界面已本地化;技能介绍与说明文档按所选语言由 AI 自动翻译,代码、命令、路径和专业术语保留原文,以原作者文档为准。
探索技能
开发与工程 · SKILL.md

reqnroll-skill

reqnroll-skill

【开发与工程】Generates production-grade Reqnroll BDD automation scripts for web (Selenium 3/4) and mobile (Appium 2) testing in C#. Supports parallel NUnit execution locally and on TestMu AI cloud. Use when the user asks to write BDD tests, automate with Reqnroll, create .feature files, write Gherkin scenarios, write step definitions, migrate from SpecFlow, or te…

独立 GitHub 快照 · 自动分类 · 以原作者文档为准

技能说明

GitHub

name: reqnroll-skill description: > Generates production-grade Reqnroll BDD automation scripts for web (Selenium 3/4) and mobile (Appium 2) testing in C#. Supports parallel NUnit execution locally and on TestMu AI cloud. Use when the user asks to write BDD tests, automate with Reqnroll, create .feature files, write Gherkin scenarios, write step definitions, migrate from SpecFlow, or test on browsers/Android/iOS. Triggers on: "Reqnroll", "BDD", "Gherkin", ".feature file", "step definition", "SpecFlow migration", "Selenium C#", "Appium C#", "TestMu", "LambdaTest", "NUnit BDD", "reqnroll.actions.json". languages:

  • C# category: bdd-testing license: MIT metadata: author: TestMu AI version: "1.0"

Overview

This skill guides QA engineers and test architects in writing production-grade Reqnroll BDD tests for web and mobile automation in C#. It covers three execution paths — Selenium 4 with manual driver management, Selenium 3 via the Reqnroll.Actions plugin, and Appium 2 for Android mobile — all targeting TestMu AI (LambdaTest) cloud infrastructure.

Reqnroll is the actively maintained open-source successor to SpecFlow. Existing SpecFlow projects can migrate by swapping the NuGet package and namespace — no step definition rewrites required.

Key Execution Pathways

Framework Selection: Distinguishes between Selenium 4 (manual DriverFactory), Selenium 3 (Reqnroll.SpecFlowCompatibility.Actions.LambdaTest plugin with IBrowserInteractions), and Appium 2 (Appium.WebDriver, AndroidDriver).

Cloud vs Local: Reads LT_USERNAME and LT_ACCESS_KEY environment variables; routes to hub.lambdatest.com (web) or mobile-hub.lambdatest.com (mobile). Reports pass/fail to LambdaTest via lambda-status JavaScript executor calls in [AfterScenario].

Parallelism: Uses [assembly: Parallelizable(ParallelScope.Fixtures)] with [assembly: LevelOfParallelism(N)] (NUnit). State is shared between step definition classes via ScenarioContext (injected by Reqnroll's DI container), not static fields.

Core Technical Patterns

Feature Files (Gherkin)

Each .feature file maps to one test class. Scenarios are tagged (@tagName) for selective filtering with dotnet test --filter "Category=tagName". Background steps run before every scenario in the file; Scenario Outlines drive data-driven testing via Examples tables.

Step Definitions

Classes are decorated with [Binding]. Constructor injection (via Reqnroll's built-in DI) receives ScenarioContext or shared context objects. One [Binding] class per concern keeps files small. Regex-based step patterns use (.*) or typed captures ((\d+)) — no attribute-level type converters needed for primitives.

Hooks

[BeforeScenario] initialises the driver (stored in ScenarioContext["driver"]) and navigates to the base URL. [AfterScenario] reads _scenarioContext.TestError (web) or TestContext.CurrentContext.Result.Outcome.Status (mobile) to emit lambda-status=passed/failed before driver.Quit().

ScenarioContext Driver Sharing

Drivers are stored as _scenarioContext["driver"] = driver and retrieved with scenarioContext["driver"] as IWebDriver. This is required for parallel execution — static driver fields cause race conditions.

Explicit Waits

WebDriverWait with Until(d => d.FindElement(locator)) replaces ImplicitWait for dynamic content. A WaitAndFind(By) helper method encapsulates the 10-second default; a WaitAndClick(By, int timeout) variant handles clickability.

Cloud Integration (TestMu / LambdaTest)

Web (Selenium 4)

var ltOptions = new Dictionary<string, object>
{
    { "build", "Build Name" },
    { "project", "Project Name" },
    { "w3c", true },
    { "selenium_version", "4.38.0" },
    { "sessionName", scenarioName },
    { "platformName", "Windows 11" }
};
var options = new ChromeOptions();
options.BrowserVersion = "latest";
options.AddAdditionalOption("LT:Options", ltOptions);
var driver = new RemoteWebDriver(
    new Uri(quot;https://{userName}:{accessKey}@hub.lambdatest.com/wd/hub"), options);

Mobile (Appium 2)

var ltOptions = new Dictionary<string, object>
{
    { "build", "Build Name" },
    { "project", "Project Name" },
    { "w3c", true },
    { "app", "proverbial-android" },         // lt:// URI or pre-uploaded alias
    { "platformName", "android" },
    { "deviceName", "Galaxy.*" },
    { "platformVersion", "14" },
    { "isRealMobile", true },
    { "autoAcceptAlerts", true },
    { "autoGrantPermissions", true },
    { "sessionName", scenarioName }
};
var appiumOptions = new AppiumOptions();
appiumOptions.AddAdditionalAppiumOption("LT:Options", ltOptions);
var driver = new AndroidDriver(
    new Uri(quot;https://{userName}:{accessKey}@mobile-hub.lambdatest.com/wd/hub"),
    appiumOptions);

Reporting Pass/Fail

// Web (AfterScenario)
if (_scenarioContext.TestError == null)
    ((IJavaScriptExecutor)driver).ExecuteScript("lambda-status=passed");
else
    ((IJavaScriptExecutor)driver).ExecuteScript("lambda-status=failed");

// Mobile (AfterScenario)
bool passed = TestContext.CurrentContext.Result.Outcome.Status == TestStatus.Passed;
((IJavaScriptExecutor)driver).ExecuteScript("lambda-status=" + (passed ? "passed" : "failed"));

Quality Checkpoints

  • Feature files use descriptive scenario names that double as the LambdaTest session name
  • ScenarioContext used for driver sharing — never static fields in parallel runs
  • [assembly: Parallelizable(ParallelScope.Fixtures)] declared once in any .cs file
  • LT_USERNAME and LT_ACCESS_KEY read from environment — never hardcoded
  • lambda-status=passed/failed emitted in every [AfterScenario] for cloud runs
  • WebDriverWait used throughout — no unconditional Thread.Sleep except transient Appium delays
  • Appium locators use MobileBy.Id (resource-id) or MobileBy.AccessibilityId before XPath
  • driver.Quit() always called in [AfterScenario] to free cloud device slots
  • dotnet test --logger "console;verbosity=detailed" surfaces per-scenario pass/fail

Reference Structure

The reference/ directory contains detailed playbook sections:

FileContents
playbook.mdFull implementation guide: project setup, all three driver modes, parallel execution, CI/CD, debugging table, best practices checklist
cloud-integration.mdLambdaTest capability reference, LT:Options fields, tunnel setup, build/session naming, test observability
selenium-4-patterns.mdSelenium 4 patterns: DriverFactory, multi-browser, ChromeOptions/FirefoxOptions/EdgeOptions, screenshot on failure
selenium-3-patterns.mdSelenium 3 patterns: Reqnroll.SpecFlowCompatibility.Actions.LambdaTest, IBrowserInteractions, reqnroll.actions.json config
appium-patterns.mdAppium 2 patterns: AndroidDriver, AppiumOptions, gesture helpers, MobileBy locators, app lifecycle
我的收藏