Now liveThe Skillselion MCP - thousands of ranked skills, loaded into your agent mid-task. No install.Get it →
Kokoszsz avatar

Qa

  • Updated March 29, 2026
  • Kokoszsz/qa-plugin

qa is a Claude Code skill for ai & agent building. It helps developers move faster with AI-assisted coding.

Key points

  • qa
  • AI & Agent Building
  • AI-coding skill

Qa by the numbers

  • Data as of Jul 7, 2026 (Skillselion catalog sync)
/plugin marketplace add Kokoszsz/qa-plugin
/plugin install qa@qa-plugin

Add your badge

Show developers this skill is listed on Skillselion. Paste this into your README.

Listed on Skillselion
Last updatedMarch 29, 2026
RepositoryKokoszsz/qa-plugin

What it does

Helps with ai & agent building tasks.

README.md

QA Plugin for Claude Code

A Claude Code plugin that analyzes any project and generates tailored QA testing infrastructure into .claude/.

Installation

/plugin marketplace add Kokoszsz/qa-plugin
/plugin install qa@qa-plugin

Quick Start

/qa:init                    # One-time setup — detects stack, picks test types, generates everything
/qa:scenarios               # Generate Gherkin user scenarios
"Write unit tests for X"    # Use generated skills to write tests
/test                       # Run tests
/qa:analyze                 # Find gaps
/qa:sync                    # Keep artifacts fresh after changes

How It Works

1. Initialize (/qa:init)

Run once per project. Init will:

  • Detect your language, framework, test runner, coverage tool
  • Scan your code for modules, controllers, repositories, UI routes
  • Recommend test types based on what it finds (unit, integration, e2e, API)
  • Ask you to confirm which test types you want
  • Generate project-specific skills, agents, and commands into .claude/

If your project has no code yet, init generates base skills only. Run /qa:sync later when you have code.

If you run /qa:init again, it tells you to use /qa:sync instead.

2. Generate Scenarios (/qa:scenarios)

/qa:scenarios                        # From code analysis
/qa:scenarios --spec docs/auth.md    # From a spec file (e.g., OpenSpec)

Produces Gherkin scenarios (Given/When/Then) describing what your app does from the user's perspective. These feed into e2e test writing.

3. Write Tests

Talk to Claude naturally — the generated skills handle the rest:

"Write unit tests for the payment service"
"Create integration tests for the user repository"
"Write e2e tests from .claude/scenarios/authentication.feature"
"Add API tests for the /api/orders endpoints"

Each skill knows your project's conventions (framework, naming, location, assertion style).

4. Run Tests (/test)

/test                    # Unit tests (default)
/test --integration      # Integration tests
/test --e2e              # E2E tests
/test --api              # API tests
/test --all              # All enabled types
/test --changed          # Only changed files
/test:coverage           # Coverage against thresholds
/test:review             # Test quality review

5. Analyze (/qa:analyze)

Deep test health check — finds coverage gaps, quality issues, untested flows. Prioritizes what to test next.

6. Keep Fresh (/qa:sync)

Run after project changes (new modules, framework migration, dependency updates). Sync will:

  • Re-detect your stack and compare against saved config
  • Show what's outdated, stale, or missing
  • Suggest enabling new test types if it detects new signals (e.g., new controllers → API tests)
  • Regenerate affected artifacts after your confirmation

Commands

Command Purpose
/qa:init One-time setup — detect stack, select test types, generate artifacts
/qa:scenarios Generate Gherkin user scenarios for e2e testing
/qa:sync Check health, update config, regenerate outdated artifacts
/qa:analyze Deep analysis of test coverage gaps and quality issues

What Gets Generated

Skills

Always generated:

  • qa-use-case-identifier — Maps code to testable user flows
  • qa-scenario-writer — Generates Gherkin scenarios
  • qa-coverage-analyzer — Interprets coverage reports
  • qa-test-reviewer — Reviews test quality

Per selected test type:

  • qa-test-writer — Unit tests (if unit selected)
  • qa-integration-test-writer — Integration tests (if integration selected)
  • qa-e2e-test-writer — E2E tests from Gherkin scenarios (if e2e selected)
  • qa-api-test-writer — API endpoint tests (if API selected)

Agents

  • qa-test-runner — Runs tests across all enabled types
  • qa-coverage-checker — Runs coverage, compares against thresholds
  • qa-test-reviewer — Reviews changed test files

Commands

  • /test — Run tests with --integration, --e2e, --api, --all, --changed flags
  • /test:coverage — Coverage analysis
  • /test:review — Test quality review

Supported Stacks

TypeScript, JavaScript, Python, Go, Rust, Java, Ruby, PHP, Elixir, Swift, Dart — with framework-specific detection for Next.js, React, Vue, Angular, Svelte, Express, NestJS, Django, Flask, FastAPI, Rails, Spring Boot, and more.

Contributing

npm test                       # Script tests (68 assertions)
npm run test:promptfoo         # Behavioral tests (requires ANTHROPIC_API_KEY)
npm run test:promptfoo:init    # Init tests only
npm run test:promptfoo:sync    # Sync tests only
npm run test:promptfoo:coordinator  # Coordinator tests only
npm run test:promptfoo:view    # View results in browser

License

MIT

Related skills

This week in AI coding

Five minutes, every Monday - the tools, releases and tactics for developers.

unsubscribe anytime.