QA Testing Skills

AI-powered QA testing prompts that work with any AI agent — Claude Code, Cursor, Windsurf, GitHub Copilot, ChatGPT, Gemini, or any LLM.

One command to install. Zero dependencies. Just markdown prompts that turn any AI into a QA architect.

Quick Start

npx @aktibaba/qa-testing-skills init

That's it. The CLI auto-detects your AI tool and installs the prompts in the right place.

Platform-specific install

npx @aktibaba/qa-testing-skills init claude      # → .claude/commands/
npx @aktibaba/qa-testing-skills init cursor      # → .cursor/rules/
npx @aktibaba/qa-testing-skills init windsurf    # → .windsurf/rules/
npx @aktibaba/qa-testing-skills init copilot     # → .github/copilot-instructions.md
npx @aktibaba/qa-testing-skills init generic     # → prompts/ (copy-paste anywhere)

Available Prompts

| Prompt | What it does | |--------|-------------| | qa-api | Design and generate API test suites (REST, GraphQL, gRPC) | | qa-ui | Browser automation and E2E tests (Playwright, Cypress, Selenium) | | qa-env | Docker-based reproducible test environments | | qa-int | Integration tests for service interactions, databases, queues | | qa-perf | Load testing, stress testing, benchmarks (k6, Locust, Artillery) | | qa-sec | Security testing — OWASP Top 10, scanning, vulnerability detection | | qa-ci | CI/CD pipeline with quality gates (GitHub Actions, GitLab CI, Jenkins) | | qa-reg | Regression suite organization — smoke, sanity, full regression tiers | | qa-rv | Test quality review — scored audit with actionable recommendations | | qa-unit | Unit testing — isolated function/method tests with mocking |

Usage Examples

Claude Code

# After running: npx @aktibaba/qa-testing-skills init claude
claude /qa-api        # Generate API tests for your project
claude /qa-ui         # Generate E2E tests
claude /qa-ci         # Setup CI pipeline with quality gates

Cursor / Windsurf

Prompts are auto-loaded as rules. Just ask:

"Use qa-api to generate tests for my Express API"

ChatGPT / Gemini / Any LLM

npx @aktibaba/qa-testing-skills init generic

Then copy-paste the relevant prompt file from prompts/ into your chat.

What Each Prompt Does

Every prompt follows the same pattern:

1. Discovery — Scans your project, detects stack, framework, existing tests
2. Strategy — Designs a risk-based test plan prioritized by business impact
3. Generation — Creates test files, configs, and helpers ready to run
4. Validation — Checks output against a quality checklist
5. Report — Provides run commands, coverage summary, and recommendations

Supported Stacks

• Languages: JavaScript/TypeScript, Python, Go, Java, C#, Ruby, PHP
• Test Frameworks: Playwright, Cypress, Jest, pytest, JUnit, Go test, xUnit, RSpec, Mocha, Vitest, Selenium
• CI/CD: GitHub Actions, GitLab CI, Jenkins, Azure Pipelines, CircleCI, Bitbucket Pipelines
• Performance: k6, Locust, Artillery, JMeter, Gatling
• Security: Semgrep, Trivy, Snyk, ZAP
• Containers: Docker, Podman, nerdctl

How It Works

Each prompt is a self-contained markdown file that instructs any AI to act as a QA architect. No runtime, no dependencies, no API keys needed.

prompts/
├── qa-api.md      # API testing prompt
├── qa-ui.md       # UI/E2E testing prompt
├── qa-env.md      # Docker environment prompt
├── qa-int.md      # Integration testing prompt
├── qa-perf.md     # Performance testing prompt
├── qa-sec.md      # Security testing prompt
├── qa-ci.md       # CI/CD pipeline prompt
├── qa-reg.md      # Regression testing prompt
├── qa-rv.md       # Test review prompt
└── qa-unit.md     # Unit testing prompt

The CLI just copies these files to the right location for your AI tool. That's the whole trick.

License

MIT