What problem does it solve? When an end-to-end test suite fails, engineers must manually investigate each failure to determine whether the test itself is broken or the production code has a real bug. This Skill automates that triage by running the tests, analyzing every failure and flaky test against production code and user stories, and producing a structured classification report. ## Core Features & Use Cases - Automated test execution: Locates and runs e2e tests for a product area using the project's test runner (Playwright, Cypress, or custom commands), capturing results, errors, and retry behavior. - Failure classification: Cross-references each failed or flaky test against production code and DefProd user stories to classify it as a test fault or production fault, with a specific recommended fix. - Structured reporting: Saves a Markdown report (e.g. AUTH-test-check.md) summarizing pass/fail/flaky counts and per-test analysis, ready to drive the follow-up fix workflow. - Use Case: After a refactor, run the skill against the BILLING area to discover three failures, learn that two are outdated selectors (test faults) and one is a genuine regression (production fault), and get a report to hand off for fixing. ## Quick Start Ask your AI agent to run the area tests for your product, for example: run the e2e tests for the AUTH area of my product and classify any failures.
