1. Docs
  2. Build tests
  3. Automate
  4. Troubleshooting

Troubleshooting

Find the message or symptom you see. Each entry says why it happens, what to check and how to fix it.

17 min read Updated CustomersAdmins

In short

On this page you can

  • Look up an Automate message by its exact text.
  • Find out why a save, recording or dry run was refused.
  • Fix archive, permission and licence refusals.
  • Solve AI generation errors.

How to use it

  1. Search this page for the words of the message (Ctrl+F).
  2. Open the entry and read Why.
  3. Work through Check, then apply Fix.
  4. Still stuck? Write to info@smartestqa.ai with the test case ID and the message.

Test cases, folders and archive#

“Company license limit has been reached. The maximum allowed usage for this operation has been exceeded.”Licence
Why
Your company’s licence can limit how many test cases you create, and your company has reached that limit. The counter counts creations; archiving or deleting does not lower it.
Check
An Admin sees the licence and its usage limits in Company Settings › Billing › License Details.
Fix
Reuse existing test cases, or ask your admin to extend the licence. See “Company license limit has been reached”.
“Your license has expired. Please renew to continue using all features.”Licence
Why
The trial or the licence period has ended. You can still open your test cases, but creating new ones fails.
Fix
An admin renews the licence in Company Settings › Billing. See Licence expired.
“Used as a template step in other test cases. Remove those steps first:”Archive
Why
The test case, or a test case inside the folder, is a template that other test cases call. Archiving it would break them.
Check
The list under the message, or the template’s ×n badge.
Fix
In each listed test case, remove the template step (or replace it) and save. Then delete again.
“You can only archive what you created, and this contains items created by other people. An admin can archive it.”Archive
Why
Users may archive only items they created themselves, including everything inside a folder. The archive dialog lists who created what.
Fix
Ask an Admin, or archive only your own test cases one by one. See Archive and restore.
“Only an admin can archive a module (it moves its folders in Analyze and Automate together).”Archive
Why
A top-level folder is a module; archiving it also archives its twin in Analyze and moves its App Map pages to Unassigned.
Fix
Ask an Admin.
“You are not authorized to delete. Only admin or the creator can delete this.”Archive
Fix
Ask the creator or an Admin to delete it.
“Only the person who archived it or an admin can restore it.”Archive
Fix
Ask the person who archived it (shown in the Archive) or an Admin.
A restored test case is missing from its pipelineArchive
Why
Archiving removes a test case from pipelines and test suites, and restoring does not put it back (“removed from n pipeline(s); restoring does not put it back”).
Fix
Add it to its pipelines again in Schedule. Its requirement links were kept.
I can’t move a module into another folder: “Only an admin can move a module folder under another module, because it removes the module.”Folders
Why
Moving a module under another folder removes the module from the App Map and moves its Analyze twin too.
Fix
Ask an Admin, who confirms “Make “name” a sub folder?”. See Move test cases and folders.
“A module with this name already exists” or “Module name is required.”Folders
Why
Top-level folders are modules, and module names must be unique and not empty. The check includes the module’s App Map and Analyze twins.
Fix
Choose another name, or put the folder inside an existing module.
“A module folder is deleted together with its module (App Map or the folder's menu).” or “This module can't be deleted yet…”Folders
Why
A module folder cannot be deleted on its own. Deleting the module from the App Map is refused while its folders in Analyze or Automate still hold items.
Fix
As an Admin, use Delete in the folder’s Options menu in Automate: the module and everything in it go to the Archive. Or move the items out first.
The tree shows “Unnamed test case #1234”Tree
Why
The test case was created with New Test Case but never named and saved.
Fix
Open it, type a name in the header and click Save, or archive it.
New Manuel Case is greyed out with “Coming soon”Tree
Why
Creating new manual test cases is not available yet. Existing manual cases still open. See Manual test cases.
“The folder summary could not be loaded.” or “Condition coverage not measured”Tree
Why
The summary could not be read, or the pages the tests use have no conditions yet because the App Map was never updated.
Fix
Click the folder again. For coverage, run Update map in Discover. See Folder summary.

Saving#

The Save button is grey: “Enter a test case name at the top to enable Save”Save
Why
The name in the editor header is empty (it has a red border).
Fix
Type a name. See Create a test case.
“Please select a Start Page URL.”Save
Why
A web test case needs a start page, for saving and for recording.
Fix
Pick one in the dialog that opens; Save or recording continues. No start page in the list? Add one. See also Quick Start.
“StartPageUrl configuration error. Please re-select the URL.”Save
Why
The start page stored with the test case is not in the form the editor expects.
Fix
Pick the start page again in the dialog that opens, then save.
“Please fill the following fields:” … DescriptionSave
Why
Name, description, and for web start page and environment, are required. The description is written by AI and read only; normally it copies the name until then. A test case created by an import, a script or MCP can have an empty description that you cannot type into.
Fix
Change the name slightly and change it back: the description copies it. Fill any other field the message names. See Save.
“Cannot save yet” — “n issue(s) found”Save
Why
Some steps have missing or invalid values. Up to five are listed as "Step" → field: message, for example “Code body cannot be empty”, “API URL is required”, “Please enter a prompt first.”, “SQL query cannot be empty.”, “Fill in the key and the JSON path, or remove this row”.
Fix
Select each listed step and fix the field in Parameters List. What each step type needs: step reference.
"<step>" → Timeout: 1-300 snSave
Why
A step’s timeout must be between 1 and 300 seconds.
Fix
Enter a value from 1 to 300 seconds.
“Duplicate keys found in step:” or “Names that produce a global parameter must be unique within the test case. Conflicts:”Save
Why
One step uses the same parameter key twice, or two steps store a value under the same variable name.
Fix
Rename one of them, for example orderNo and orderNo2, and update the steps that read it. See Steps that produce a variable.
“This key is already registered as a global parameter, please change the key value.”Save
Why
Another test case already produces a global parameter with this key.
Fix
Use a more specific key (checkoutOrderNo), or read the other test case’s value with {{tc:<caseId>.name}}.
“Parameter key can only contain letters, digits and underscore”Save
Fix
Remove spaces, dashes, dots and braces from the key: order_no, not order-no or {{orderNo}}.
“Test case not be updated”Save
Why
The server refused the save. This one message covers several causes.
Check
  1. A load-test rule: JMeter or k6 steps mixed with browser steps. See Load-test scenarios.
  2. A template step that would create a loop, or that calls a test case of another project.
  3. A mail step whose mail integration is missing or belongs to another project.
  4. Two mobile or desktop Get Text, Get Attribute or Store Variable steps using the same variable key.
  5. You lost access to the project, or your session expired.
Fix
Fix the cause, then save again. If none applies, reload the page (your unsaved changes are lost) and repeat the change in smaller parts to find the step. Still failing: write to info@smartestqa.ai with the test case ID.
“Test Case Parameters Changed” opens on Save, or “Some schedulers could not be updated”Save
Why
You changed parameters of a test case that pipelines use; the dialog syncs them.
Fix
See Test Case Parameters Changed. For pipelines that could not be updated, open them in Schedule and check their parameters.
A rename or locator change showed up in other test casesSteps
Why
A step’s name is its action’s name, and Action Detail edits the shared action. Both change every test case that uses the action. Action Detail edits are saved at once.
Fix
Change it back the same way. Next time check the usage chip and use Save as Copy. See Rename a step.
Replace Action cleared my valuesSteps
Why
Replace Action resets the step’s parameters, because the new action may need different ones.
Fix
If you have not saved, leave without saving to get them back. Otherwise enter them again. See Replace the action.
“Save as copy is not supported for Assertion, JMeter, API, and AI action types.”Steps
Why
These step types (also k6, database and global parameter) have their own action per step already.
Fix
Edit the step’s values in Parameters List; the change affects only this step.
Steps disappeared after a colleague savedSteps
Why
Test cases are not locked. The last Save wins, with the step list of whoever saved last.
Check
History in Test Case Detail shows who saved when.
Fix
Add the missing steps again. Agree who edits which test case. See Two people editing.
“Are you sure to exit without saving?”Save
Fix
Save and Exit keeps your changes; Exit without saving drops them; Cancel stays. See Leaving with unsaved changes.
An environment is greyed out: “No URL defined for this environment”Environment
Fix
Add the address for that environment to the start page in Project Panel, or pick another environment. See Environment and start page.

Recording and adding steps#

“No URL configured for environment: env”Recording
Why
The start page has no address for the test case’s environment.
Fix
Switch the environment chip, or add the address in Project Panel › Start Page URLs.
The Chrome Web Store opens, or “Browser extension still not detected. Please make sure the Smartest extension is installed and enabled, then try again.”Recording
Why
The page did not get an answer from the extension.
Fix
Install or enable the extension, press F5 on the SmartestQA tab and click Record again. See Quick Start.
“No new step found. Please refresh the page.”Recording
Why
After recording, the editor waited 20 seconds and found no new steps on the server. Either nothing was recorded, or saving took longer.
Fix
Press F5 and reopen the test case. If you had unsaved changes, save them first. If the steps are still missing, record again and check that you are logged in to SmartestQA in the same browser.
“An error occurred while loading steps.”Recording
Fix
Reload the page and reopen the test case: the recorded steps are already saved on the server.
Recorded steps went missing after I switched test caseRecording
Why
The extension saves into the test case you started from. The editor shows them only when that test case is open.
Fix
Open the test case you recorded into.
The Record button is missingRecording
Why
The test case is a load-test scenario (JMeter or k6), where recording is hidden.
Fix
Record in a separate test case. See Load-test scenarios.
Mobile or desktop steps I recorded are goneRecording
Why
Inspector steps are added to the editor only (“n step(s) added to the test case. Don't forget to save.”) and were not saved.
Fix
Record again and click Save. The actions themselves were created and are in Discover’s Action List.
“No SmartTest-Desktop agent found. Install and start the desktop app on the computer where the application runs.”Recording
Fix
Install and start the agent on that computer (Windows or macOS) and wait until it is online. See Install an agent.
“Cannot Add Recursive Test Case”Steps
Why
Adding this template would make a test case call itself, directly or through another template.
Fix
Use a different template, or split the shared steps into a new template both can use. See Add a template.
JMeter or k6 is disabled, or “This scenario is a load test: only JMeter, k6 and API steps can be added.”Steps
Why
A load test lives in a test case of its own, with only JMeter, k6 and API steps.
Fix
Create a separate test case for the load test. See Load-test scenarios.
The same step was added twiceSteps
Why
The Action List and Template dialogs stay open and each click adds a step, so a double click adds two.
Fix
Remove the extra step.
“Failed to add step.”, “Failed to remove step.” or “No step was selected. Please select a step and try again.”Steps
Fix
Click the step first, then try again. If it repeats, save, reload the page and retry.
“Failed to update action”Steps
Why
The action could not be saved from Action Detail.
Fix
Check the action’s fields and try again, or edit it in Discover.

Variables and data#

The test typed {{name}} literallyVariables
Why
No step produced the variable before this step, and no stored value or data source key of that name exists. Unresolved variables are used as written.
Check
The spelling (names are case-sensitive), that the producing step comes earlier and is active, and that it passed in the log.
Fix
Fix the name or the step order. See Where a variable gets its value.
A step used a value from an earlier runVariables
Why
Values that steps produce are also stored as global parameters. When the producing step did not run this time, the stored value from an earlier run was used.
Fix
Make sure the producing step runs before it, and check its log line. See Where a variable gets its value.
The dry run used only one data rowVariables
Why
Dry runs use the first row of a dataset. Iterating over rows is a pipeline feature.
Fix
Set the execute count in the pipeline. See Run a test case once per data row.
“This parameter belongs to a later step and may not have a value yet”Variables
Fix
Move the producing step above this one, or pick another value. The same applies to “This parameter is produced by this step, so it has no value inside the step itself”.

Dry runs and the live log#

I can’t start a dry run: “Please save the test case before running”Dry run
Why
The test case has unsaved changes. A dry run runs the saved version.
Fix
Click Save, then Dry Run.
“… code is empty for Python step(s): …. Dry Run was not started.”Dry run
Why
A Python step has no code for the engine the project uses (Selenium or Playwright).
Fix
Write the code in that engine’s tab of the step, or deactivate the step.
“To perform a Dry Run, you need to have the Agent installed locally and the WebSocket connection established…”Dry run
Why
You chose to run on your agent, but there is no agent registered for you.
Fix
Install the agent, or choose Run on Server for a web test.
“Your agent appears to be offline…” or “Your agent is offline. Start the agent app to use this option.”Dry run
Why
Your agent is installed but not connected: closed, asleep or without network.
Fix
Start the agent and wait for the green dot in Device Farm › Test Agents. See Agent offline.
“The test server queue is full. Please try again in a moment.” or “The test server is not reachable right now. Please try again later.”Dry run
Why
Run on Server has 20 runs waiting already, or the server runner is not reachable.
Fix
Try again in a minute, or run on your agent. See Quick Start.
“Failed to start test execution!”Dry run
Fix
Save, reload the page and try again. If it repeats, try the other run target and write to info@smartestqa.ai with the test case ID.
“Appium server not found — Appium server must be running to execute mobile tests…”Dry run
“The test could not start — a step's code or IF/ELSE/END structure is invalid…”Dry run
Why
A code step does not compile, or an IF has no END. No step was run.
Fix
Check the code steps and their IF / ELIF / ELSE / END order. See the step reference.
The live log says DisconnectedDry run
Why
The panel lost its connection to the live log; the run itself may still be going.
Fix
Check your network and wait. The live log of a run is kept for 30 minutes. Run again if the panel does not recover.
“Screenshot expired (kept for 1 hour).” or “Screenshot could not be loaded.”Dry run
Why
Dry-run screenshots are deleted after one hour.
Fix
Dry run again, or run the test case in a pipeline to keep screenshots with the report.
The tree icon did not change after a dry runDry run
Why
The icon shows the last pipeline run. Dry runs never change it, and they do not appear in Recent runs or Report.
Fix
Run the test case in a pipeline. See Status icon and Run modes.
I can’t stop a dry runDry run
Why
Dry runs have no stop. Closing the panel only closes the panel.
Fix
Wait for it to end; the live log shows when it is Completed. For long tests, deactivate the steps you do not need while you work on one part.
An action’s locators changed after a dry runDry run
Why
Auto-heal found the control with a new locator and wrote it to the action. A review was created.
Fix
Approve or reject the review in Maintenance; rejecting restores the old locators. See Diagnose a failed step.
The test passes in a dry run but fails in a pipelineDry run
Check
  1. The pipeline’s own parameter values for this test case.
  2. Which dataset row the failing iteration used.
  3. Where the pipeline runs: the server cannot reach systems inside your network.
  4. The environment, and earlier test cases in the pipeline that change the data.
Fix
See Run modes.

AI#

AI buttons are disabled, or “AI is not enabled or API key is missing for this project.”AI
Why
Enable AI or AI Document Analysis is off, or the project has no working AI connection. The tooltip names the missing option. A message starting “AI API key is not configured” means the key is missing.
Fix
An Admin turns them on and sets up the AI connection. See Turn on AI.
“No actions found for this project. Please record actions first.” or “No pages found — record actions first.”AI
Why
The AI builds steps from recorded actions, and there are none (or none on the selected pages: “No actions found for the selected pages.”).
Fix
Record the flow once, or scan pages with Discover Your App.
“AI could only partially cover your request — missing actions for …”AI
Why
Some parts of your prompt have no recorded action; they became AI steps.
Fix
Record the missing controls and generate again, or replace the AI steps with actions. See Generate from Prompt.
“AI could not produce any usable steps for this prompt…”, “AI returned no usable test cases…”, “AI produced no usable scenarios…” or “The AI could not produce a valid result…”AI
Fix
Try again. Rephrase with more detail: the page, the data and the expected result. Make sure the pages involved have a purpose in the App Map.
“Please enter a prompt first.”AI
Fix
Type what the test should do before you click Generate Steps, or fill the prompt of the AI step.
Generate Steps replaced my stepsAI
Why
Generate Steps replaces the whole step list, name and description in the editor.
Fix
If you have not saved: leave the test case with Exit without saving. If you saved, the old steps are gone; rebuild them or record again.
Expand Scenarios moved my test case, or created a new moduleAI
Why
Create puts the variations and the original in a new folder under the original’s folder. At the root that folder is a module.
Fix
Drag the test cases where you want them and archive the empty folder. See Expand Scenarios.
“AI request limit reached. Please wait a moment and try again.”, “Could not reach the AI service…” or “The configured AI model is no longer available…”AI
Why
The AI provider is rate-limiting requests, cannot be reached, or the configured model was withdrawn. “Expansion failed” and “Error saving expansion” are the general errors of Expand Scenarios.
Fix
Wait a minute and try again. For a withdrawn model, an Admin picks another model in the AI settings.
The description only repeats the nameAI
Why
The AI has not written it yet: descriptions are written when someone runs Update map in the App Map.
Fix
Run Update map and apply. See AI descriptions.

Didn’t find your answer?Write to support with your project name and what you see on screen.

info@smartestqa.ai
© 2026 Selenity Technology · SmartestQAsmartestqa.aiThis page as MarkdownChecked against smarttest-fe predev 2a8c79cc, smarttest-be 633c2a03, smarttest-testsuite runner, research/automate-editor.md, research/automate-steps-and-runs.md