---
sourceDocument: Australia IT Service Management
sourceDocumentLink: https://www.servicenow.com/docs/r/it-service-management

 Release :

    - australia

ft:locale :

    - en-US

ft:publication_title :

    - Australia IT Service Management

ft:clusterId :

    - itsm

bundleId :

    - itsm

workflow :

    - Technology


---

# Assess quality of a change request

# IT Service Management AI agent collection assess quality of a change request agentic workflow {#ariaid-title1}

Release version: Australia  
Updated March 12, 2026  
![](https://www.servicenow.com/docs/portal-asset/ico-clock) 11 minutes to read
Summarize  
![AI sparkle icon](https://servicenow.com/docs/portal-asset/ai-sparkle-icon) Summarized using AI  
This content was generated using new OpenAI-powered functionality. Results are provided on an as is basis and are not guaranteed to be accurate or complete.  

## Summary of IT Service Management AI Agent Collection: Assess Quality of a Change Request Agentic Workflow

The Assess Quality of a Change Request agentic workflow helps ServiceNow customers evaluate the quality of change requests by analyzing key fields and generating improvement suggestions.
It uses either an active change policy document or similar closed change requests as references to assess the change request.
The workflow must be run manually and requires thesnitsmaia.snaiachgqualityrole, included in theitilandsnchangewriteroles.
Show full answer Show less  

## Key Features

* **AI Agent Versions:** Two modes are available:
  * **Version 1 (Supervised):** Prompts users for confirmation before suggesting and setting field values, then records quality scores and work notes.
  * **Version 2 (Autonomous):** Automatically assesses and records quality scores and work notes without setting field values or user confirmation.
* **Change Policy Documents:** Define rating criteria for change request fields and guide agent assessments. Only one active policy per scope is allowed. Customers can create, activate, deactivate, and copy policies via the Change Policy Control table.
* **Field Assessments:** The agent evaluates short description, description, implementation plan, backout plan, test plan, risk and impact analysis, and justification, rating each from Excellent to Incomplete. Suggestions are provided for fields rated below Excellent.
* **Custom Fields Support:** Customers can customize extraction prompts for out-of-box or custom fields by overriding the `POLICYEXTRACTIONKEYS` in the `ChangeQualityUtil` script.
* **Integration with ServiceNow Otto Panel:** Users can initiate assessments by entering prompts referencing change request numbers. The AI agent returns detailed assessment results, ratings, and suggestions directly in the panel.
* **Change Quality Scores Tracking:** Quality assessments are recorded in the `AI Change Quality Scores` table, capturing overall and per-field scores, explanations, and policy references if applicable.
* **Analytics and Visualization:** Platform Analytics enables tracking change quality trends over time, with options to group by assignment group, model, or change model template, and pin reports to dashboards for continuous monitoring.

## Practical Use and Benefits

This workflow empowers ServiceNow customers to systematically evaluate and improve change request quality, increasing transparency and consistency in change management processes. By leveraging AI-driven assessments based on policy documents or historical changes, teams can identify gaps in documentation and planning early, reducing risk and improving implementation success.

Customers benefit from configurable AI agent modes that fit their operational preferences---choosing between supervised field updates or autonomous assessments. The ability to customize field evaluation criteria and integrate with the Otto panel streamlines adoption and enhances usability.

Tracking quality scores over time supports continuous improvement initiatives, helping organizations monitor compliance with change policies and identify areas for training or process refinement.

## Getting Started

* Assign the required role (`snitsmaia.snaiachgquality`) to users who will run the workflow.
* Enable the workflow display toggle in AI Agent Studio to activate the ServiceNow Otto panel integration.
* Create or maintain active change policy documents to guide assessments, or rely on fallback to similar closed changes.
* Configure AI agent versions according to your preference for supervised or autonomous operation.
* Use the ServiceNow Otto panel to manually run change request assessments and review AI-generated quality scores and improvement suggestions.
* Leverage Platform Analytics to monitor change quality trends and support your change management governance.  
Use the assess quality of a change request agentic workflow to assess the quality of a change request and generate suggestions to improve the information in the fields. The workflow uses an active change
policy document if one applies, or falls back to similar closed change requests.  
Note:  
Change request plans AI agent can be used to set the values in the change request. For more information, see .

## AI agent execution modes {#now-assist-itsm-aiagents-assess-quality-change-request-workflow__section_b54_clt_kkc}

The change quality assessor AI agent is available in two versions: version 1 (supervised mode) and version 2 (autonomous mode).

In version 1, supervised mode, the AI agent assesses the change request and prompts the user for confirmation. After the initial confirmation, the agent adds field suggestions to the change request. After a second confirmation, it
records the change score in the Change Quality Scores table and saves the work notes to the change request.

In version 2, autonomous mode, the AI agent acts without pausing for user confirmation. The agent assesses the change request and suggests improvements automatically. It records the change score in the Change Quality Scores table
and saves the work notes to the change request without setting values on change request fields. Use the Change request plans AI agent to update values on the change request field. For more information, see .

## Change quality assessor AI agent overview {#now-assist-itsm-aiagents-assess-quality-change-request-workflow__section_qrv_fhh_lgc}

The change quality assessor AI agent analyzes the quality of a change request and suggests values for change request fields to improve it.

When the workflow runs, the agent first checks for an active change policy document using the **Get Change Quality Policy Document** tool:

* Policy document found: The agent rates the change request using the policy document and suggests values only for the fields defined in it.
* No policy document found: The agent falls back to similar closed change requests to assess the change request.

{#now-assist-itsm-aiagents-assess-quality-change-request-workflow__ul_mbz_ft3_tjc}

The agent assesses the following areas and assesses each one:

* Short description
* Description
* Implementation plan
* Backout plan
* Test plan
* Risk and impact analysis
* Justification

{#now-assist-itsm-aiagents-assess-quality-change-request-workflow__ul_nbz_ft3_tjc}  
The assessment report covers the overall change request and each area listed. It rates information as Excellent, Very good, Good, Fair, Poor, Very poor, or Incomplete. If all fields are rated Excellent, no suggestions are provided. The report is added to the Work notes field of the change request. You can also view the report in the `ai_change_quality_score` table.  
Note:  
The assess quality of a change request agentic workflow has no trigger and must be run manually. To modify the workflow, duplicate it and adjust the settings to meet your requirements. To clone an agentic workflow that is available by default, you must run the semantic index for similarChangeRequests. The Business Rule (\[Chg Quality\] Trigger semantic index) does not run automatically for cloned workflows. Activate it manually or run the script in the business rule to confirm that the index runs. When you modify an agentic workflow, AI agent, or tool, update all instructions accordingly.

## Create a change policy document {#now-assist-itsm-aiagents-assess-quality-change-request-workflow__section_d21_453_tjc}

A change policy document defines how the agent rates each field of a change request. Change policy documents are stored in the Change Policy Control table. If no active change policy document applies, the workflow falls back to the
similar changes path.  
To create a change policy document:

1. Navigate to All and search for and open Change Policy Control.
2. Select New.
3. Enter a description and then set either the Change model field or the Change type field.
4. Attach the policy document to the record.
5. Select Submit.
{#now-assist-itsm-aiagents-assess-quality-change-request-workflow__ol_create_change_policy_doc}

When you save the record, the Ingest policy document business rule runs automatically. It calls to extract the policy criteria from the attached document, populates the Policies field
with the results, and selects the Active check box.  
Keep the following in mind when working with change policy documents:

* To deactivate a policy, clear the Active check box.
* Only one policy can be active for a scope at a time. Creating a new active policy automatically deactivates any existing active policy for that scope.
* To activate an existing policy, first deactivate the currently active one for that scope.
* To reuse a policy for another change request, select Copy Policy on an existing Change Policy Control record.
{#now-assist-itsm-aiagents-assess-quality-change-request-workflow__ul_czb_x53_tjc}

## Update custom fields using extraction prompts {#now-assist-itsm-aiagents-assess-quality-change-request-workflow__section_uwp_gdz_tjc}

Important:  
The extraction prompt for each field is defined in the `ChangeQualityUtilSNC` script. To support custom fields, override the prompt in the `ChangeQualityUtil` script instead.

`ChangeQualityUtilSNC` includes the following field entries for each out-of-box field: `justification`, `implementation_plan`, `backout_plan`, `test_plan`,
and `risk_impact_analysis`. To add custom fields or change how the agent evaluates an existing field, override `POLICY_EXTRACTION_KEYS` in the `ChangeQualityUtil` script. Changes there
are preserved across upgrades.

One entry is included as a starting point for customization:

u_custom_field
:   A placeholder for a custom field. Copy it, set `name` to the element name of your field, and write a `description` that explains what the field should contain. Add one entry per custom field.

The following example shows a `ChangeQualityUtil` override with a `u_custom_field` entry:

    POLICY_EXTRACTION_KEYS: [
        {
            "name": "justification",
            "description": "The business or technical reason explaining why the change is necessary. Look for references to a specific problem, incident, risk, compliance requirement, or performance goal. Should explain the consequence of not making the change. Examples: 'Resolving incident INC0012345 causing nightly job failures', 'Remediating CVE-2024-1234 on the authentication service', 'Meeting PCI-DSS audit requirement by 31 March deadline'."
        },
        {
            "name": "implementation_plan",
            "description": "The step-by-step sequence of actions required to execute the change. Look for numbered or ordered steps, each describing who does what, on which system, and when. Should reference specific scripts, config files, commands, or components. Examples: 'Step 1 - Prior to cutover, disable the scheduler service on app-server-01', 'Step 2 - Run migration script db_migrate_v4.sql on PROD-DB-02 during maintenance window', 'Step 3 - Restart the Orders API and confirm health check returns HTTP 200'."
        },
        {
            "name": "backout_plan",
            "description": "The specific steps to restore the previous state if the change fails or produces unacceptable results. Look for concrete rollback actions (not just 'revert the change'), trigger conditions that define when to initiate rollback, who is responsible, and whether rollback is feasible within the change window. Examples: 'If health check fails after Step 3, on-call engineer to restore previous WAR file from /backup/releases/v2.1.3 and restart service', 'Trigger: error rate exceeds 5% within 10 minutes of deployment --- rollback owner: J. Smith (primary implementer)'."
        },
        {
            "name": "test_plan",
            "description": "The specific tests or checks performed after implementation to confirm the change succeeded and nothing unintended broke. Look for named test cases, expected outcomes, who performs the testing, which environment (production post-deployment required), and a clear pass/fail criterion. Examples: 'Post-deployment: verify login endpoint returns HTTP 200 using test account svc-smoketest@acme.com', 'Confirm nightly batch job completes without errors in PROD within 15 minutes of deployment --- pass criterion: zero error entries in /var/log/batch/run.log'."
        },
        {
            "name": "risk_impact_analysis",
            "description": "An assessment of what could go wrong if the change fails, who or what would be affected, and how severely. Look for identified risks, affected users or services, likelihood/severity language, and blast radius estimates. Should be distinct from the implementation plan --- focused on failure scenarios, not planned actions. Examples: 'Low probability but high impact --- a failed schema migration could render the Orders API unavailable for up to 2 hours, affecting approximately 300 internal users', 'Risk of session timeouts during restart affecting users currently logged in, estimated 50 concurrent sessions'."
        },
        {
            "name": "u_custom_field",
            "description": "Explanation of what the custom field is"
        }
    ],

## Access the workflow {#now-assist-itsm-aiagents-assess-quality-change-request-workflow__section_kf2_wvz_kgc}

The workflow requires the `sn_itsm_aia.sn_aia_chg_quality` role. This role is included in the `itil` and `sn_change_write` roles.  
To make the workflow available in the panel:

1. Navigate to AllAI Agent StudioCreate and manage.
2. Open the Assess quality of a change request agentic workflow.
3. In the left navigation, select Select channels and status.
4. For Engage via the ServiceNow Otto panel, turn on the Display toggle.
5. Select Save and test.
{#now-assist-itsm-aiagents-assess-quality-change-request-workflow__ol_cqv_xvz_kgc}  
Note:  
Turning on the Display toggle is the only configuration required. The remaining configuration is available out of the box. Turning on this toggle also triggers a business rule that activates the text index the workflow requires for change requests.

## Configure change quality assessor AI agent versions {#now-assist-itsm-aiagents-assess-quality-change-request-workflow__section_chg_quality_version_settings}

Select the active version for each change quality assessor child AI agent. The version controls whether the agent surfaces recommendations only or also sets change request fields.

The following child AI agents each support two versions:

* Change quality assessor using change policy documents AI agent
* Change quality assessor using similar changes AI agent

Each agent supports the following versions:

* **Version 1 (supervised)**: The agent asks for confirmation before setting a field.
* **Version 2 (autonomous)**: The agent surfaces recommendations but does not set change request fields.

To set the active version:

1. Navigate to AllAI Agent StudioCreate and manage.
2. Open the child AI agent to configure.
3. On the Define the specialty page, select View versions.
4. In the View all versions dialog, select a version from the list.
5. Turn on the Set as active toggle.
6. Select Save.

{#now-assist-itsm-aiagents-assess-quality-change-request-workflow__ol_emc_cmt_kkc}  
Note:  
On the Add tools and information page, each tool shows an Execution mode of Autonomous or Supervised. Regardless of the active version, the Set Change Field and Set chosen Change Fields tools always run in supervised mode.

## Assess a change request from the ServiceNow Otto panel {#now-assist-itsm-aiagents-assess-quality-change-request-workflow__section_qrb_4v3_tjc}

To assess a change request, open the change request and enter a prompt in the ServiceNow Otto panel, such as assess quality of \<change request number\>. You can enter the change request number from the list of suggestions, add
keywords, or type a custom prompt.  
The agent returns a quality assessment in the panel. The assessment includes:

* An overall quality rating, for example, Very Good.
* A summary of the assessed fields with a justification for each rating.
* Suggested improvements for fields that are not rated Excellent.
{#now-assist-itsm-aiagents-assess-quality-change-request-workflow__ul_md1_tv3_tjc}

The results that the AI agent generates depend on the configured agent version.  
Version 1 (supervised mode): In Version 1, the agent asks for confirmation before it updates any field. On confirmation of a suggested value, the agent updates the field, saves the change request, and marks the update as AI Generated. The agent then asks whether to record the quality summary. On confirmation, the agent adds the summary to the Work notes field and creates a record in the AI Change Quality Scores table.  
Note:  
Select Version 1 for the agent to suggest and write field values. For more information, see Configure change quality assessor AI agent versions section in this topic.

Version 2 (autonomous mode): In Version 2, the agent automatically records the quality rating and the explanation as a work note on the change request. The agent also creates a
record in the AI Change Quality Scores table. This record stores the change request, the explanation, the per-field score, the rating, and the numerical score. The agent does not update any fields on the change request, such as
Impact, Priority, or State. Use the Change request plans AI agent to update values on the change request field. For more information, see .

The system keeps only the most recent rating for a change request. A second assessment of the same change request replaces the existing AI Change Quality Scores record instead of adding a record. If the
policy control record that a rating used is later deactivated, the existing score record continues to reference that policy control record. This reference persists as long as the change request is not reassessed. The score record
still shows which policy version applied and when it applied. The AI generated results are generated based on the agent version setup.

For more information, see [Request the generative AI capabilities in ITSM by using the ServiceNow Otto panel](https://www.servicenow.com/docs/AnJIDfOoTCaVTN3DW2Cxdg "Use the ServiceNow Otto panel to request the contextual generative AI capabilities in IT Service Management (ITSM) such as a chat summary, incident summary, or incident resolution notes in a conversational manner. You can also add comments and work notes. These capabilities provide you with a quick resolution to issues.").

## Change quality scores {#now-assist-itsm-aiagents-assess-quality-change-request-workflow__section_u2c_ww3_tjc}

The workflow records the quality summary in the AI Change Quality Scores \[`ai_change_quality_score`\] table. Each record stores the overall score, the rating, the explanation, and the per-field
scores for the change request. When a change policy document is used to rate the change request, the record also stores a reference to the Change Policy Control record. This allows change management teams to track how scores trend
over time as a policy document changes.

To view the recorded scores, search for AI Change Quality Scores. The table includes the following information:
{#now-assist-itsm-aiagents-assess-quality-change-request-workflow__table_chg_quality_scores__entry__2}

| Field | Description |
|-|-|
| Change Policy Control | The change policy document used to rate the change request. This field is empty when the change request is rated through the similar changes path. |
| Change Request | The change request that was rated. |
| Explanation | The natural language explanation of the overall score. |
| Rating | The overall quality rating, for example, Very Good. |
| Score | The overall numeric score, from 0 through 100. |
| Per field score | The per field rating values for the assessed fields. |
[ ]

{#now-assist-itsm-aiagents-assess-quality-change-request-workflow__table_chg_quality_scores}

If a record already exists for a change request, the workflow overwrites it with the latest assessment.

## Visualize change quality scores {#now-assist-itsm-aiagents-assess-quality-change-request-workflow__section_fzd_ghz_tjc}

Use Platform Analytics to track trends in the `ai_change_quality_score` table. A line chart of average score by month shows whether change quality is improving or declining.

Hover over a point to see the score for that month. Select Go to record list to open the records behind that data point.

Use Group by to break the trend down by a change request dimension, such as Assignment group, Model, or Change Model Template. This lets
you compare change quality across teams or change types.

When the chart is grouped, hover over a point to see the score and percentage for each group. Select Go to record list to view the change requests behind any value.

Select Add to Dashboard to pin the report to a dashboard for ongoing monitoring.

*[\>]: and then