Test File Errors
No tests found in your file
This message means that Cypress was unable to find tests in the specified file. You’ll likely get this message if you have an empty test file and have not yet written any tests.
We found an error preparing your test file
This message means that Cypress encountered an error when compiling and/or bundling your test file. Cypress automatically compiles and bundles your test code so you can use ES2015, CoffeeScript, modules, etc.
You’ll typically receive this message due to:
- The file not existing
- A syntax error in the file or one of its dependencies
- A missing dependency
When the error is fixed in your test file, your tests will automatically re-run.
Support File Errors
Support file missing or invalid
The supportFolder
option was removed from Cypress in version 0.18.0
and was replaced by module support and the supportFile
configuration option.
Cypress used to automatically include any scripts in the supportFolder
before your test files. However, automatically including all the files in a certain directory is somewhat magical and unintuitive, and requires creating globals for the purpose of utility functions.
Use modules for utility functions
Cypress supports both ES2015 modules and CommonJS modules. You can import/require npm modules as well as local modules:
import _ from 'lodash' import util from './util' it('uses modules', function () { expect(_.kebabCase('FooBar')).to.equal('foo-bar') expect(util.secretCode()).to.equal('1-2-3-4') })
Use supportFile to load scripts before your test code
It’s still useful to load a setup file before your test code. If you are setting Cypress defaults or utilizing custom Cypress commands, instead of needing to import/require those defaults/commands in every test file, you can use the supportFile
configuration option.
To include code before your test files, set the supportFile
path. By default, supportFile
is set to look for one of the following files:
cypress/support/index.js
cypress/support/index.coffee
Just like with your test files, the supportFile
can use ES2015+ (or CoffeeScript) and modules, so you can import/require other files as needed.
Command Errors
Cypress cannot execute commands outside a running test
This message means you tried to execute one or more Cypress commands outside of a currently running test. Cypress has to be able to associate commands to a specific test.
Typically this happens accidentally, like in the following situation.
describe('Some Tests', function () { it('is true', function () { expect(true).to.be.true // yup, fine }) it('is false', function () { expect(false).to.be.false // yup, also fine }) context('some nested tests', function () { // oops you forgot to write an it(...) here! // these cypress commands below // are running outside of a test and cypress // throws an error cy.visit('http://localhost:8080') cy.get('h1').should('contain', 'todos') }) })
Move those Cypress commands into an it(...)
block and everything will work correctly.
If you are purposefully writing commands outside of a test, there is probably a better way to accomplish what you’re trying to do. Read through the Examples, chat with someone in our chat, or Open an issue.
cy...()
failed because the element you are chaining off of has become detached or removed from the dom
Getting this errors means you’ve tried to interact with a “dead” DOM element - meaning it’s been detached or completely removed from the DOM.
Cypress errors because it can’t interact with “dead” elements - just like a real user could not do this either. Understanding how this happens is very important - and it is often easy to prevent.
Let’s take a look at an example below.
Application HTML
<body> <div id="parent"> <button>delete</button> </div> </body>
Application JavaScript
$('button').click(() => { // when the <button> is clicked // we remove the button from the DOM $(this).remove() })
Test Code causing error
cy.get('button').click().parent()
We’ve programmed our application above so that as soon as the click
event happens, the button is removed from the DOM. When Cypress begins processing the next command (.parent()
) in the test above, it detects that the yielded subject (the button) is detached from the DOM and throws the error.
We can prevent Cypress from throwing this error by rewriting our test code.
Fixed Test Code
cy.get('button').click() cy.get('#parent')
The above example is an oversimplification. Let’s look at a more complex example.
In modern JavaScript frameworks, DOM elements are regularly re-rendered - meaning that the old element is thrown away and a new one is put in its place. Because this happens so fast, it may appear as if nothing has visibly changed to the user. But if you are in the middle of executing test commands, it’s possible the element you’re interacting with has become “dead”. To deal with this situation you must:
- Understand when your application re-renders
- Re-query for newly added DOM elements
- Guard Cypress from running commands until a specific condition is met
When we say guard, this usually means:
- Writing an assertion
- Waiting on an XHR
cy....()
failed because the element cannot be interacted with
You may see a variation of this message for 4 different reasons:
- The element is not visible
- The element is being covered by another element
- The element’s center is hidden from view
- The element is disabled
Cypress runs several calculations to ensure an element can actually be interacted with like a real user would. If you’re seeing this error, the solution is often obvious. You may need to guard your commands (due to a timing or an animation issue).
There have been situations where Cypress does not correctly allow you to interact with an element that should be interactable. If that’s the case, Open an issue.
If you’d like to override these built-in checks, provide the {force: true}
option to the action itself. Refer to each command for their available options, additional use cases, and argument usage.
Ignore built-in error checking
cy.get('[disabled]').click({force: true}).
Be careful with this option. It’s possible to force your tests to pass when the element is actually not interactable in your application.
cy....()
failed because the element is currently animating
By default Cypress detects if an element you’re trying to interact with is animating. This check ensures that an element is not animating too quickly for a real user to interact with the element. This also prevents some edge cases where actions, such as .type()
or .click()
, happened too fast during a transition.
Cypress will continuously attempt to interact with the element until it eventually times out. If you’d like to force Cypress to interact with the element there are a few options:
- Pass
{force: true}
. This disables all error checking - Pass
{waitForAnimations: false}
to disable animation error checking - Pass
{animationDistanceThreshold: 20}
to decrease the sensitivity of detecting if an element is animating. By increasing the threshold this enables your element to move farther on the page without causing Cypress to continuously retry.
cy.get('#modal button').click({ waitForAnimations: false })
You can globally disable animation error checking, or increase the threshold by modifying the configuration in your configuration.
cypress.json
{ "waitForAnimations": false, "animationDistanceThreshold": 50 }
The test has finished but Cypress still has commands in its queue
Let’s examine several different ways you may get this error message. In every situation, you’ll need to change something in your test code to prevent the error.
Flaky tests below!Several of these tests are dependent on race conditions. You may have to run these tests multiple times before they will actually fail. You can also try tweaking some of the delays.
Simple Example
This first test below will pass and shows you that Cypress tries to prevent leaving commands behind in the queue in every test.
Even though we return a string in our test, Cypress automatically figures out that you’ve queued commands above and does not end the test until all cy commands have finished.
// This test passes! it('Cypress is smart and this does not fail', function () { cy.get('body').children().should('not.contain', 'foo') // <- no return here return 'foobarbaz' // <- return here })
The example below will fail because you’ve forcibly terminated the test early with mocha’s done
.
// This test errors! it('but you can forcibly end the test early which does fail', function (done) { cy.get('body') .then(() => { done() // forcibly end test even though there are commands below }) .children() .should('not.contain', 'foo') })
Complex Async Example
What’s happening in this example is that because we have NOT told Mocha this is an asynchronous test, this test will pass immediately then move onto the next test. Then, when the setTimeout
callback function runs, new commands will get queued on the wrong test. Cypress will detect this and fail the next test.
describe('a complex example with async code', function() { it('you can cause commands to bleed into the next test', function() { // This test passes...but... setTimeout(() => { cy.get('body').children().should('not.contain', 'foo') }, 10) }) it('this test will fail due to the previous poorly written test', function() { // This test errors! cy.wait(10) }) })
The correct way to write the above test code is using Mocha’s done
to signify it is asynchronous.
it('does not cause commands to bleed into the next test', function (done) { setTimeout(() => { cy.get('body').children().should('not.contain', 'foo').then(() => { done() }) }, 10) })
Complex Promise Example
In the example below, we forget to return the Promise
in our test. This means the test passes synchronously but our Promise
resolves in the next test.
This also causes the commands to be queued on the wrong test. We will get the error in the next test that Cypress detected it had commands in its command queue.
describe('another complex example using a forgotten "return"', function () { it('forgets to return a promise', function () { // This test passes...but... Cypress.Promise.delay(10).then(() => { cy.get('body').children().should('not.contain', 'foo') }) }) it('this test will fail due to the previous poorly written test', function () { // This test errors! cy.wait(10) }) })
The correct way to write the above test code would be to return our Promise
:
it('does not forget to return a promise', function () { return Cypress.Promise.delay(10).then(() => { return cy.get('body').children().should('not.contain', 'foo') }) })
cy.visit()
failed because you are attempting to visit a second unique domain
See our Web Security documentation.
Cypress.addParentCommand()
/ Cypress.addDualCommand()
/ Cypress.addChildCommand()
has been removed and replaced by Cypress.Commands.add()
In version 0.20.0
, we removed the commands for adding custom commands and replaced them with, what we believe to be, a simpler interface.
Now you can create parent, dual, and child commands using the same Cypress.Commands.add()
command.
Please read our new documentation on writing custom commands.
Cypress detected that you invoked one or more cy
commands in a custom command but returned a different value.
Because cy
commands are asynchronous and are queued to be run later, it doesn’t make sense to return anything else.
For convenience, you can also omit any return value or return undefined
and Cypress will not error.
In versions before 0.20.0
of Cypress we automatically detected this and forced the cy
commands to be returned. To make things less magical and clearer, we are now throwing an error.
Cypress detected that you invoked one or more cy
commands but returned a different value.
Because cy commands are asynchronous and are queued to be run later, it doesn’t make sense to return anything else.
For convenience, you can also omit any return value or return undefined
and Cypress will not error.
In versions before 0.20.0
of Cypress we automatically detected this and forced the cy
commands to be returned. To make things less magical and clearer, we are now throwing an error.
Cypress detected that you returned a promise from a command while also invoking one or more cy commands in that promise.
Because Cypress commands are already promise-like, you don’t need to wrap them or return your own promise.
Cypress will resolve your command with whatever the final Cypress command yields.
The reason this is an error instead of a warning is because Cypress internally queues commands serially whereas Promises execute as soon as they are invoked. Attempting to reconcile this would prevent Cypress from ever resolving.
Cypress detected that you returned a promise in a test, but also invoked one or more cy
commands inside of that promise.
While this works in practice, it’s often indicative of an anti-pattern. You almost never need to return both a promise and also invoke cy
commands.
cy
commands themselves are already promise like, and you can likely avoid the use of the separate Promise.
Passing cy.route({stub: false})
or cy.server({stub: false})
is now deprecated.
You can safely remove: {stub: false}
.
CypressError: Timed out retrying: Expected to find element: ‘…’, but never found it. Queried from element: <…>
If you get this error in a case where the element is definitely visible in the DOM, your document might contain malformed HTML. In such cases, document.querySelector()
will not find any elements that appear after the point where the HTML is malformed. Even if you feel certain your HTML is not malformed anywhere, check it anyway (line by line in the dev tools). Especially if you’ve exhausted all other possibilities.
CLI Errors
You passed the --record
flag but did not provide us your Record Key.
You may receive this error when trying to run Cypress tests in Continuous Integration. This means that you did not pass a specific record key to: cypress run --record
.
Since no record key was passed, Cypress checks for any environment variable with the name CYPRESS_RECORD_KEY
. In this case, that was also not found.
You can get your project’s record key by locating it in your settings tab in the Test Runner or in the Dashboard Service.
You will want to then add the key to your config file or as an environment variable.
The cypress ci
command has been deprecated
As of version 0.19.0
and CLI versions 0.13.0
, the cypress ci
command has been deprecated. We did this to make it clearer what the difference was between a regular test run and a recorded test run.
Previously to record runs you had the environment variable: CYPRESS_CI_KEY
or you wrote:
cypress ci abc-key-123
You need to rewrite this as:
cypress run --record --key abc-key-123
If you were using the environment variable CYPRESS_CI_KEY
, rename it toCYPRESS_RECORD_KEY
.
You can now run and omit the --key
flag:
cypress run --record
We will automatically apply the record key environment variable.
A Cached Cypress Binary Could not be found
This error occurs in CI when using cypress run
without a valid Cypress binary cache installed on the system (on linux that’s ~/.cache/Cypress
).
To fix this error, follow instructions on caching the cypress binary in CI, then bump the version of your CI cache to ensure a clean build.
Incorrect usage of --ci-build-id
flag
You passed the --ci-build-id
flag but did not provide either a --group
or --parallel
flag.
The --ci-build-id
flag is used to either group or parallelize multiple runs together.
Check out our guide on parallelizing runs and when to use the --ci-build-id
option.
The --ci-build-id
, --group
, or --parallel
flags can only be used when recording
You passed the --ci-build-id
, --group
, or --parallel
flag without also passing the --record
flag.
These flags can only be used when recording to the Dashboard Service.
Please review our parallelization documentation to learn more.
We could not determine a unique CI build ID
You passed the --group
or --parallel
flag but we could not automatically determine or generate a ciBuildId
.
In order to use either of these parameters a ciBuildId
must be determined.
The ciBuildId
is automatically detected if you are running Cypress in most CI providers. Please review the natively recognized environment variables for your CI provider.
You can avoid this check in the future by passing an ID to the --ci-build-id
flag manually.
Please review our parallelization documentation to learn more.
Group name has already been used for this run
You passed the --group
flag, but this group name has already been used for this run.
If you are trying to parallelize this run, then also pass the --parallel
flag, else pass a different group name.
Please review grouping test runs documentation to learn more.
Cannot parallelize tests across environments
You passed the --parallel
flag, but we do not parallelize tests across different environments.
This machine is sending different environment parameters than the first machine that started this parallel run.
In order to run in parallel mode each machine must send identical environment parameters such as:
- Specs
- Operation system name
- Operating system version
- Browser name
- Major browser version
Please review our parallelization documentation to learn more.
Cannot parallelize tests in this group
You passed the --parallel
flag, but this run group was originally created without the --parallel
flag.
You cannot use the --parallel
flag with this group.
Please review our grouping test runs documentation to learn more.
Run must pass --parallel
flag
You did not pass the --parallel
flag, but this run’s group was originally created with the --parallel
flag.
You must use the --parallel
flag with this group.
Please review our parallelization documentation to learn more.
Cannot parallelize tests on a stale run
You are attempting to pass the --parallel
flag to a run that was completed over 24 hours ago.
You cannot run tests on a run that has been complete for that long.
Please review our parallelization documentation to learn more.
Run is not accepting any new groups
The run you are attempting access to is already complete and will not accept new groups.
When a run finishes all of its groups, it waits for a configurable set of time before finally completing. You must add more groups during that time period.
Please review our parallelization documentation to learn more.
Page Load Errors
Cypress detected a cross-origin error happened on page load
For a more thorough explanation of Cypress’s Web Security model, please read our dedicated guide to it.
This error means that your application navigated to a superdomain that Cypress was not bound to. Initially when you cy.visit()
, Cypress changes the browser’s URL to match the url
passed to cy.visit()
. This enables Cypress to communicate with your application to bypasses all same-origin security policies among other things.
When your application navigates to a superdomain outside of the current origin-policy, Cypress is unable to communicate with it, and thus fails.
There are a few simple workarounds to these common situations:
-
Don’t click
<a>
links in your tests that navigate outside of your application. Likely this isn’t worth testing anyway. You should ask yourself: What’s the point of clicking and going to another app? Likely all you care about is that thehref
attribute matches what you expect. So make an assertion about that. You can see more strategies on testing anchor links in our “Tab Handling and Links” example recipe. -
You are testing a page that uses Single sign-on (SSO). In this case your web server is likely redirecting you between superdomains, so you receive this error message. You can likely get around this redirect problem by using
cy.request()
to manually handle the session yourself.
If you find yourself stuck and can’t work around these issues you can just set this in your cypress.json
file. But before doing so you should really understand and read about the reasoning here.
// cypress.json { "chromeWebSecurity": false }
Cypress detected that an uncaught error was thrown from a cross-origin script.
Check your Developer Tools Console for the actual error - it should be printed there.
It’s possible to enable debugging these scripts by adding the crossorigin
attribute and setting a CORS
header.
Browser Errors
The Chromium Renderer process just crashed
Browsers are enormously complex pieces of software, and from time to time they will inconsistently crash for no good reason. Crashes are just a part of running automated tests.
At the moment, we haven’t implemented an automatic way to recover from them, but it is actually possible for us to do so. We have an open issue documenting the steps we could take to restart the renderer process and continue the run. If you’re seeing consistent crashes and would like this implemented, please leave a note in the issue.
If you are running Docker
there is a simple one line fix for this problem documented here.
Test Runner errors
Cannot connect to API server
Logging in, viewing runs, and setting up new projects to record requires connecting to an external API server. This error displays when we failed to connect to the API server.
This error likely appeared because:
- You do not have internet. Please ensure you have connectivity then try again.
- You are a developer that has forked our codebase and do not have access to run our API locally. Please read more about this in our contributing doc.
Cypress detected policy settings on your computer that may cause issues
When Cypress launches Chrome, it attempts to launch it with a custom proxy server and browser extension. Certain group policies (GPOs) on Windows can prevent this from working as intended, which can cause tests to break.
If your administrator has set any of the following Chrome GPOs, it can prevent your tests from running in Chrome:
- Proxy policies:
ProxySettings, ProxyMode, ProxyServerMode, ProxyServer, ProxyPacUrl, ProxyBypassList
- Extension policies:
ExtensionInstallBlacklist, ExtensionInstallWhitelist, ExtensionInstallForcelist, ExtensionInstallSources, ExtensionAllowedTypes, ExtensionAllowInsecureUpdates, ExtensionSettings, UninstallBlacklistedExtensions
Here are some potential workarounds:
- Ask your administrator to disable these policies so that you can use Cypress with Chrome.
- Use the built-in Electron browser for tests, since it is not affected by these policies. See the guide to launching browsers for more information.
- Try using Chromium instead of Google Chrome for your tests, since it may be unaffected by GPO. You can download the latest Chromium build here.
- If you have Local Administrator access to your computer, you may be able to delete the registry keys that are affecting Chrome. Here are some instructions:
- Open up Registry Editor by pressing WinKey+R and typing
regedit.exe
- Look in the following locations for the policy settings listed above:
HKEY_LOCAL_MACHINE\Software\Policies\Google\Chrome
HKEY_LOCAL_MACHINE\Software\Policies\Google\Chromium
HKEY_CURRENT_USER\Software\Policies\Google\Chrome
HKEY_CURRENT_USER\Software\Policies\Google\Chromium
- Delete or rename any policy keys found. Make sure to back up your registry before making any changes.
- Open up Registry Editor by pressing WinKey+R and typing
Uncaught exceptions from your application
WIP. We’ll be adding more here soon.
For now, please visit the Catalog of Events page for examples how to turn off catching uncaught exceptions.