# Qa Testing Kit

> >

- **Type:** Skill
- **Install:** `agentstack add skill-softwareonehn-qa-testing-kit-qa-testing-kit`
- **Verified:** Yes — security-reviewed for prompt injection and unsafe behavior
- **Seller:** [SoftwareOneHN](https://agentstack.voostack.com/s/softwareonehn)
- **Installs:** 0
- **Category:** [Web & Browser](https://agentstack.voostack.com/c/web-and-browser)
- **Latest version:** 0.1.0
- **License:** MIT
- **Upstream author:** [SoftwareOneHN](https://github.com/SoftwareOneHN)
- **Source:** https://github.com/SoftwareOneHN/qa-testing-kit

## Install

```sh
agentstack add skill-softwareonehn-qa-testing-kit-qa-testing-kit
```

Requires the [AgentStack CLI](https://agentstack.voostack.com/docs/cli). Works with Claude Code, Cursor, and any MCP-compatible agent.

## About

# QA Testing Kit

Lifecycle-driven QA workflow — from requirements analysis to test reports, with state tracking and impact analysis.

## QA Lifecycle

```
requirements → test-plan → test-cases → automation → execution → report
```

## Key Capabilities

- Requirements analysis from website/Jira
- Test case design (QUICK mode & FULL RBT 6-step)
- Automation script generation (Playwright, Selenium, Appium)
- Structured test data generation (positive, negative, boundary, combinatorial)
- Flaky test analysis & fix
- Stable locator generation
- Automation framework scaffolding
- Jira/Xray integration (fetch requirements, push results)
- State-aware continuation (knows current position, recommends next step)
- Impact analysis (assess changes before mutations)

## Routing — Intent to Command

When the user says something, route to the correct workflow:

| User Intent | Command | Reference File |
|---|---|---|
| Generate requirements from website | `requirements` | [references/workflow-requirements.md](references/workflow-requirements.md) |
| Generate test cases quickly | `test-cases` (QUICK) | [references/workflow-testcases-quick.md](references/workflow-testcases-quick.md) |
| Full 6-step RBT test cases | `test-cases` (FULL RBT) | [references/workflow-testcases-rbt.md](references/workflow-testcases-rbt.md) |
| Convert TC to automation | `automation` | [references/workflow-automation.md](references/workflow-automation.md) |
| Create framework / scaffold | `framework` | [references/workflow-framework.md](references/workflow-framework.md) |
| Generate test data | `test-data` | [references/workflow-test-data.md](references/workflow-test-data.md) |
| Generate locators | `locator` | [references/workflow-locator.md](references/workflow-locator.md) |
| Flaky test analysis | `flaky-analysis` | [references/workflow-flaky-tests.md](references/workflow-flaky-tests.md) |
| API tests from Swagger | `api-tests` | [references/workflow-api-tests.md](references/workflow-api-tests.md) |
| Cross-module combinatorial | `cross-module` | [references/workflow-cross-module.md](references/workflow-cross-module.md) |
| Fetch from Jira | `jira-fetch` | [references/workflow-jira-integration.md](references/workflow-jira-integration.md) |
| Push results to Xray | `xray-push` | [references/workflow-jira-integration.md](references/workflow-jira-integration.md) |
| Change existing artifact | `impact` | See Impact Analysis below |
| Continue / next step | `next` | See State-Aware Continuation below |

**Priority when ambiguous:**
- Prefer `impact` if user mentions changing an existing artifact
- Prefer `test-cases` QUICK if user says "generate TC" without "6-step" or "RBT"
- Prefer `automation` if user mentions "Playwright" or "Selenium" with "test"

## Defaults

- Language: Vietnamese for explanations, English for code
- Architecture: Page Object Model (POM) mandatory
- Viewport: 1920×1080 (desktop debug)
- Wait strategy: Smart waits ONLY (NEVER use Thread.sleep / waitForTimeout)
- Test data: Random + traceable (`__`)

## Output Directory Structure

```
qa-output/{slug}/
├── PROJECT.md
├── project-memory.md
├── 01_requirements/requirements.md
├── 02_test-plan/test-plan.md
├── 03_test-cases/{module}/test-cases.md
├── 04_automation/
│   ├── locator-collection.md
│   └── src/pages/, tests/, utils/
├── 05_execution/task.md, results.md
└── 06_report/report.md
```

## Behavior Rules

### Fail-Closed
- Never silently choose slug or module. Ask if ambiguous.
- If prerequisite missing → print exact path + command, then stop.

### Overwrite Protection
Before mutating existing artifacts:
1. Check if target path exists
2. If exists → ask before overwriting
3. No explicit approval → stop

### Impact-First Rule
When user requests changes to existing requirements or test cases → run impact analysis FIRST.
Only mutate after user approves.

### Auto-Heal (Rule E3)
When test FAILS during execution:
- Self-read log → analyze → fix → re-run (max 5 rounds)
- NEVER ask user during fix process
- Only ask when: business rule conflict, server inaccessible, or 5 rounds exhausted

### Security
- NEVER read .env files directly
- NEVER print credentials
- NEVER commit files containing credentials
- Credentials only via environment variables

## State-Aware Continuation

When user says "continue" or "next step":
1. Inspect `qa-output/{slug}/` directory
2. Check which artifacts exist
3. Find first incomplete phase in lifecycle order
4. Recommend next action with context

## Impact Analysis

When user requests changes to existing artifacts:
1. Capture change request
2. Trace downstream impact (requirements → test-cases → automation → test-data)
3. Present impact matrix with risk level
4. Wait for user approval before mutating

## Tech Stack

| Platform | Framework | Language | Runner |
|---|---|---|---|
| Web | Playwright | TypeScript/Java/Python | Playwright Test/TestNG/Pytest |
| Web | Selenium | Java/Python | TestNG/Pytest |
| Mobile | Appium | Java | TestNG |
| API | REST Assured | Java | TestNG |
| API | Playwright API | TypeScript | Playwright Test |

## Framework-Specific Rules

Load the appropriate reference when working with a specific framework:
- Playwright: [references/rules-playwright.md](references/rules-playwright.md)
- Selenium: [references/rules-selenium.md](references/rules-selenium.md)
- Appium: [references/rules-appium.md](references/rules-appium.md)
- General automation: [references/rules-automation.md](references/rules-automation.md)
- Locator strategy: [references/rules-locator-strategy.md](references/rules-locator-strategy.md)

## MCP Server (Jira/Xray)

The skill includes an MCP server for Jira/Xray integration. See [references/workflow-jira-integration.md](references/workflow-jira-integration.md) for setup and usage.

Tools provided:
- `fetch_jira_issue` — Fetch single Jira issue by key
- `fetch_jira_issues` — Fetch multiple issues (project/type/JQL)
- `fetch_epic_children` — Fetch Epic children
- `xray_authenticate` — Authenticate with Xray Cloud
- `xray_import_results` — Import test results to Xray
- `test_jira_connection` — Test Jira API connectivity

## Source & license

This open-source skill is cataloged on AgentStack and links to its original source — we do not rehost the code.

- **Author:** [SoftwareOneHN](https://github.com/SoftwareOneHN)
- **Source:** [SoftwareOneHN/qa-testing-kit](https://github.com/SoftwareOneHN/qa-testing-kit)
- **License:** MIT

Install and usage instructions live in the source repository linked above.

## Pricing

- **Free** — Free

## Security capabilities

Automated source analysis of v0.1.0 — what this tool can access:

- **Network access:** no
- **Filesystem access:** no
- **Shell / process execution:** no
- **Environment & secrets:** yes
- **Dynamic code execution:** no

*"Yes" means the capability is present in the source — more access means more to trust, not that it is unsafe.*


## Versions

- **0.1.0** — security scan: passed — Imported from the upstream source.

## Links

- Listing page: https://agentstack.voostack.com/l/skill-softwareonehn-qa-testing-kit-qa-testing-kit
- Seller: https://agentstack.voostack.com/s/softwareonehn
- Browse the marketplace: https://agentstack.voostack.com/browse

---
Listed on AgentStack — the marketplace for AI agent skills and MCP servers. Every listing is security-reviewed. Creators keep 70%.