testing-expert_skill

Overview

This skill is a testing expert focused on improving test reliability, diagnosing flaky tests, and guiding testing architecture decisions across Jest, Vitest, Playwright, and Testing Library. It provides pragmatic advice on test structure, mocking strategies, async timing, coverage analysis, CI performance, and cross-framework debugging. Use it proactively for migrating frameworks, stabilizing suites, and designing test boundaries for a Vue-based application.

How this skill works

It inspects the project environment to detect test frameworks, configuration files, and test file patterns, then adapts recommendations to match existing conventions and CI differences. The skill classifies problems into categories (structure, mocking, async, coverage, integration, CI) and applies targeted strategies ranging from quick fixes to complete architecture changes. It emphasizes reproducible validation steps and one-shot test runs to verify fixes without entering long-running watch modes.

When to use it

• Stabilizing flaky or intermittent test failures
• Designing or refactoring test structure and boundaries (unit/integration/e2e)
• Choosing and applying correct mocking strategies and cleanup
• Optimizing test performance and CI reliability
• Addressing async timing issues, fake timers, or race conditions

Best practices

• Detect frameworks and configs first; match current conventions before changing patterns
• Keep unit tests fast and isolated; move slow or external-dependent scenarios to integration or e2e
• Mock external boundaries only; prefer spies/stubs where verifying interactions matters
• Always clear/reset mocks and global state in beforeEach/afterEach hooks
• Use deterministic async patterns (await, findBy, runInBand when debugging) and mock timers when appropriate
• Measure meaningful coverage with focused collectCoverageFrom and thresholds, not raw line counts

Example use cases

• Convert brittle Jest tests to Vitest for faster ESM-friendly execution in a Vite-based Vue app
• Diagnose a flaky test by running serial repeats, adding proper awaits, and isolating DB transactions
• Design test folder layout separating unit, integration, and e2e tests with shared utils
• Implement CI optimizations: caching, sharding, retries, and Dockerized test databases for parity
• Improve coverage quality by focusing on behavior and critical error paths, then adding thresholds