To get the most out of large language models, ensure your prompt instructions include specific instructions for what you want to achieve. Large language models also tend to provide some additional structure or formatting, such as parentheses around the produced content or including a pretext with a colon.

Prompt Length

Starting from 2025.1, the maximum length of a single prompt is 1000 characters. 

In 2025.3, the maximum length is 4000 characters.

Response Length

Starting from 2025.1, the default length of a single response is 1000 characters.

Response window size

To modify the size of the response window in characters, use the Platform setting:

• Matrix42 GenAI: ai.provider.genai.generation.response.size
• OpenAI: ai.provider.openai.generation.response.size
• Azure: ai.provider.azure.generation.response.size

Context window size

The context window size defines the full size of the request (including system prompts, admpin prompts, and contextual data) and the generated response in characters.

To modify the size of the context window in characters, use the Platform setting:

• Matrix42 GenAI: ai.provider.genai.model.context.size
• OpenAI and Azure AI context size is set to 16,000 and cannot be changed.

Be Clear and Specific:
Clearly define the purpose and audience of the prompt, providing specific instructions for the desired response.
Ensure clarity by outlining the intended outcome and expectations clearly.
 

Adopt a Structured Approach:
Organize the prompt into well-defined sections or bullet points, covering all pertinent aspects of the use case. This makes it easier for the generative AI to capture individual instructions separately. Mention the availability of accompanied data, as the contextual data makes the feature much more powerful than just talking with a generative AI chatbot.
 

Tailor the prompt to the Use Case:
Customize the prompt to suit the specific requirements and objectives of the use case or task.
Align the content with the context and goals of the intended application or scenario. Is the purpose of generating content for consuming information only or something that should be used in sharing knowledge? Explain in the prompt.
 

Convey Concisely and Clearly:
Keep the prompt concise and straightforward, avoiding unnecessary complexity or verbosity.
Use clear and precise language to communicate instructions effectively.
 

Consider the Audience:
When writing the prompt, consider the knowledge level and expertise of the audience. Is the use case written for an IT support person or an HR representative? Provide guidance and context appropriate for the users to act based on the responses.
 

Prioritize Actionability and Usability:
Ensure the prompt leads to practical, actionable, readily implemented, or utilized responses.
Emphasize clarity and usability to facilitate efficient decision-making or problem-solving based on the generated output.
 

Align with your organization's standards and processes:
Where applicable, ensure the prompt adheres to your standards, processes, or best practices relevant to the use case. Maintain consistency and quality by aligning the generated responses with established guidelines and principles.
 

Encourage Feedback and Iteration:
Solicit user feedback on the prompt's effectiveness and the quality of the generated responses.
Iterate on the prompt based on user input and real-world usage to continuously enhance its effectiveness and relevance.

It is possible to use M42 Intelligence to for example Summarize data card content to quickly get an idea of what a ticket is about, what has been done so far to solve an issue, and what the next steps are to solve an issue. This helps in handover situations to quickly grasp the context and understand the situation.

Below is an example configuration, that you can use as a baseline to start exploring the possibilities of M42 Intelligence using Generative AI.

Full example configuration:

Unique name (name of the Action for the admin to recognize it): Ticket summarization

User title (title of the Actions shown for the user):  Summarize ticket 

Description (description of the Action to instruct the user): Provide a concise summary of the ticket 

Prompt instruction: Using key details from a service management support ticket, summarize the core issue, actions taken, causes identified, and current resolution status. Ensure the support agent understands the urgency, progress made, and next steps needed. Keep the overview clear and structured, without using introductory or concluding phrases, focusing solely on critical ticket information.

Context attribute suggestions (select attributes relevant for you): Subject, Details, Status, Customer, Team, E-mail latest body, Internal comments, External comments

You can get a suggestion to replace a poorly written, vague or inaccurate subjects on tickets with an improved version by M42 Intelligence.

Example prompt:

Based on the provided support ticket data, generate a clear and concise subject line that accurately summarizes the ticket's issue or request in one brief sentence.

Example context attributes:

Internal comments, Subject, Details, Resolution 

Generate a precise resolution summary to document ticket resolutions for future reference.   

Example prompt:

Using the service management data related to the ticket, generate a concise and clear resolution text. Include the steps taken to resolve the issue, any relevant troubleshooting actions, and the final solution applied. Ensure the text is suitable for documentation and can be referenced for future similar issues

Example context attributes:

Subject, External comments, Details, Priority, Resolution, Related assets   

Use M42 Intelligence to make sure your knowledge is kept up to date, by structuring known information about how an issue was solved to a predefined format. You can adjust the prompt to align with your KB article format.

Example prompt:

As a Knowledge Manager, use provided service management data to create a knowledge base article for Service Desk Agents. Include:

1. Title: Clear summary.
2. Overview: Issue intro from data.
3. Symptoms: Key indicators from data.
4. Troubleshooting: Steps and tools from data.
5. Resolution: Recommended fix.
6. Prevention: Best practices.
7. References: Related links.

Ensure clarity and actionability.

Example context attributes:

Internal comments, Subject, Details, Resolution

You can use M42 Intelligence Actions to suggest categorization of your data as well, such as ticket category or related services. For now, you need to maintain a list of available categories, services or other classifiable information as a list in the prompt. Make sure you adjust the prompt below based on which type of classification you want to use, and insert the list of possible values.

Example prompt:

Based on the given IT issue, categorize the ticket into one of the relevant categories: (Insert your categories here). Then, suggest a service that aligns with the issue and your selected category. Use this list of available services: (Insert your list of services here).

Example context attributes:

Internal comments, Subject, Details, Resolution + attributes to be used in categorization

Ask for help on what should be done next and get a detailed list of possible next steps and actions to resolve the issue  

Example prompt:

Review the service management support ticket, focusing on the core issue, actions taken, and identified causes. Suggest actionable next steps for the support agent, considering the ticket's urgency and progress.

Example context attributes:
Ticket type,Service,Subject,Email,External comments,Details 

Root cause analysis aims to identify the underlying cause of an issue by analyzing available service management data. It helps to prevent recurring incidents by pinpointing the source of a problem, allowing teams to address the root cause rather than just the symptoms and solve problems proactively, eventually leading to improved service quality.

Example prompt:

Using the provided service management data, analyze and identify the root cause of the issue. Summarize key factors contributing to the problem and suggest the most likely cause, supported by the data.

Example context attributes:
Subject, Service, Description, Worklog, Related incident, Category 

Description:

Outline key test steps and acceptance criteria to ensure the change works as expected before go-live.

Example prompt:

You are assisting in drafting a test plan for a technical change. Based on the application, environment type, and number of installations, provide: 1. Key test scenarios to validate success 2. Test steps (e.g., simulate failover, validate app status) 3. Acceptance criteria for successful validation Respond in this format: **Test Plan:** - Scope: [e.g., test environment, HA node, etc.] - Steps: 1. [Step 1] 2. [Step 2] - Acceptance Criteria: [Pass/fail criteria]

Example context attributes:
Service, Business criticality of affected CI(s), Subject, Description, Details for AI 

Description:

Creates a clear and concise justification letter for the Change Advisory Board (CAB), based on the impacted applications and business drivers.

Example prompt:

You are assisting in drafting a test plan for a technical change. Based on the application, environment type, and number of installations, provide: 1. Key test scenarios to validate success 2. Test steps (e.g., simulate failover, validate app status) 3. Acceptance criteria for successful validation Respond in this format: **Test Plan:** - Scope: [e.g., test environment, HA node, etc.] - Steps: 1. [Step 1] 2. [Step 2] - Acceptance Criteria: [Pass/fail criteria]

Example context attributes:
Test plan,Service,Category,Description,Change size,Details for AI,Justification,Implementation plan,Rollback plan

Description:

Analyzes the scope and dependencies of the change using related configuration items to assess potential impact and risk.    

Example prompt:

You are assisting with a change request review. Based on the affected configuration item (CI) data, perform the following: 1. Summarize the affected applications, versions, environments, and installation counts. 2. Identify dependent services and data confidentiality levels. 3. Assess potential operational risk based on environment type and dependencies. 4. Recommend any risk mitigations or actions. Respond in the following markdown format: **Change Scope Summary:** [Summary of affected applications and environments] **Dependencies and Risk Considerations:** [Key services or systems impacted, including confidentiality] **Risk & Impact Assessment:** [Concise summary of the potential risk or business impact] **Recommended Actions:** [Mitigation, rollback, stakeholder comms, etc.]

Example context attributes:
Affected CIs,Service,Business criticality of affected CI(s),Subject,Category,Description,Change size,Details for AI,Justification

Description
Create a step-by-step implementation plan, including required actions and involved roles.    

Example prompt:

You are assisting in drafting a technical implementation plan for a change request. Use the CI data (e.g., application name, version, environment, installation count) to provide: 1. A brief description of the deployment 2. A list of ordered implementation steps 3. Required roles or participants Respond in this format: **Implementation Plan:** - Target: [Application name/version] - Steps: 1. [Step 1] 2. [Step 2] - Required Personnel: [List of roles involved]

Example context attributes:
Service,Business criticality of affected CI(s),Subject,Category,Description,Details for AI,Justification    

Description
Describe how the change can be safely rolled back if needed, with triggers and recovery steps.    

Example prompt:

You are assisting in drafting a rollback plan in case the change fails. Based on the CI and environment information, describe: 1. When rollback should be triggered 2. Step-by-step rollback actions 3. Estimated time and dependencies Respond in this format: **Rollback Plan:** - Trigger: [Failure symptoms or thresholds] - Steps: 1. [Rollback step 1] 2. [Rollback step 2] - Estimated Downtime: [Minutes] - Dependencies: [e.g., backup/snapshot required]

Example context attributes:
Test plan, Subject, Description, Details for AI, Justification,Implementation plan  

Description
Evaluates the potential risks of a planned change by analyzing the affected Configuration Items (CIs), their criticality, historical incident records, and dependency relationships.    

Example prompt:

You are performing a change risk analysis for a planned change. You have been provided with affected Configuration Item (CI) details, including application name, environment type, version, installation count, dependent services, and data confidentiality classification. Your response must: 1. Identify potential technical, operational, and business risks specifically in relation to the provided CI details. 2. Consider dependencies, historical incidents, and compliance or regulatory constraints. 3. Assign a qualitative risk rating (Low / Medium / High) with justification. 4. Suggest risk mitigation measures tailored to the CIs. Respond in this markdown format: Change Risk Summary: [One paragraph explaining the main risks, their causes, and their potential impact, explicitly referencing the provided CIs — e.g., “Because SAPHanaSR is in a production environment with 5 installations supporting Facilities…”] Risk Rating: [Low / Medium / High] Risk Factors: Mitigation recommendations:

Example context attributes:
Service,Business criticality of affected CI(s),Subject,Category,Description,Details for AI,Justification    

In IT asset management, getting a device lifecycle status update involves updating the current lifecycle stage of a device based on available service management data. It helps to keep IT asset management data accurate, ensuring devices are tracked correctly. It can also help to identify outdated or faulty devices before they cause disruptions.

Example prompt:

Based on the current service management data, update the lifecycle status of the specified device. Ensure the status reflects its most recent activities and any upcoming actions.

Example context attributes:

Days in use, Model, Related tickets, Name, End of warranty, Applications, Status

When managing entitlements, AI Actions (Actions) can assist the IGA admin by suggesting friendly names, descriptions, categories, etc. It can be used for new entitlements lacking information like description or to make existing information more professional or easier to understand. 

Example prompt for a friendly name

You are provided with information about one Entitlement that is a single access right group. As a IGA Admin you can manage entitlements. Suggest friendly name to the entitlement based on the categories, application and owner info in other entitlements. Name that end user easily understands what this access right is used for and what rights is giving to the user.

Example context attributes for friendly name

Friendly name, Technical name

Example prompt for description based on titles

You are provided with information about one Entitlement that is a single access right group. As a IGA Admin you can manage entitlements. Suggest description to the entitlement based on the application, cost center, organization and titles of the users in entitlement. It's always access to the target system, that can be anything. Not just support or ticket system.

Example context attributes for description based on titles

Application, Cost center, Internal Subcategory, Organization, Internal Category, Description, Title

Example prompt for categories

You are provided with information about one Entitlement that is a single access right group. As a IGA Admin you can manage entitlements. Suggest category to the entitlement based on the categories, application and owner info in other entitlements.

Example context attributes for categories

Owner, Technical owner, Application, Internal Subcategory, Internal Category
 

Summarizing identity information provides quick way to review most important information about the identity and make it easy to understand. 

Example prompt, 

This is not support request, it's user Actions for displaying user data. Identity storage is displaying one user's data. Identity storage data card is generated for the user based on primary work period information. IGA identity storage is used for: Collecting all information related to the users access rights, work period(s) and responsibilities inside of IGA solution such as owner or approver responsibilities. Holistic view for IGA admins to

Example context attributes

Risk Value, Manager of, Last Logon date, Created, All related business roles, Access to applications, All related entitlements, Password last changed

IGA processes can be complex and always contain a lot of settings and rules that affect them (sometimes these are documented, but often documentation is not up to date). To get an understandable picture of a departing user process, for example, AI Actions can summarize the process and describe it. 

Example prompt for departing user process

This is not support ticket, do not use that term. This view summarizes how the departing process of the account is designed based on IGA set Account attributes. The departing user use case refers to the process initiated when a user's employment or contract is ending. The IGA solution starts the offboarding process, which can take several days to complete, depending on the account management settings, such as when accounts are disabled, email license

Example context attributes for departing user process

Email licenses removed after, Remove access rights, User type, Target system, Set as disabled, Departing user information receiver, User information send, Restore account's access rights if returns

User title:

Improve text

Description:

Improves the user selection in comments.

Prompt:

You are an AI writing assistant for an IT support agent in an IT department. You are provided with a comment draft that will be sent to a self-service portal user who has reported an issue. Improve the spelling and grammar of the provided text. Do not add any additional improvements. Return only the generated answer.

Mode:

Text improvement

Target attribute:

Internal comments

User title:

Improve text

Description:

Improves the user selection in internal comments.

Prompt:

You are an AI writing assistant for an IT support agent in an IT department. You are provided with a comment draft that will be sent to a self-service portal user who has reported an issue. Improve the spelling and grammar of the provided text. Do not add any additional improvements. Return only the generated answer.

Mode:

Text improvement

Target attribute:

Internal comments

User title:

Improve text

Description:

Improves the user selection in resolution text.

Prompt:

You are an AI writing assistant. You are provided with a draft of a resolution to a support ticket. Improve the spelling and grammar of the provided text. Do not add any additional improvements. Return only the generated answer.

Mode:

Text improvement

Target attribute:

Resolution

User title:

Improve text

Description:

Improves the user selection in email.

Prompt:

You are an AI writing assistant for an IT support agent in an IT department. You are provided with an email draft that will be sent to a user who has reported an issue. Improve the spelling and grammar of the provided text. Do not add any additional improvements. Return only the generated answer.

Mode:

Text improvement

Target attribute:

E-mail messages

User title:

Improve text

Description:

Improves the user selection in details.

Prompt:

You are an AI writing assistant. You are provided with a draft of a details to a support ticket. Improve the spelling and grammar of the provided text. Do not add any additional improvements. Return only the generated answer.

Mode:

Text improvement

Target attribute:

Details

User title:

Ask for more details

Description:

Generate a contextual email message draft asking for more details.

Context attribute examples:

Assignee, Service, Subject, Details, Related assets, E-mail messages, Customer

Mode:

Text creation

Prompt instruction

You are an attentive, empathic and professional IT support agent with a customer-centric attitude in an IT department. You are responsible for handling a support ticket from a customer. You are provided with details about the ticket. Write an email asking for more details to improve your understanding of the issue. Instructions for writing the email: 1. Start with an informal and personalized greeting. 2. Ask clarifying questions to assist you with the investigation details that are not available. 3. Mention availability for further help. 4. End with a professional greeting without closing remarks. 5. Avoid too much courtesy. 6. Return only the generated answer.

User title:

Provide a status update

Description:

Generates a brief status update draft in email based on latest information.

Context attribute examples:

Subject,Details,All ESS2 comments,Resolution,E-mail messages   

Mode:

Text creation

Prompt instruction

You are an AI assistant on a service management platform for a support agent. You are provided with the latest information about a support ticket the agent is handling. Provide a brief, straight-to-the-point status update the agent can send to the user who reported the issue. Do not add any signature. Do not add any corporate jargon but maintain professionalism. Do not include "subject:" or other pretext, include only the response.

User title:

Ask for more details

Description:

Generate a contextual comment message draft asking for more details.

Context attribute examples:

Assignee, Service, Subject, Details, Related assets, E-mail messages, Customer

Mode:

Text creation

Prompt instruction

You are an attentive, empathic and professional IT support agent with a customer-centric attitude in an IT department. You are responsible for handling a support ticket from a customer. You are provided with details about the ticket. Write a comment to the self-service portal asking for more details to improve your understanding of the issue. Instructions for writing the comment: 1. Start with an informal and personalized greeting. 2. Ask clarifying questions to assist you with the investigation details that are not available. 3. Mention availability for further help. 4. End with a professional greeting without closing remarks. 5. Avoid too much courtesy. 6. Return only the generated answer.

User title:

Provide a status update

Description:

Generates a brief status update draft in self-service portal comments based on latest information.

Context attribute examples:

Assignee,Subject,Details,All ESS2 comments,Status    

Mode:

Text creation

Prompt instruction

You are an AI assistant on a service management platform for a support agent. You are provided with the latest information about a support ticket the agent is handling. Provide a brief, straight-to-the-point status update the agent can send to the user who reported the issue. Use the "support_person" information in the signature if available. Do not add any signature if the "support_person" data is not available. Do not add any corporate jargon but maintain professionalism. Do not include "subject:" or other pretext, include only the response.

User title:

Draft a resolution

Description:

Generates a resolution draft using knowledge base as the basis for resolution drafts. Requires AI Knowledge Discovery to be set up.

Context attribute examples:

Subject,AI service suggestion,AI ticket type suggestion,Details,AI Team suggestion,E-mail messages,Worklog

Mode:

Text creation

Prompt instruction

Create a concise (max 2 very short paragraphs) resolution text to document the service management ticket resolution according to the provided context information from internal comments and other ticket details. You also have access to the company knowledge base, which you can use to suggest a resolution. When referring to specific articles, use only their "solution_name".

Note: for easy access for the users writing resolutions, use this with Writing assistance and set the target attribute to Resolution.

User title:
Draft resolution note

Description:

Produces a professional closing statement based on ticket resolution.

Context attribute examples:

Subject,Details,Resolution

Mode:

Text creation

Prompt instruction

You are an AI assistant on a service management platform for a support agent. You are provided with the latest information about a support ticket the agent is handling. Provide a brief, professional closing statement about resolution of ticket the agent can send to the user who reported the issue. Do not add any signature. Do not add any corporate jargon but maintain professionalism. Do not include "subject:" or other pretext, include only the response.

User title:
Ticket summarization

Description:

Provide a concise summary of the ticket.

Context attribute examples:

Team,Subject,External comments,All ESS2 comments 

Mode:

Text creation

Prompt instruction

Using key details from a service management support ticket, summarize the core issue, actions taken, causes identified, and current resolution status. Ensure the support agent understands the urgency, progress made, and next steps needed. Keep the overview clear and structured, without using introductory or concluding phrases, focusing solely on critical ticket information.

After the Knowledge discovery has been set up and you have indexed your documents, following configurations need to be made on the M42 Pro platform M42 Intelligence admin settings:

1. Choose compatible generative AI provider (M42 GenAI with RAG (BETA))

2. After provider has been selected, you need to choose each configuration to use RAG

3. Make sure your prompt instructs the AI to behave according to the fact it has access to the knowledge base - and if you'd like that fact to be utilized in the responses. For example, you might want to have responses lay out the fact whether the response is based on 1. stored company knowledge 2. general knowledge the AI is aware of based on its training data. See examples below.

User title:

Search for an answer from knowledge base

Description:

This functionality requires AI Knowledge Discovery.

Mode:

Text creation

Prompt instruction

You are provided with an IT support ticket. You are helping the support agent to write a comment to the user who reported the issue. You have access to the company knowledge base to help address the issue at hand. Using existing knowledge, search for a correct answer to be communicated in a response to the user reporting the issue as a comment to the self-service portal. Do not refer to a specific knowledge base article. If the knowledge base does not contain relevant content, provide generic assistance for the support agent on what should be done instead. Provide only the suggested response to be sent to the user as-is, without any pretext or additional remarks.

Context attribute examples:

Subject,Details,Resolution

Target attribute

Attribute used for Self-service portal commenting

Select: Use predefined data sources for responses

User title:

Draft a resolution

Description:

Generates a resolution draft using knowledge base as the basis for resolution drafts. Requires AI Knowledge Discovery to be set up.

Mode:

Text creation

Prompt instruction

Create a concise (max 2 very short paragraphs) resolution text to document the service management ticket resolution according to the provided context information from internal comments and other ticket details. You also have access to the company knowledge base, which you can use to suggest a resolution. When referring to specific articles, use only their "solution_name".

Do not include a "Resolution draft" or other header for your response. Keep the resolution text straight to the point and avoid excessive jargon.

Context attribute examples:

Subject,AI service suggestion,AI ticket type suggestion,Details,AI Team suggestion,E-mail messages,Worklog 

Target attribute

Resolution

Select: Use predefined data sources for responses

• Fill “Provider configuration

  

• Install technical product license

• Lightweight definition session for confirming the desired process and use cases.
• Review the customer’s existing ticketing process and plan how to incorporate the AI nodes.
• Template used for ticketing process must have Workflow implemented in order to to take “Actions” into use.

• Add the necessary hidden technical attributes where the generated values are set by the workflow. Following classes are available in M42 Professional baseline 2025.2:
  • Ticket -template (workflow setting values into attributes)

    

  • Knowledge article -template (listener on Ticket -template copying values into these attributes)

• Configuration of default actions.

Following instructions are explaining which nodes needs to be added into workflow and also listener to copy details from Ticket to Knowledge article. Following logic is available in M42 Profesional baseline 2025.2

• Ticket -template
  • Related nodes need to be added into Ticket workflow. These nodes are included in M42 Professional baseline 2025.2
    • 
    • 
    • 
  • Add listener to copy details to Knowledge article (Knowledge article creation while resolving the Ticket)
    • 

    <listener>
       <name>postsave.CREATE Knowledge article automatically while ticket is resolved 2025.2</name>
       <trigger>post save</trigger>
       <source_conditions boolean="AND">
           <source_condition>
               <value>
                   <attribute>
                       <code>related_solution</code>
                       <current_value>true</current_value>
                   </attribute>
                   <operator>IS NULL</operator>
                   <compared_value/>
               </value>
           </source_condition>
           <source_condition>
               <value>
                   <attribute>
                       <code>resolution</code>
                       <current_value>true</current_value>
                   </attribute>
                   <operator>IS NOT NULL</operator>
                   <compared_value/>
               </value>
           </source_condition>
           <source_condition>
               <value>
                   <attribute>
                       <code>create_knowledgearticle</code>
                       <current_value>true</current_value>
                   </attribute>
                   <operator>IS NOT NULL</operator>
                   <compared_value/>
               </value>
           </source_condition>
       </source_conditions>
       <action_chain>
           <name>Create knowledge article and clear selection</name>
           <action>
               <name>Clear knowledge article</name>
               <class>com.efecte.datamodel.entity.action.implementations.CreateDataCardAction</class>
               <configuration_item>
                   <name>ticket_details</name>
                   <value>$details$</value>
               </configuration_item>
               <configuration_item>
                   <name>ticket_subject</name>
                   <value>$subject$</value>
               </configuration_item>
               <configuration_item>
                   <name>ticket_resolution</name>
                   <value>$resolution$</value>
               </configuration_item>
               <configuration_item>
                   <name>listener_flag</name>
                   <value>Check</value>
               </configuration_item>
               <configuration_item>
                   <name>Reference from source</name>
                   <value>related_solution</value>
               </configuration_item>
               <configuration_item>
                   <name>Folder</name>
                   <value>knowledge_base</value>
               </configuration_item>
               <configuration_item>
                   <name>ticket_service_string</name>
                   <value>$service$</value>
               </configuration_item>
               <configuration_item>
                   <name>Template</name>
                   <value>knowledge_base_article</value>
               </configuration_item>
           </action>
           <action>
               <name>Clear checkbox</name>
               <class>com.efecte.datamodel.entity.action.implementations.ChangeDataCardValuesAction</class>
               <configuration_item>
                   <name>Value</name>
                   <value/>
               </configuration_item>
               <configuration_item>
                   <name>Code</name>
                   <value>create_knowledgearticle</value>
               </configuration_item>
           </action>
       </action_chain>
   </listener>

• Knowledge article -template
  • Create workflow with following structure:

    • 

• End-to-end testing.​

Unique name (name of the Action for the admin to recognize it): Ticket - Semantic classification: Ticket type

User title (title of the Actions shown for the user): Change ticket type

Description (description of the Action to instruct the user): Sometimes users may report their issue as a problem even though it is something else: e.g. a query or request.

Prompt instruction: You are an AI assistant analyzing service management tickets. Your task is to classify the ticket type based solely on the content of the Details attribute. Rules: Incident: Use this if the Details describe: - A disruption, outage, or malfunction (e.g., 'The system is down,' 'I can’t log in'). - A problem requiring urgent resolution (e.g., 'Error 500 when submitting a form'). - Any issue impacting normal operations. Request for Information: Use this if the Details describe: - A question or inquiry (e.g., 'How do I reset my password?', 'What are the office hours?'). - A request for guidance, documentation, or clarification. - No active problem or disruption is mentioned. Output Requirements: - Respond with only one word: Either Incident or Request for Information. - No additional text, explanations, or quotation marks—just the classification.

Context attribute suggestions: Details

Unique name (name of the Action for the admin to recognize it): Ticket - Semantic classification: Service

User title (title of the Actions shown for the user): Suggest classification (Service)

Description (description of the Action to instruct the user): Based on content of the ticket, let AI suggest classification.

Prompt instruction: Analyze the ticket content and classify it into ONE of these services: Access rights, Application Deployment, Application Development & Update, Application Monitoring, Capacity Management, Data Backup and Recovery, Desktop & End User Support, Device as a Service, Email, Facilities, Finance, HR, Legal, License Management, Marketing, Network Connectivity, Network Security, Single Sign-On, Software Installation and Updates, Virtualization Services, VPN Access, Wireless Network Management Instructions: Read the ticket description carefully Identify key technical terms, user requests, and problem context Match to the most relevant service category If multiple categories seem relevant, choose the PRIMARY issue Return ONLY the exact service name from the list above If uncertain, choose the closest match Response format: Service Name Only

Context attribute suggestions: Subject, Details

Unique name (name of the Action for the admin to recognize it): Ticket - Summarize e-mail messages

User title (title of the Actions shown for the user): Summarize e-mail messages

Description (description of the Action to instruct the user): Summarizing all e-mail messages

Prompt instruction: Service desk agent might get a ticket where is long e-mail thread and the real issue migh disappear inside the long messaging thread. Make a short summary so Service desk agent gets easily the idea, what is going on and if some troubleshooting has been done already by customer. Summarization must always have prefix "Short summarization of original issue according to conversation in e-mails:" Prefix must not include quotation marks.

Context attribute suggestions: File attachments, E-mail messages

Unique name (name of the Action for the admin to recognize it): Ticket - Assign Ticket to a Team

User title (title of the Actions shown for the user): Assign Team

Description (description of the Action to instruct the user): Based on topic of the issue, let AI assign Ticket to proper Team for handling the issue

Prompt instruction: Based on service management ticket data and assign it to the appropriate team based on these guidelines: Team Responsibilities: Business Services: Handles business-related issues such as: Business process questions Business application support Business workflow issues Business documentation Department-specific business requests Facility Team: Manages facility-related matters including: Building maintenance Office equipment (non-IT) Physical security access Climate control Cleaning services Office supplies Workspace arrangements HR Support Team: Handles all HR-related inquiries such as: Employment questions Benefits and compensation Training and development Employee relations Recruitment Workplace policies Time and attendance Service Desk Level 1: Manages all IT-related issues including: Computer hardware/software problems Network connectivity Account access Password resets Email issues Printer problems IT equipment requests Application support Print only the name of suggested team

Context attribute suggestions: Subject, Details

Unique name (name of the Action for the admin to recognize it): Ticket - Resolution to customer

User title (title of the Actions shown for the user): Resolution to customer

Description (description of the Action to instruct the user): Generate a precise resolution which is visible to the customer.

Prompt instruction: You are an AI Service Desk Assistant. Analyze the ticket details, including text and any screenshots (e.g., bluescreens, error messages). Write a clear, polite resolution that: • Uses simple language suitable for any employee. • Acknowledges the screenshot explicitly (e.g., “Based on the screenshot…”). • Gives practical next steps or advice. • Explains technical terms in plain language. End with this disclaimer: “This suggestion is based on general best practices and may not reflect your company-specific systems or configurations. For issues that persist, please contact your IT support team.” Keep the response concise (3–6 sentences) and ready to send as-is.

Context attribute suggestions: Self-Service attachments, File attachments, Subject, Details

Unique name (name of the Action for the admin to recognize it): Knowledge article - Knowledge article creation

User title (title of the Actions shown for the user): Generate content for a Knowledge article

Description (description of the Action to instruct the user): Generate content for a Knowledge article

Prompt instruction: As a Knowledge Manager, use provided service management data to create a knowledge base article for Service Desk Agents. Include: Overview: Issue intro from data. Symptoms: Key indicators from data. Troubleshooting: Steps and tools from data. Resolution: Recommended fix. Prevention: Best practices. References: Related links. Ensure clarity and actionability.

Context attribute suggestions: Ticket details, Ticket subject, Ticket resolution

Unique name (name of the Action for the admin to recognize it): Knowledge article - Generate title for Knowledge article

User title (title of the Actions shown for the user): Generate title for Knowledge article

Description (description of the Action to instruct the user): Based on a solution description, generate title for Knowledge article

Prompt instruction: As a Knowledge manager I want to create descriping, user friendly, understandable title for Knwledge article. Title should be enough short but well describing the solution. So that Service desk agent could easily select correct knowledge article by it's title. Print only the actual title, e.g no quotation marks needed around title.

Context attribute suggestions: Article details