framework
A TypeScript/Playwright UI and API test structure developed with Claude AI and MCP-assisted workflows.
Scope of the work
A separate Playwright repository that describes using AI/MCP support for test design and debugging on the Automation Exercise application.
README-described AI-assisted scenario and Page Object design; UI/API test organization.
Turning AI suggestions into reviewable tests
The README says Claude AI supported scenario, Page Object, and locator suggestions, while MCP supported a context-aware development flow. This process description is the repository owner's statement; detailed tool-session logs are not included.
The code includes page objects, helper functions, and separate UI/API test folders. This separation keeps test flows reviewable apart from page interactions.
UI and API scope
UI files cover sign-in, product, cart, and registration flows. The cart test checks product count and product name after adding an item, while the API file validates product and brand list responses.
The README describes helpers for advertisements, pop-ups, and dynamic interface behavior. No measured stability improvement is inferred from those descriptions.
Limits and next verification
The README lists GitHub Actions, cross-browser testing, and visual regression as future work. They are not presented as completed features.
Without a shareable, dated test report, no claim is made about the framework's current pass status or production maturity.
Execution context
The README documents npx playwright test and separate examples for the UI and API folders. It also includes the Playwright HTML report command. This page does not claim a current result because the repository was not run for the portfolio.
The path to evidence
These links lead to public repository files. The presence of code does not mean the tests currently pass.
- Approach · README ↗AI/MCP process and planned improvements
- Cart UI tests ↗Adding a product and cart assertions
- API tests ↗Product and brand list responses
What do we not know yet?
There is no detailed AI/MCP session log or current test report. Maturity and stability statements in the README are not treated as independent results here.