This chapter provides subscriber-facing troubleshooting procedures and frequently asked questions for AIUNIFY PROS SUITE.
It covers problems involving:
When a Tool does not work as expected, troubleshoot in this order:
This order helps distinguish an account problem from a Tool-specific problem.
The Sign-In form requires:
The current form validates both before submitting.
If no Email is entered, the form can display:
Email is required
If the value is not formatted as an Email address:
Invalid email address
If Password is empty:
Password is required
The Login form contains these validation rules before authentication is attempted.
If the login request fails, the page can display:
Login failed
with the error returned by the authentication service or the fallback:
Invalid credentials.
If an unexpected error occurs during the login process:
Error
Something went wrong. Please try again.
Verify:
Then retry Sign In.
The account may require Email Verification.
When this applies, the Sign-In interface switches to:
Verify Your Email
Current message:
You must verify your email address to log in. Please check your inbox for a verification link.
Use:
Resend Verification Email
After a successful resend:
Verification Sent
A new verification link has been sent to your email.
The Email Verification state can also display:
A new verification link has been sent!
The verification page provides a resend workflow when a link is no longer usable.
Current guidance:
If your link has expired, enter your email below to request a new verification link.
A failed link can display:
Verification Failed
The subscriber can request another verification Email.
Current fallback:
Connection error. Please try again.
Select:
Forgot password?
from Sign In.
The current Password-Recovery process follows:
→ OTP
→ New Password
→ Success
The Forgot Password page states:
Enter your email address and we'll send you an OTP to reset your password.
Enter the account Email and select:
Send Reset OTP
The current recovery workflow sends a:
6-digit verification code.
Enter all six digits and select:
Verify Code
If verification fails:
Invalid OTP
with:
The OTP provided is incorrect or expired.
Use:
Didn't receive the code? Resend
Successful resend:
OTP Resent
A new OTP has been sent to your email.
Successful verification:
OTP Verified
You can now reset your password.
After successful OTP verification, the workflow displays:
Set new password
Successful completion displays:
Password Reset!
Your password has been changed successfully. You can now use your new password to log in.
Select:
Return to Login
and sign in using the new Password.
After correct Email and Password, an account with 2FA enabled can display:
Two-Factor Required
Please enter your authentication code.
The current 2FA page asks for:
a 6-digit security code
from the subscriber's authenticator app.
Select:
Use a recovery code instead
The current interface describes Recovery Codes as:
XXXX-XXXX
eight-character codes.
The 2FA Settings page states that each Recovery Code can only be used once.
AIUNIFY recommends storing them somewhere secure, such as a Password Manager.
The 2FA interface can display:
Please enter your verification code
Recovery mode can display:
Please enter a recovery code.
The shared authentication layer has the fallback:
2FA verification failed.
If Recovery Codes were regenerated, the previous codes are invalidated.
The Settings interface explicitly warns:
Regenerating recovery codes will invalidate your current ones.
The current Sidebar filters Tool visibility according to:
A Tool can be absent because it is not currently assigned to the subscriber's account or plan.
The current access logic treats Enterprise accounts differently from standard assigned-Tool restrictions.
A system-configured Free Tool can bypass the normal plan-assignment check.
A Tool can also be explicitly disabled for the user.
Protected Tools can display:
Access Restricted.
Current message:
Your organization has disabled access to this feature.
Current message:
This AI tool is not included in your current plan. Please upgrade to access it.
When the restriction is plan-related rather than organizationally disabled, the screen can provide:
Upgrade Plan
The same Access Restricted screen provides:
Return to Dashboard.
Some disabled protected routes can redirect directly back to:
/dashboard
instead of showing the Tool.
Tokens are the AIUNIFY PROS SUITE account's usage units for supported AI operations.
They should not be confused with raw LLM input/output token counts.
Some specialized generation services return:
Insufficient tokens. Please top up your balance.
Audio Isolation can return:
Insufficient tokens for audio processing.
Specialized Website and Presentation workflows can receive:
Insufficient tokens
or a service-provided Token error.
Review:
Standard ToolPage can calculate Cost from:
and can alter the displayed Cost for certain Models.
Standard ToolPage currently applies a 1.5× display multiplier when the selected Model provider is OpenAI.
For example, Website Builder and PPT Playground explicitly charge separate Initial Generation and Follow-Up Edit amounts rather than using Standard ToolPage.
AIUNIFY refreshes the subscriber account after many successful AI/API operations.
The current shared authentication layer also schedules a silent account refresh after successful non-authentication API POST requests.
That automatic refresh is currently scheduled after approximately:
800 milliseconds
for qualifying successful POST operations.
Reloading the application causes the authentication context to fetch the current account state again through /api/auth/me.
Some Models require an API Key.
If the required Key is unavailable:
Model Not Available
with:
This model requires an API key. Please configure it in Settings.
Open:
Settings → API Keys
Current Settings provides fields for:
After Save:
API Keys configuration saved
Your API keys have been updated successfully.
The current authentication context calls refreshModels() when API Keys change.
If the currently stored Model is no longer in the available Model list and the selection is not a manually preserved selection, AIUNIFY returns to the configured default Model.
Settings describes the Gemini Key as required for Gemini 2.5 Flash and other Google Models where a subscriber Key is required.
Settings identifies the OpenAI Key as required for GPT-4o and other OpenAI Models.
Settings identifies the Anthropic Key for Claude Models.
Settings contains a Mistral API-Key field.
Settings contains a Groq API-Key field.
Settings contains an NVIDIA API-Key field.
Certain specialized services can return an error such as:
Gemini API Key not configured
when the required provider-side configuration is unavailable.
Not every specialized Tool uses the same BYOK pathway as Standard ToolPage.
A subscriber's saved Key therefore does not prove that every independent media/provider service has valid server-side configuration.
First confirm that Input contains non-whitespace text.
Standard ToolPage's exact validation is:
Input required
Please enter a prompt to continue.
Standard ToolPage can display:
Error
with either a service-specific error or:
Failed to generate content. Please try again.
For a Standard Tool failure, verify:
If no usable streamed response is returned, a Standard Tool has no separate recovery history.
Retry after checking Model/Token configuration.
Most Standard ToolPage workspaces keep only the current local Output.
Generating again replaces the active result.
Standard ToolPage does not provide universal persistent Save.
This is expected behavior rather than a broken Save button.
Standard ToolPage does not provide a universal Output editor.
Use Copy or Download, or modify the Prompt and generate another result.
After a successful Standard Tool copy:
Copied!
Content copied to clipboard.
Standard ToolPage downloads generated content as:
[tool-id]-output.txt
TXT is the current built-in Standard ToolPage format.
Only specialized Tools with their own export functions provide other formats.
Source-confirmed examples include:
where documented in their chapters.
Do not assume every AIUNIFY Tool accepts the same:
Several image-capable subscriber workflows explicitly reject Images larger than:
5 MB
with:
Please upload an image smaller than 5MB.
Some interfaces display:
File too large
or:
File Too Large
with:
Please upload an image smaller than 5MB.
Where a Tool is configured for Images only, it can reject other formats with:
Please upload an image file.
A separate source-confirmed document submission workflow accepts:
and limits files to:
10 MB.
That workflow can display:
Unsupported File Format
Only PDF, DOC, and DOCX formats are supported.
Always follow the specific upload controls and limits shown by the Tool being used.
Several AI Marketing generators use asynchronous generation.
A request starts a task and the page polls for completion.
Common notifications include:
Generation started. Please wait...
Current Video Generator can state:
Video generation started. This may take 2-3 minutes.
This is interface copy, not a guaranteed completion time.
Reels can display:
AI Reel generation started. This involves script, voiceover, and scene creation.
An asynchronous provider task can ultimately return:
Generation failed
or the provider's supplied error.
The client logs polling errors when it cannot retrieve task status.
A polling error is different from a confirmed provider-side generation failure.
If the page shows an active Generation state, allow that task's status workflow to resolve before assuming the request never started.
Some AI Marketing generators explicitly detect provider safety failures.
One source-confirmed message is:
Prompt flagged by safety filter. Try rephrasing.
Another current interface can state:
Your prompt was flagged by the AI safety filter. Try rephrasing — avoid mentioning real people, brands, or sensitive topics.
Rewrite the Prompt so that it clearly describes the intended permissible creative result without unnecessary sensitive or ambiguous material.
Do not assume all generators apply the same provider or identical filtering behavior.
The current Image Generator can display a service error or:
Failed to start generation ([status]).
Another source-confirmed fallback is:
Failed to generate image.
Confirm:
Then retry if appropriate.
Music Generator can return:
Failed to generate music.
The Music status action can return:
Unauthorized
if the session is unavailable.
Confirm that the subscriber remains signed in before starting or checking the task.
The specialized Website workflow attempts to deduct:
500 Tokens
before generating a new Website.
If that operation fails, the generation is not started.
Follow-up AI Website modifications require:
10 Tokens
per source-defined edit request.
Current message:
AI did not generate any code.
Current message:
AI did not generate any valid code for this modification.
Retry with a more explicit Website-design or modification instruction.
For targeted Visual Edits, clearly describe what should change on the selected element.
The current Website workspace can display:
Failed to load project.
Check that:
The source does not establish additional user-facing recovery diagnostics beyond the returned server error.
The Website Projects page can display:
No projects yet. Create your first AI website!
That is an empty-state message, not necessarily an error.
The current save workflow can display:
Failed to save website ([status])
or a server-provided error.
Another current error format is:
Error saving website: [error].
Successful Save displays:
Website Saved Successfully!
Website Builder supports persistent Projects, but unsaved changes should not be assumed to exist in the Saved Project record.
The current Publish dialog requires a Subdomain before proceeding.
Current message:
Please enter a subdomain
The current Subdomain Input is automatically converted to lowercase and strips characters other than:
a-z0-9-The workflow can display:
Failed to publish website ([status])
or the server-returned error.
It can also display:
Error publishing website: [error]
Current success:
Website Published to Subdomain!
Current errors include:
Failed to unpublish website
or:
Error unpublishing website
Current success:
Website unpublished!
PPT Playground requires:
ppt-playground Feature accessThe route uses FeatureGuard.
The specialized Presentation Studio deducts:
500 Tokens
before generating a new Deck.
Follow-up Design Chat requests deduct:
10 Tokens.
If no Slide content can be parsed from the AI response, the current Presentation workspace can fall back to pre-built template Slides.
Using pre-built template slides for this topic.
A completely failed AI Slide response can activate the fallback deck generator.
Review the Deck rather than assuming all displayed claims were generated specifically from verified research.
Current message:
Failed to save presentation.
Current fallback:
Error saving presentation.
Current message:
Presentation saved successfully!
Save is disabled if:
Export requires a non-empty Presentation Deck.
While the browser export runs:
Exporting...
is displayed.
Save Presentation stores the AIUNIFY Project.
Export PPTX creates a PowerPoint file.
These are separate actions.
Game Developer can display:
Generation Failed
with the returned error.
Game Developer's optional Image Upload currently enforces:
5 MB
and can display:
Please upload an image smaller than 5MB.
Unlike Standard ToolPage, Game Developer has a source-confirmed cancel mechanism that aborts the active request and returns the generator to idle state.
Stop/Cancel availability is Tool-specific.
First identify which type of workspace you used.
Not every Tool saves results.
Most Standard ToolPage generators do not create persistent Projects.
Website Builder does create persistent Saved Projects when Save is used.
Presentation Studio supports Saved Presentations.
Game Developer supports Saved Games.
Certain Media Tools provide their own:
where documented in their respective chapters.
Downloading a file to your device is separate from creating a persistent AIUNIFY Project.
Copying Output to the clipboard does not create an AIUNIFY Project.
Some persistent-content workflows can return:
Delete failed.
A source-confirmed delete workflow can display:
Network error.
For example, the Website Project confirmation explicitly states:
Are you sure you want to delete this project? This action cannot be undone.
Deleting a Project should not be used merely to solve a loading problem unless losing that Project is acceptable.
A current subscriber workflow uses:
Network Error
Failed to connect to the server.
At the user-interface level, this means the request could not successfully communicate with the expected application service.
Retry the operation after confirming that the browser remains connected and the AIUNIFY application is responsive.
These are different failure classes:
| SymptomFirst Check | |
| Cannot Sign In | Email/Password/Verification/2FA |
| Tool Missing | Assigned Tool/disabled feature |
| Access Restricted | Plan or organization restriction |
| Generate Disabled | Required Input |
| Model Not Available | Provider API Key |
| Insufficient Tokens | Token balance/Tool Cost |
| File Too Large | Tool-specific upload limit |
| Generation Failed | Prompt/provider/Token/service |
| Project Missing | Whether Tool actually saves |
| Save Failed | Persistent Project operation |
| Export Disabled | Required content exists |
| Publish Disabled | Required Subdomain/Input |
| No History | Verify whether Tool supports persistence |
Yes.
Subscriber Tools are designed around an authenticated account.
The application no longer considers the current browser session authenticated, or the protected Route requires authentication.
Some disabled-feature routes are explicitly redirected to /dashboard.
Global Search applies Tool-availability filtering similar to the Sidebar and omits unavailable Tools.
No.
Availability can depend on the subscriber's:
No.
Yes.
Standard ToolPage can use current Tool-specific AI limits and Model-related display logic.
Some source files contain:
The live Tool display should take priority for actual operation.
For Standard ToolPage, yes.
The current source applies a 1.5× display multiplier for OpenAI-provider Models.
Website Builder is a specialized workflow with an explicit 500-Token Initial Generation deduction.
The current specialized Website edit workflow explicitly deducts 10 Tokens for follow-up AI modifications.
No.
The current Settings/model architecture contains multiple providers.
The Model may require a provider API Key.
Current interface meaning:
This model requires an API key. Please configure it in Settings.
The current AuthContext refreshes Models after API-Key state changes.
The current client saves:
selectedModelisManualModelto browser local storage.
Settings → API Keys
No.
The Model selector checks whether a required Key is already available through system configuration or the subscriber's provider-key configuration.
Not every specialized Tool uses the same AI provider pathway.
No.
The Key can still be invalid, expired, restricted, or unsuitable for the requested provider/service.
The subscriber source does not contain a universal successful-Key guarantee.
Most Standard ToolPage workspaces do not include persistent Save.
Persistent History is not a universal Standard ToolPage capability.
Yes.
Yes, as TXT for Standard ToolPage.
Not in the shared Standard ToolPage.
The shared Standard ToolPage supports:
Ctrl + Enter
rather than plain Enter for its keyboard generation shortcut.
Reset intentionally clears the current Input and local Output.
Preserve useful content before Reset.
No.
The current Research Agent Standard ToolPage does not source-verify a live Web-browsing Tool connection.
Not through its current Standard ToolPage implementation.
Yes.
Especially verify:
where consequential.
No.
No.
No.
It generates copy.
No.
It generates variants.
The current Marketing Agent client does not source-verify external advertising execution.
No.
It generates Email content.
No.
No.
It creates preparation text.
No.
It generates Script versions intended for approximate 30-, 60-, and 90-second uses.
Its current implementation is a Standard text-generation workspace.
No.
The current client does not source-verify a deterministic contrast calculator.
No.
The current Tool audits supplied text through AI.
No.
The current Tool generates a textual specification rather than a visual wireframe canvas.
No.
The current client generates the visible Confidence as a random number from 70–99%.
The current visible Bars use fixed templates based on the simplified Overall classification.
No.
It generates Goal-planning text.
No.
Not through its current ToolPage.
The current four AI Agent pages do not source-verify persistent background Jobs.
No.
Their current implementation uses Standard ToolPage.
No.
No.
Not through its current Standard Agent workspace.
No.
It is the Website Builder's specialized editing workspace.
Yes.
Yes, saved Website Projects can be loaded by Project ID.
Yes, through the source-verified Subdomain Publish workflow.
Yes.
Yes.
Use Visual Edits and select the relevant element.
If no individual element is selected, the AI receives the broader page-modification workflow.
For surgical changes, use Visual Edits.
Exactly:
8 Slides
for the current core AI generation architecture.
Yes.
Yes.
Use:
Export PPTX
Yes.
Yes.
The Presentation workspace supports Play mode.
The current eight-Slide architecture does not source-verify general add/delete/reorder controls.
Several Media Tools start asynchronous provider tasks and poll for completion instead of returning the final asset in one immediate response.
The request has been accepted for processing.
It does not mean the final Media asset is already complete.
The asynchronous task reported a failure or the client received an error state.
No.
Possible user-visible failure categories include:
No.
Not automatically.
No.
Asset-management actions differ by specialized Tool.
The Dashboard only retrieves specific source-supported recent asset classes; it is not a universal archive of every PROS SUITE generation.
Some current Galleries do not implement:
This is a current feature limitation rather than a troubleshooting error.
No.
It is a specialized AIUNIFY workspace with the source-verified behaviors documented in Chapter 18.
No.
Browser capability should not be inferred simply because another Tool contains words such as:
Typical signs:
Typical signs:
Typical signs:
Typical signs:
Typical signs:
Typical signs:
Typical signs:
Typical signs:
Typical signs:
Verify that the Dashboard can be opened normally.
If the Tool is missing or restricted, resolve access before troubleshooting generation.
Confirm the balance exceeds the Tool's live displayed requirement.
Make sure the selected Model remains available.
If the Model requires BYOK, verify its provider Key has been configured.
Provide every required field.
Confirm:
Retry after correcting the identified issue.
Do not replace the displayed service error with an assumption.
Before Reset, regeneration, deletion, or replacement, preserve anything important through the Tool's available Save, Copy, or Download function.
Check account availability first.
Read the displayed error.
Persistence is feature-specific.
Deletion may be permanent.
A new generation can replace the active Output.
Verify it.
Generation is not the same as:
unless the Tool specifically implements that action.
Record the exact Tool being used.
Record the page Route where practical.
Example:
/website
or:
/ai-marketing/image-generator
Preserve the wording displayed by AIUNIFY.
Identify what you were doing:
Record the Model used.
Record the visible balance if the issue could involve Tokens.
For upload issues, record:
Do not include confidential file contents unless necessary.
For a Saved Project issue, preserve the Project name and relevant identifier shown in the workflow.
Troubleshooting does not require sharing:
Meaning: No usable Prompt was supplied.
Action: Enter Prompt text.
Meaning: Selected Model requires unavailable API-Key configuration.
Action: Configure the appropriate provider Key in Settings.
Meaning: Tool is disabled or not included in current access.
Action: Review account/plan Tool access.
Meaning: Account balance is below the required operation Cost.
Action: Review balance and live Tool Cost.
Meaning: Uploaded file exceeds that Tool's current limit.
Action: Use a smaller supported file.
Meaning: File type is not accepted by that specific workflow.
Action: Convert/use one of the supported formats.
Meaning: AI/provider generation did not complete successfully.
Action: Review the associated error, Prompt, Model, Tokens, and provider state.
Meaning: The application request could not communicate successfully with its expected server endpoint.
Action: Retry after confirming the application/browser connection.
Meaning: Saved Website Project could not be retrieved.
Action: Confirm session and Project selection; preserve any server-returned error.
Meaning: Website persistence request failed.
Action: Do not assume changes have been saved. Retry Save after reviewing the error.
Meaning: Presentation persistence request failed.
Action: Keep the current Deck open and retry rather than assuming the Project was stored.
AIUNIFY PROS SUITE subscriber troubleshooting can usually be organized into:
The current source confirms user-facing workflows for:
Tool access is controlled by account and system configuration.
The current FeatureGuard can distinguish:
Your organization has disabled access to this feature.
from:
This AI tool is not included in your current plan. Please upgrade to access it.
A missing or restricted Tool should therefore not automatically be treated as an application failure.
Token shortages are a separate failure class.
Specialized workflows can explicitly return:
Insufficient tokens. Please top up your balance.
Standard Tool costs can depend on live configuration, catalogue settings, and selected Model, while specialized workflows can use their own operation-specific Token deductions.
The current Model selector prevents selection of certain Models without required configuration and displays:
Model Not Available
This model requires an API key. Please configure it in Settings.
Subscriber Settings currently contain provider-Key fields for Gemini, OpenAI, Anthropic, Mistral, Groq, and NVIDIA.
The most common Standard Tool validation is:
Input required
Please enter a prompt to continue.
A generation failure can display:
Failed to generate content. Please try again.
Standard ToolPage generally provides Copy, TXT Download, and Reset—but not universal persistent Save or History.
Upload requirements are Tool-specific.
Several Image-enabled workflows enforce a 5 MB Image limit, while other specialized document workflows can use different file types and limits.
Subscribers should therefore follow the actual Tool's accepted formats rather than applying one Suite-wide upload rule.
Image, Video, Avatar, Logo, Manga, Thumbnail, Reels, Music, and other specialized media workflows can use asynchronous provider tasks.
A status such as:
Generation started. Please wait...
means the request entered processing.
It is not the same as a completed result or a failure.
Some specialized AI Marketing workflows can explicitly return:
Prompt flagged by safety filter. Try rephrasing.
That should be handled by revising the Prompt rather than repeatedly submitting the identical request.
A key troubleshooting distinction throughout PROS SUITE is:
Not every generated result is a Saved Project.
Standard ToolPage content is generally transient.
Specialized workspaces such as Website Builder and Presentation Studio implement explicit persistent Save functions.
Save stores supported Project data within the source-verified AIUNIFY persistence workflow.
Download creates a file for the subscriber.
Copy transfers text to the clipboard.
These three actions should not be treated as equivalent.
Website Builder and PPT Playground have their own Token, Save, editing, and error workflows.
Website Builder can report:
while Presentation Studio can report:
The central operating rule for Chapter 48 is:
Read the exact message AIUNIFY PROS SUITE displays, identify which layer failed, and correct that layer first. Authentication problems should not be treated as Token problems; Tool-access problems should not be treated as Model problems; Model/API-Key problems should not be treated as generation problems; and the absence of Save, History, publishing, browsing, or external execution should not be treated as an error when that capability is not implemented in the current Tool.