Skip to main content
Back to NLP Topics Standard responses are pre-defined NLP interpreter messages the AI Agent displays based on events, triggers, user inputs, or system conditions. You can customize all defaults and add channel-specific overrides. Go to Automation AI > Conversation Intelligence > Conversation Management > Standard Responses. For each response category, the trigger or user input is listed alongside the default response(s). The NLP interpreter randomly selects from multiple responses at runtime.

Response Categories

CategoryDescription
StatementsResponses triggered by AI Agent or user actions — for example, task cancellation or invalid input.
QueriesQuestions the AI Agent asks for more information — for example, Which field do you want to change?
Errors & WarningsMessages displayed during errors — for example, End date cannot be before the start date.
QuestionsGeneral questions posed to the user that don’t fit other categories.
ChoicesYes/No prompts — for example, Would you like to receive push notifications?
GreetingInitial responses when the app starts or when a user returns after inactivity.

Statements

Responses triggered by actions from the AI Agent or user.
MessageWhen It Appears
When the user confirms to close the conversation in a language and switch to another languageUser ends a conversation in one language and starts the next in a different language.
When only field prompt is present in the OOBInternal field prompt for action and alert tasks. Not recommended for customization by enterprises.
When language preference is setUser switches language using cheat lang or picks from language options.
When task is not accessible to userDeveloper tries to access an app task they don’t have permission for.
User confirms alert setupAn alert is set up successfully (Alert Task only).
User discards/abandons the current activityUser says discard, or the task fails due to max retries, invalid LoV, or cURL error.
User implies to end the conversationUser says bye, good night, etc. Shown when the Follow-up Intents list is presented at dialog end.
Location provided by user is invalidAI Agent cannot validate input for a Location-type entity.
Amending an Alert fieldUser says Yes to amend alert parameters (Alert Task only).
Amending a task which has no fieldsTask has no entities to amend (Alert, Action, and Information Tasks).
Ask at least one optional fieldUser did not provide any optional field value when at least one is required (Action Task only).
Topic retry limit exceededDeveloper mode only. Dialog looped through the same nodes beyond the allowed limit.
Universal fallback responseAI Agent cannot generate a response for any reason, or maximum node execution limit exceeded.
Server returned authorization failureAuthorization failure occurred during script or API execution.
User selects a valid option from a listUser selects an action/alert/info task from the listed options.
User selects an actionUser selects an action task; AI Agent confirms and asks for required parameters.
Alert is identified, ask for alert timeDisplayed when Trigger Interval is set to At a specific time.
Alert is identified, ask for alert time intervalDisplayed when Trigger Interval is set to Every X minutes.
User selects an alertUser selects an alert task; AI Agent confirms and requests required parameters.
Task discardedUser types discard mid-task.
App choosing the default accountShown during alert amend when AI Agent cannot parse the user’s response.
Confirmation Dialog for Task consolidatedAI Agent confirms captured parameters before executing an Action/Info task.
Confirmation Dialog for Alert consolidatedAI Agent confirms all details before setting up an alert (if confirmation is enabled).
Confirmation Dialog for Alert consolidated without fieldsAlert has no fields; AI Agent confirms before execution (if confirmation is enabled).
The alerts will be mutedDeveloper has muted alert notifications.
Confirmation Dialog when alerts are mutedShown when Mute alert notifications is enabled for an RSS alert.
Alert frequency descriptionConfirmation of alert query frequency for web-service alert tasks.
Not understanding the input at confirmationUser’s response during alert frequency confirmation does not match the expected input.
No results found for the search stringTypeahead field returned no results.
Successful task setupAlert task set up successfully.
I am not sure what you meanOne or more low-confidence intents found. Usually shown as FAQ suggestions.
Ambiguity in intent recognitionAI Agent failed to identify user intent; prompts user to rephrase.
User has selected daily alert — ask for frequency or specific timeBoth frequency interval and specific time options are available.
App needs to present filter optionsMultiple filters available during alert setup.
App needs to present filter options with exit optionSame as above, but includes an option to skip.
User finished configuring filtersFilter configuration for an alert task completed.
User asks Who am IUser types Who am I.
User asks for helpUser types Help; AI Agent lists available tasks.
Skipped valuesPart of confirmation message for Action/Info tasks — shows skipped (optional) parameters.
Ask user to change dependent fieldA dropdown depends on another field, but the Platform cannot fetch data from that field.
When there is no dataUser chooses to change account but responds no to the AI Agent’s confirmation.
When date given in invalid formatUser enters an unrecognizable date format. Valid formats: YYYY-MM-DD, DD-MM-YYYY, 1 Jan 1982, Next Friday.
Follow-up sentence post task discardAI Agent confirms task was discarded and offers to restart.
Confirmation Dialog for task with skipped fieldsConfirmation before executing an Action/Info task when some optional fields were not provided.
Listing all tasksDisplayed when user types Help and no task matches.
Listing all alertsDisplayed when user types Help and the app has only alert tasks.
Listing all FAQsDisplayed when user types Help — includes Knowledge Graph collections.
Listing all alerts after tasksShown when app only has alert tasks and intent is not identified.
Listing all alerts after tasks and FAQsShown when all task types are configured and user types Help.
Lists all actions and FAQsNo matching task found; lists actions and FAQs.
Lists all alerts and FAQsNo matching task found; lists alerts and FAQs.
When app has no tasksDeveloper tries to run an app without any tasks.
When a server call failsWebhook or service node call failed; task is discarded.
User prompt when a task is discarded due to server call failureAPI call did not complete successfully.
Specific app is not availableApp is not yet published.
Field title and value in confirmationPart of the confirmation step — shows captured field values for alert, action, or info tasks.
Recursive dialog foundAI Agent recursively called the same dialog and exceeded the limit (100).
Loop limit reached for a nodeSame node invoked repeatedly beyond the allowed number of times.
Message component doesn’t have a messageMessage or Service node returned empty content.
User says thanksUser thanks the AI Agent.
User says OK, fine, yesShown after task completion when the user acknowledges positively.
User says no, nopeSmall talk response to negative acknowledgments.
User says great, awesomeShown after task completion when the user responds positively.
User says that helped, that was usefulShown after positive feedback (great, cool, fair enough, that helped).
User says I am doneShown when user indicates they are finished (done, that’s all, that’s it).
Successfully added filterA filter was added to an alert.
After successfully adding a filter, showing filter options againShown when a filter is applied to an alert (Alert Task only).
User gave invalid input for filterFilter name doesn’t match any configured filter.
Resume paused taskUser chooses to resume on-hold task. Shown only when Resume option is Notify the user with a message that the on-hold task is being resumed.
Notify the user that we have canceled the paused taskUser chose not to resume the on-hold task during confirmation.
Notify the user that we have canceled the current task to switch to anotherHold option is Discard the current task and switch to new.
When an interruption is identified but hold is not allowedNew intent detected during task execution; new intent added to follow-up list (current task continues).
Service call timed outThird-party API call (webhook or service node) timed out.
Dialog has no transitionsDialog task has only an Intent node and no further flow defined.
Manage accountUser says manage accounts, add phone number, delete account, etc.

Queries

Questions the AI Agent asks users for more information.
MessageWhen It Appears
Ask if user would like to switch to a new language (multiple detected)Multiple new languages detected mid-conversation; user asked to select preferred language.
Ask if user would like to switch to a new language (single detected)A different language detected; user offered option to switch or continue.
Follow-up sentence post task execution or discardTask completed or discarded; user offered option to restart.
Confirmation before discardUser types Discard or Discard All during task execution (applies to select clients).
User types Change or similar wordsUser changes their mind during alert filter setup; alert setup is discarded.
User prompted to enter text to searchTypeahead field prompt (Action, Alert, Info tasks).
Ask the user to enter alert frequencyFrequency can be interval or specific time (Alert only).
Ask the user to enter alert intervalFrequency is an interval, e.g., every 4 hours (Alert only).
Ask the user to enter alert timeFrequency is a specific time, e.g., daily at 9:00 am.
Alert identified — ask for frequencyPrompted based on schedule options defined by the alert developer.
Prompting user to enter name for alert instanceUser installs alert in Kore messaging rooms.
User needs to select from a static dropdownFilter field type is static dropdown.
User needs to select from a search results listFilter field type is dynamic dropdown.
Request for email address for webhook instructionsWebhook alert setup — developer email needed for configuration instructions.
Query to get the alert instance descriptionUser prompted to provide a description during alert setup.
Tell me the alphabet or unique word for the actionDisplayed before a list of options when no custom template is used.
I am not sure what you mean after an invalid choiceUser’s follow-up doesn’t match any of the initially identified ambiguous alerts.
Asking user to pick an alertA list of alerts is shown; user prompted to select one.
Asking user to pick an alert after an errorUser selected an invalid option from the alert list.
When user responded with invalid answer to confirmation dialogUser said they want to amend but the AI Agent didn’t understand what — confirmation repeated.
Typeahead field prompt (first field in task)First field in task is a typeahead type.
Ask user what to rename the alertUser chose to change/amend the alert name.
After user amends a parameter — ask if anything else to changeUser amended a parameter; AI Agent asks if further changes are needed.
No results found for search string — ask to try againTypeahead field query returned no results.
Follow-up sentence post discard — Part 2At alert confirmation, user said no twice.
Listing ambiguous tasksUser prompted to pick a task from a list; response led to ambiguity.
Ask if user wants to continue the paused taskShown after the interrupting task completes (Resume option: Get confirmation before resuming).
Ask if user would like to switchHold & Resume is set to Allow the end user to select the behavior; AI Agent confirms before switching.

Errors & Warnings

Messages displayed when errors or validation issues occur.
MessageWhen It Appears
Agent transfer fallback responseAgent unavailable, wait time exceeded, or agent transfer failed.
User exceeds maximum retries for an entity (task on hold)User exceeded max retries while another task is on hold.
Dialog task associated to FAQ is not availableThe dialog linked to a FAQ was suspended, deleted, or not assigned to the user.
User needs to authorize or re-authorizeEnd-user authorization or re-authorization is required. The authorization URL is appended dynamically.
User authorization has failed or is yet to authorizeTask requires authorization; AI Agent provides an authorization link.
User’s intent recognized, but authorization failsService call needs authorization and no account has been authorized yet.
User authorization failed more than 3 timesUser entered wrong credentials three times in a row.
User provided value is not uniqueUser’s input causes ambiguity when selecting from a partially matched intent list.
App did not find any options based on user’s choiceOption set is empty, or no account credentials are available.
Discarding task — no options found for list or lookupUser input not found in service call response (dynamic dropdown or typeahead field).
Task identified, but server call failedKore account for the third-party API expired or failed.
User enters invalid emailInvalid email address entered for an alert, action, or info task.
User enters invalid countryInvalid country name entered.
User enters invalid URLInvalid URL entered.
User enters invalid phone numberInvalid phone number entered.
User enters invalid numberNon-numeric input entered for a number field.
Invalid date formatInvalid date format entered.
Missing dateNo date provided for a date field.
Ask user to enter time for a fieldTime missing for a DateTime entity in a dialog task.
Ask user to enter timeUser provided only a date for a DateTime entity (e.g., scheduled appointment).
Invalid time — ask to try again for a fieldUser’s time input for a DateTime entity is missing or invalid.
Invalid time — ask to try againUser entered an invalid time value (e.g., Aug 29 at 25).
Missing date/timeError prompt for DateTime entities in Actions and Alerts.
You have chosen a wrong choiceSub-response used when a user selects an invalid option from a list.
Cannot understand your requestAI Agent fails to identify intent; lists available tasks and FAQs.
App is selected, user asked for reauthorizationAPI call failed due to expired credentials.
Listing all tasks after wrong choiceUser selected an invalid option; app has only Actions/Dialogs/Info tasks.
Listing all alerts after wrong choiceUser selected an invalid option; app has only Alert tasks.
Listing all FAQs after wrong choiceUser selected an invalid option; app has only FAQs.
Listing all FAQs — not understanding requestApp has only a Knowledge Graph and fails to understand intent.
Listing all actions — not understanding requestApp has only Action tasks and fails to understand intent.
Listing all actions and alerts — not understanding requestApp has only Action and Alert tasks and fails to understand intent.
Listing all actions and FAQs — not understanding requestApp has only Actions and a Knowledge Graph and fails to understand intent.
Listing all alerts — not understanding requestApp has only Alert tasks and fails to understand intent.
Listing all alerts and FAQs — not understanding requestApp has only Alert tasks and a Knowledge Graph and fails to understand intent.
Listing all tasks — not understanding requestApp fails to understand intent; shows all available tasks.
Listing all tasks and alerts after wrong choiceUser selected an invalid option; app has tasks and alerts.
Listing all alerts, tasks, and FAQs after wrong choiceUser selected an invalid option; app has all task types.
Lists all actions and FAQs after wrong choiceApp has only Actions and a Knowledge Graph; user selected an invalid option.
Lists all alerts and FAQs after wrong choiceApp has only Alerts and a Knowledge Graph; user selected an invalid option.
Task discarded — user exceeded maximum allowed attemptsUser entered an invalid entity format five times.
Entity component in dialog has no error promptEntity node has no error prompt defined.
Entity component in dialog has no message attributeEntity node has no message prompt defined.
User input exceeds maximum sentence or word limitInput exceeds the allowed character or word count.

Questions

General questions the AI Agent poses to users that don’t fit other categories.
MessageWhen It Appears
When user input matches and frequency is askedUser input led to ambiguity while setting an alert frequency.
Ask user to input alert frequencyPrompt for frequency at which the alert task will run.

Choices

Yes/No prompts used during alert and task setup.
MessageWhen It Appears
Search results displayed (type-ahead search)User input triggered a typeahead search in an alert filter (Alert Task only).
Confirming alert frequencyApp confirms current alert trigger frequency with the user.
Repeat confirmation of default alert frequencyUser input doesn’t match the expected frequency response.
Confirmation to set up alert with frequencyConfirm the trigger frequency for an alert.
Prompting user to set up filtersApp asks whether the user wants to configure alert filters.
Prompting user to set up filters as soon as alert is identifiedSame as above — triggered immediately on alert identification.
Asking user to set up push notificationsUser wants to amend the alert notification setting.
Repeat — asking user to set up push notificationsUser’s response to notification setting doesn’t match yes/no.
Amending the descriptionUser wants to change the alert description.
Amending the frequencyUser wants to change the alert trigger frequency.
Amending notification settingsUser wants to change whether notifications are enabled.
Prompting user to choose days of the weekUser selecting Weekday/Weekend for trigger frequency.
Query to choose interval for frequencyUser prompted to select a frequency interval.
Query to send webhook setup instructions via emailApp checks if developer wants webhook configuration instructions emailed.
Do you want to change the alert name?User wants to amend the alert name.
Query to choose operator while setting up a filterUser prompted to select an operator for an alert filter.
Options to pick frequency for alert notification daysAll interval options available (Daily, Weekday, Weekend).
When the user needs to choose from provided choicesUser prompted to pick from a dropdown in alerts or actions.
Confirm if user wishes to reuse a pre-saved contextPrompted when a pre-saved context is available for reuse.
Show possible intervals for alert scheduleUser entered a specific time, but only intervals are valid.
Show closest frequency for alert scheduleUser’s input doesn’t match available frequencies; nearest option shown.
Show closest time for alert scheduleUser’s input doesn’t match available times; nearest option shown.
When changing a fieldUser selected No during alert parameter confirmation.
Prompting to set up filters after alert identifiedUser provides input to set up an alert that has filters configured.
Ask if setup instructions should be sent via email (first prompt)First prompt in alert setup is to ask about emailing setup instructions.
While choosing optionsShows options from the Platform for a typeahead entity in an action or alert.
After selecting task — ask about scheduleAlert selected from list; AI Agent asks how often the user wants to be notified.
Wrong choice for task scheduleUser entered invalid frequency for an alert; AI Agent asks again.
After selecting task — ask for field valueFirst field in an action or alert is a dropdown.
User says no for alert setup confirmationUser chose not to proceed with the alert configuration; offered to amend.
After selecting task — user selects wrong choice in dialogUser entered an invalid choice from a non-LoV list.
User enters wrong choice for alert frequencyAlert frequency value doesn’t match options defined by the developer.
User enters wrong value for a dropdownInput doesn’t match any value in the List of Values.
List ambiguous LoV valuesMultiple partial matches for a List of Values entity; user asked to select.
User enters invalid choice for an actionUser’s response doesn’t match the presented task list.
Ambiguous dateUser entered an ambiguous date (e.g., 6/9); both interpretations shown.
Ambiguous airportUser entered an airport name that matches multiple airports (e.g., London).
Ambiguous locationGeocode service returned multiple location possibilities.
More than one account to pick fromUser has multiple third-party accounts; asked to select one.
Intent recognized, but more than one account to pick fromSame as above, when intent is identified.
Custom lookup returned more than one valueMultiple results returned from a custom list query.
Chosen list is empty after selecting a valid choiceDependent dropdown has no options because the parent field value returned empty results.
User selects an invalid choice (dependent dropdown)Same as above; user’s value for the dependent field is invalid.
Present follow-up tasks at end of dialogShown at end of dialog when Follow-up Task setting is Yes, ask the user to select a follow-up task.
Ask user to select entity to amend (ambiguous entity value)Multiple entities share the same value during dialog execution; user asked to clarify which to amend.
Max on-hold quantity exceeded — ask user to cancel a taskUser tried to put more tasks on hold than the configured maximum.
User provides intent when prompted for String/Description entityUser input matched an intent when prompted for a string/description entity; AI Agent asks how to proceed.

Add and Edit Responses

To add a new standard response:
  1. Select a category tab and hover over the target situation.
  2. Click Add Response.
  3. In the Add Channel Overrides dialog:
    • Select a channel (or Any Channel for all channels).
    • For channel-specific responses, select a task and optionally a field.
    add channel override
  4. In the Response editor:
    • Simple Mode — Enter plain text with optional formatting (bold, italics, hyperlinks, lists).
    • Advanced Mode — Enter JavaScript for custom channel-specific messages. Use response.message for the default message string and response.choices for the options array.
  5. Click Save.

Channel Override Examples

Slack:
var message = {};
var attachments = [];
var attachment = {
    'text': "Exception occurred during API execution",
    'color': '#FFCC00',
    'attachment_type': 'default'
};
attachments.push(attachment);
message.attachments = attachments;
print(JSON.stringify(message));
Facebook:
var message = { text: response.message };
var buttons = [];
for (var i = 0; i < response.choices.length; i++) {
    var choiceItem = response.choices[i];
    if (choiceItem.length > 20) {
        choiceItem = choiceItem.substring(0, 18) + '..';
    }
    buttons.push({
        content_type: 'text',
        title: choiceItem,
        payload: choiceItem
    });
}
message.quick_replies = buttons;
print(JSON.stringify(message));
SMS:
var message = {};
var indexArray = ['a','b','c','d','e','f','g','h','i','j','k','l','m','n','o','p','q','r','s','t','u','v','w','x','y','z'];
var text = response.message + '\n';
for (var i = 0; i < response.choices.length; i++) {
    text = text.concat(indexArray[i] + ') ' + response.choices[i] + '\n');
}
message.text = text;
print(JSON.stringify(message));
Email:
var message = {};
var text = response.message + ' ';
for (var i = 0; i < response.choices.length; i++) {
    text = text.concat('<u>' + response.choices[i] + '</u><br>');
}
message.text = text;
print(JSON.stringify(message));

Customizing Ambiguity Display with getChoices()

Use getChoices() to retrieve ambiguous intent choices and display them in a custom format (for example, as buttons):
var info = getChoices();
var message = {
    "type": "template",
    "payload": {
        "template_type": "button",
        "text": "Did you mean",
        "subText": "You can now customize response for did you mean",
        "buttons": []
    }
};
for (var i = 0; i < info.length; i++) {
    message.payload.buttons.push({
        "type": "postback",
        "title": info[i],
        "payload": "payload1"
    });
}
print(JSON.stringify(message));