QA AI Toolkit

🎭 Test Authoring

Playwright E2E Testing

Author resilient Playwright end-to-end tests with the Page Object Model, fixtures, stable user-facing locators, and web-first assertions. Includes a live-browser Playwright MCP server for inspecting user journeys before writing tests.

Skill Agent MCP Server

Playwright E2E POM

npx qa-ai-repo add playwright-e2e copy

🥒 Test Authoring

Playwright → BDD Converter

Convert imperative Playwright tests into BDD — auto-generate Gherkin .feature specs written in business language, and wire step definitions to Playwright fixtures and page objects with the playwright-bdd runner. Behavior is verified before originals are deleted.

Skill Agent

Gherkin BDD Playwright

npx qa-ai-repo add playwright-bdd copy

📸 Test Authoring

Visual Regression Testing

Reliable screenshot diffing with Playwright's toHaveScreenshot — determinism first (disabled animations, frozen clocks, pinned fonts, container-generated baselines), tuned thresholds, and masked dynamic regions so diffs catch real bugs, not noise.

Skill Agent

Screenshots Playwright CI

npx qa-ai-repo add visual-regression copy

♿ Test Authoring

Accessibility Testing

Test web UIs against WCAG 2.2 AA — axe-core automation wired in as the floor, plus the manual review automation can't do: keyboard-only navigation, focus management, and screen-reader semantics. Guards against aria-label spam and treating a clean axe run as full coverage.

Skill Agent MCP Server

WCAG 2.2 axe-core a11y

npx qa-ai-repo add accessibility-testing copy

🤝 Test Authoring

API Contract Testing

Catch breaking API changes before they ship — consumer-driven contracts with Pact and spec-first validation with OpenAPI tooling, across REST, GraphQL, and event boundaries. Wires provider verification and blocking can-i-deploy gates into CI.

Skill Agent

Pact OpenAPI CI Gates

npx qa-ai-repo add api-contract-testing copy

🎲 Suite Health

Flaky Test Triage

Detect, score, and root-cause flaky tests — compute per-test flake scores from reruns and CI history, classify each flake (timing, shared state, external deps, animations, randomness), fix at the root, and quarantine with an owner and SLA — never blanket retries.

Skill Agent

Flakiness CI Triage

npx qa-ai-repo add flaky-test-triage copy

🧬 Suite Health

Jest Coverage & Mutation Testing

Measure whether your unit tests catch real bugs — Jest code coverage plus Stryker mutation testing to prove tests do more than execute lines. Kill survived mutants and gate CI on mutation score, not coverage percentage alone.

Skill Agent

Jest Stryker Mutation

npx qa-ai-repo add jest-coverage-mutation copy

🔍 Suite Health

UI Test Overuse Auditor

Scan Playwright, WebdriverIO, and Selenium suites in any language to find where UI tests are overused — collapse validation matrices and permission loops into parameterized API or unit tests, keeping only genuine browser journeys at the top of the pyramid.

Skill Agent

Audit Selenium Migration

npx qa-ai-repo add ui-test-auditor copy

🧭 Strategy

QA Strategy

Interview a team about its product, stack, size, and release cadence, then generate a tailored risk-based QA strategy — automation plan, quality gates, tooling, metrics with precise formulas, and a phased rollout. Every recommendation traces to a stated input and lands as a measurable gate.

Skill Agent

Strategy Risk-Based Metrics

npx qa-ai-repo add qa-strategy copy

🔺 Strategy

Full-Stack Test Pyramid

Analyze an application end to end — frontend, backend, and middleware — and produce a complete test pyramid: what to test at each layer and level, which tools to use, runnable per-level example tests, and CI wiring for which suite gates which stage.

Skill Agent

Architecture Test Levels CI

npx qa-ai-repo add test-pyramid copy

⚡ Performance

Performance Testing

Prove speed and scale targets under realistic load — k6 load, stress, soak, and spike tests plus Lighthouse web vitals. Defines SLOs (p95/p99 latency, error rate, throughput, LCP/INP/CLS) first, then turns every threshold into a pass/fail CI gate. Percentiles, never averages.

Skill Agent

k6 Lighthouse SLOs

npx qa-ai-repo add performance-testing copy

🧪 AI & LLM

LLM Evaluation (DeepEval)

Prove an LLM, RAG, or agent feature is correct with numbers a release can gate on — DeepEval suites for faithfulness, answer relevancy, hallucination, and tool-call correctness, built on a versioned golden dataset and wired into CI as pytest exit codes.

Skill Agent

DeepEval RAG LLM Evals

npx qa-ai-repo add llm-eval copy