From Selenium .NET

This guide ports upstream’s docs/migrate/selenium.mdx (tester-army/e2e at 7b1b8b27, tester-army/e2e#1441). Upstream’s example is Java and its target is TypeScript. This guide maps Selenium WebDriver for .NET (Selenium.WebDriver, with NUnit or xUnit) to C# e2e tests.

e2e replaces a Selenium suite with C# tests. Locators find nodes by role and label, and assertions retry until they pass or time out. Run both suites and move one flow at a time.

Agent prompt

Migrate this project's Selenium .NET tests to e2e, one flow at a time. Docs: the "From Selenium .NET" page of the e2e-dotnet docs site (docs-site/src/content/docs/migrate/selenium.mdx).

1. Read the migration guide. It maps Selenium config and WebDriver calls to e2e.
2. e2e tests are NUnit or xUnit v3 tests. If the Selenium suite uses another language, ask me where the new test project should live.
3. If the solution has no e2e test project, run `e2e init <directory> --yes` beside the Selenium project (add `--framework xunit` for xUnit v3). It creates the test project and example test, writes e2e.config.json, and installs the skill. In an existing test project it adds the missing package references to the .csproj.
4. Read the e2e skill (`e2e guide` prints it, `e2e guide <topic>` prints a topic) and follow it for every step below.
5. If you ran init, ask me which model to use for agent steps. If it needs a sign-in, give me the `e2e login <provider>` command and wait until I confirm.
6. Set the target's app.url to the URL every driver.Navigate().GoToUrl used, and app.command to the command that starts the app if the suite starts it. Move passwords and tokens into credentials or secrets.
7. List the page objects and the flows that use them, and ask me which to move first. Suggest one page object and its flows.
8. Rewrite those flows as test classes that derive from E2ETest, with the guide's tables. Prefer role, label, and test id queries over CSS and XPath. Replace every wait and Thread.Sleep with a locator assertion. Narrow queries that match several nodes. Turn the page object's setup code into a [SetUp] (NUnit) or InitializeAsync (xUnit).
9. Keep navigation and the final assertions deterministic, with App.OpenAsync and Expect.That. Where a run of scripted interactions breaks often, propose an Agent.ActAsync goal for it instead.
10. Run the new classes with `dotnet test --filter` and fix them until they pass. Keep the Selenium tests until I confirm they can go.
11. Summarize which flows moved, which stayed in Selenium, and why.

For a migration that spans several sessions, the migrate-to-e2e skill records what moved in a progress file and keeps the old suite running until you retire each test. See Coding agents.

Port the config

Install e2e with the web guide. Only agent calls need a model, so the example below runs without one.

Selenium	e2e
The URL in every `Navigate().GoToUrl`	`app.url` on the target, then relative paths
Starting the app in a script	`app.url` with `app.command`, see Web
One driver per browser	One target per browser, `WebEngineOptions.BrowserName = BrowserName.Firefox`
Selenium Grid	The framework’s parallelism; `WebEngineOptions.Connect` attaches to a remote Chromium over CDP, and a browser provider leases hosted browsers
`Manage().Window.Size`	`WebEngineOptions.Viewport`
Implicit waits	`actionTimeout` and `assertionTimeout` set action and assertion deadlines
A page object for sign-in	`[E2ESetup]` and `[E2ESession]`, see Signing in

Rewrite one flow

Start with one page object and the flows that use it.

driver.Navigate().GoToUrl("http://localhost:3000/login");
var wait = new WebDriverWait(driver, TimeSpan.FromSeconds(10));
wait.Until(ExpectedConditions.ElementIsVisible(By.Id("username"))).SendKeys("admin@example.test");
driver.FindElement(By.Id("password")).SendKeys(Environment.GetEnvironmentVariable("ADMIN_PASSWORD"));
driver.FindElement(By.XPath("//button[contains(., 'Sign in')]")).Click();
wait.Until(ExpectedConditions.UrlContains("/dashboard"));
Assert.That(driver.FindElement(By.CssSelector(".greeting")).Text, Is.EqualTo("Welcome back, admin!"));

using E2E.NUnit;

public sealed class SignInTests : E2ETest
{
    [Test]
    public async Task Signs_in()
    {
        var admin = Credentials.User("admin");
        await App.OpenAsync("/login");
        await Screen.GetByLabel("Username").FillAsync(admin.Username);
        await Screen.GetByLabel("Password").FillAsync(admin.Password);
        await Screen.GetByRole("button", "Sign in").TapAsync();
        await Expect.That(Browser).ToHaveURLAsync("/dashboard");
        await Expect.That(Screen.GetByRole("status")).ToHaveTextAsync("Welcome back, admin!");
    }
}

For xUnit v3, use E2E.XUnit.V3 and [Fact]. The body is the same. Locator actions wait for their targets, so no explicit wait is left. The password comes from E2E_USER_ADMIN_USERNAME and E2E_USER_ADMIN_PASSWORD or the credentials key, see Signing in.

Map each WebDriver call

Selenium	e2e
`driver.Navigate().GoToUrl(url)`	`App.OpenAsync(path)`
`driver.Navigate().Back()`, `.Refresh()`	`App.BackAsync()`, `Browser.ReloadAsync()`
`By.Id("x")`	`Browser.Locator("#x")`; or `WebEngineOptions.TestIdAttribute = "id"` and `Screen.GetByTestId("x")`
`By.Name("x")`	`Browser.Locator("[name=\"x\"]")`, not portable; prefer a role or label
`By.LinkText`, `By.PartialLinkText`	`Screen.GetByRole("link", name)`, `exact: false` for part of the name
`By.CssSelector`, `By.XPath`	`Browser.Locator("css or xpath")`, not portable
`FindElements(...).Count`	`await locator.CountAsync()`
`FindElements(...)[n]`	`locator.Nth(n)`
`.Click()`	`.TapAsync()` or `.ClickAsync()`
`.SendKeys("text")`	`.FillAsync("text")` replaces the content
`.SendKeys(Keys.Enter)`	`.PressAsync("Enter")`
`.Clear()`	`.ClearAsync()`
`new SelectElement(el).SelectByText`	`.SelectOptionAsync("label")`
`.Text`, `.GetAttribute()`, `.Displayed`	`.TextContentAsync()`, `.GetAttributeAsync()` wait for the node then read one time, `.IsVisibleAsync()` reads at once; `ToHaveTextAsync()` retries
`wait.Until(ElementIsVisible(...))`	`await Expect.That(locator).ToBeVisibleAsync()`
`wait.Until(d => el.Text == "x")`	`await Expect.That(locator).ToHaveTextAsync("x")`
`wait.Until(UrlContains(...))`	`await Expect.That(Browser).ToHaveURLAsync(new Regex("..."))`
`new Actions(driver).MoveToElement(el)`	`.HoverAsync()`
`Actions.DragAndDrop(a, b)`	`a.DragToAsync(b)`
`Actions.KeyDown(Keys.Control).SendKeys("a")`	`.PressAsync("ControlOrMeta+a")`
`SwitchTo().Frame()`	`Browser.FrameLocator("#frame")`
`SwitchTo().Alert().Accept()`	`await Browser.OnDialogAsync(DialogHandler.Accept)` before the action
`SwitchTo().Window()`	No equivalent. Popups are unsupported
`ExecuteScript`	`Browser.EvaluateAsync<T>`
`Manage().Cookies.AddCookie`	`Browser.SetCookiesAsync`
`GetScreenshot()`	`App.ScreenshotAsync("label")`, redacted
`Thread.Sleep`	Use a locator assertion for the condition you need
Page objects	A class that takes `Screen` and `Browser`, or a goal for `Agent.ActAsync`, see Migrating from Playwright .NET

FindElement takes the first match. e2e fails with LOCATOR_AMBIGUOUS on several, so narrow with Filter(...), First(), or Nth(n).

Core concepts: goals, assertions, and locators instead of scripted selectors.
Working with the browser: hosted browsers and CDP recovery.

From Selenium .NET

Agent prompt

Port the config

Rewrite one flow

Map each WebDriver call

Next