To configure different scan options, use the chainable methods of the AxeBuilder (reference docs):

For a complete working sample project that uses this library, see the C# sample in microsoft/axe-pipelines-samples.

AxeBuilder Reference

AxeBuilder.Analyze()

Runs an axe accessibility scan of the entire page using all previously chained options and returns an AxeResult representing the scan results.

AxeBuilder.Analyze(IWebElement element)

Runs an axe accessibility scan scoped using all previously chained options to the given Selenium IWebElement. Returns an AxeResult representing the scan results.

Not compatible with AxeBuilder.Include or AxeBuilder.Exclude; the element passed to Analyze will take precedence and the Include/Exclude calls will be ignored.

AxeBuilder.Include(string cssSelector)

Scopes future Analyze() calls to include only the element(s) matching the given CSS selector.

Include may be chained multiple times to include multiple selectors in a scan.

Include may be combined with Exclude to scan a tree of elements but omit some children of that tree. For example:

Include is not compatible with Analyze(IWebElement) - the Analyze argument will take precedence and Include will be ignored.

This overload of Include only supports CSS selectors which refer to elements which are in the topmost frame of the page and not contained within a shadow DOM. To specify a selector in an iframe or a shadow DOM, see the overload that accepts an AxeSelector.

AxeBuilder.Include(AxeSelector axeSelector)

Scopes future Analyze() calls to include only the element(s) matching the given AxeSelector.

Include may be chained multiple times to include multiple selectors in a scan.

Include may be combined with Exclude to scan a tree of elements but omit some children of that tree. For example:

This overload of Include supports complex AxeSelectors which specify elements inside iframes and/or shadow DOMs:

AxeBuilder.Exclude(string cssSelector)

Scopes future Analyze() calls to exclude the element(s) matching the given CSS selector.

Exclude may be chained multiple times to exclude multiple selectors in a scan.

Exclude may be combined with Include to scan a tree of elements but omit some children of that tree. For example:

Exclude is not compatible with Analyze(IWebElement) - the Analyze argument will take precedence and Exclude will be ignored.

This overload of Include only supports CSS selectors which refer to elements which are in the topmost frame of the page and not contained within a shadow DOM. To specify a selector in an iframe or a shadow DOM, see the overload that accepts an AxeSelector.

AxeBuilder.Exclude(AxeSelector axeSelector)

Scopes future Analyze() calls to exclude the element(s) matching the given CSS selector.

Exclude may be chained multiple times to exclude multiple selectors in a scan.

Exclude may be combined with Include to scan a tree of elements but omit some children of that tree. For example:

Exclude is not compatible with Analyze(IWebElement) - the Analyze argument will take precedence and Exclude will be ignored.

This overload of Exclude supports complex AxeSelectors which specify elements inside iframes and/or shadow DOMs:

AxeBuilder.WithRules(params string[] axeRuleIds)

Causes future calls to Analyze to only run the specified axe rules.

For a list of the available axe rules, see https://dequeuniversity.com/rules/axe/3.3. The "Rule ID" at the top of each individual rule's page is the ID you would want to pass to this method.

WithRules is not compatible with WithTags or DisableRules; whichever one you specify last will take precedence.

WithRules is not compatible with the deprecated raw Options property.

AxeBuilder.DisableRules(params string[] axeRuleIds)

Causes future calls to Analyze to omit the specified axe rules.

For a list of the available axe rules, see https://dequeuniversity.com/rules/axe/3.3. The "Rule ID" at the top of each individual rule's page is the ID you would want to pass to this method.

DisableRules is compatible with WithTags; you can use this to run all-but-some of the rules for a given set of tags. For example, to run all WCAG 2.0 A rules except for color-contrast:

DisableRules is not compatible with WithRules; whichever one you specify second will take precedence.

DisableRules is not compatible with the deprecated raw Options property.

AxeBuilder.WithTags(params string[] axeRuleTags)

Causes future calls to Analyze to only run axe rules that match at least one of the specified tags.

A "tag" is a string that may be associated with a given axe rule. See the axe-core API documentation for a complete list of available tags.

WithTags is compatible with DisableRules; you can use this to run all-but-some of the rules for a given set of tags. For example, to run all WCAG 2.0 A rules except for color-contrast:

WithTags is not compatible with WithRules; whichever one you specify second will take precedence.

WithTags is not compatible with the deprecated raw Options property.

AxeBuilder.WithOptions(AxeRunOptions options)

Note: in most cases, the simpler WithRules, WithTags, and DisableRules can be used instead.

Causes future calls to Analyze to use the specified options when calling axe.run in axe-core. See the axe-core API documentation for descriptions of the different properties of AxeRunOptions.

WithOptions will override any values previously set by WithRules, WithTags, and DisableRules.

WithOptions is not compatible with the deprecated raw Options property.

Skip iFrames

If you don't want to run Axe on iFrames you can tell Axe skip with AxeRunOptions.

Prevents Axe core from getting injected into page iFrames.

Causes future calls to Analyze to use the specified options when calling axe.run in axe-core. See the axe-core API documentation for descriptions of the different properties of AxeRunOptions.

AxeBuilder.WithOutputFile(string filePath)

Causes future calls to Analyze to export their results to a JSON file, in addition to being returned as an AxeResult object as usual.

The output format is exactly the same as axe-core would have produced natively, and is compatible with other tools that read axe result JSON, like axe-sarif-converter.

AxeBuilder.AxeBuilder(webDriver, axeBuilderOptions)

This constructor overload enables certain advanced options not required by most Deque.AxeCore.Selenium users. Currently, its only use is to allow you to use a custom axe-core implementation, rather than the one that is packaged with this library.

AxeBuilder.UseLegacyMode(boolean legacyMode = true)

Set the frame testing method to "legacy mode". In this mode, axe will not open a blank page in which to aggregate its results. This can be used in an environment where opening a blank page causes issues.

With legacy mode turned on, axe will fall back to its test solution prior to the 4.3 release, but with cross-origin frame testing disabled. The frame-tested rule will report which frames were untested.

Important: Use of .UseLegacyMode() is a last resort. If you find there is no other solution, please report this as an issue. It will be removed in a future release.

Working with AxeResult objects

In most cases, you would run an axe scan from within a test method in a suite of end to end tests, and you would want to use a test assertion to verify that there are no unexpected accessibility violations in a page or component.

We strongly recommend FluentAssertions, a NuGet package that does a good job of producing actionable error messages based on the AxeResult output from a scan. That said, if you prefer a different assertion library, you can still use this library; it does not require any particular test or assertion framework.

If you start with no accessibility issues in your page, you can stay clean by validating that the Violations list is empty:

If you already have some accessibility issues & you want to make sure that you do not introduce any more new issues till you get to a clean state, you can use Exclude to remove problematic elements from a broad scan, and a combination of Include and DisableRules to perform a more scoped scan of the element with known issues:

Migrating from Selenium.Axe (SeleniumAxeDotnet)

This project acts as a drop-in replacement for most of the functionality from the Selenium.Axe NuGet package (SeleniumAxeDotnet). To migrate:

1. Update your test .csproj file's <PackageReference> for Selenium.Axe to Deque.AxeCore.Selenium
2. Add a new <PackageReference> to Deque.AxeCore.Commons at the same version number as Deque.AxeCore.Selenium
3. Update all using Selenium.Axe; statements in your tests to using Deque.AxeCore.Selenium; and/or using Deque.AxeCore.Commons;

This project does not include a replacement for Selenium.Axe's built-in HTML report functionality. We expect it to be split out into a separate standalone library (usable from either this package or Deque.AxeCore.Playwright) at a later date, but there is currently no direct replacement.

Finally, there are a few minor breaking changes which won't impact most Selenium.Axe users, but which may require updates if you use certain advanced features:

1. The .Target and .XPath properties of AxeResultNode or AxeResultRelatedNode are now strongly-typed AxeSelector objects. Most Selenium.Axe users do not refer to these properties explicitly, but if your tests do, you will probably want to do so via their ToString() representations.
2. AxeRunOptions.FrameWaitTimeInMilliseconds was renamed to AxeRunOptions.FrameWaitTime to match the equivalent axe-core API. The usage is unchanged; it still represents a value in milliseconds.
3. AxeResult.TestEngineName and AxeResult.TestEngineVersion were replaced by a separate AxeTestEngine object containing Name and Version properties. You will have to replace usages of AxeResult.TestEngineName and AxeResult.TestEngineVersion with AxeResult.TestEngine.Name and AxeResult.TestEngine.Version, respectively.
4. The already-deprecated AxeBuilder.Options property was removed; replace it with WithOptions, WithRules, WithTags, and/or DisableRules.
5. The AxeResult.Error property was removed; any errors that would have appeared here are instead reported as exceptions.
6. The AxeBuilder.Include and AxeBuilder.Exclude overloads which accept more than one parameter have changed:
   • If you were using it to refer to an element inside a nested frame, replace it with the Include/Exclude overloads which accept an AxeSelector: //old new AxeBuilder(webDriver).Include("#some-iframe", "#element-in-nested-frame").Analyze(); // new new AxeBuilder(webDriver).Include(new AxeBuilder("#element-in-nested-frame", new List<string> { "#some-iframe" })).Analyze();
   • If you were using it to refer to multiple elements in the main frame of the page under test, replace it with multiple Include/Exclude calls (one per element): // old // This wasn't doing what you meant it to and may have been hiding issues; this // would have looked for #bar elements inside an iframe with selector #foo, *not* // for sibling #foo and #bar elements in the main frame of a page. new AxeBuilder(webDriver).Include("#foo", "#bar").Analyze(); // new new AxeBuilder(webDriver).Include("#foo").Include("#bar").Analyze();

Contributing

Refer to the general axe-core-nuget CONTRIBUTING.md.

License

This package is distributed under the terms of the MIT License.

However, note that it has a dependency on the (MPL licensed) Deque.AxeCore.Commons NuGet package.

Acknowledgements

This package builds on past work from the SeleniumAxeDotnet project (see NOTICE.txt). We thank all of its contributors for their work.