Generic CRUD Operations

Overview

ServiceNow MCP provides generic tools that work on ANY table in the platform. This skill teaches you how to perform Create, Read, Update, and Delete (CRUD) operations across 160+ ServiceNow tables without needing specialized tools for each.

• What problem does it solve? Provides a single, consistent approach to table operations
• Who should use this skill? Developers, administrators, and integrators working with ServiceNow data
• What are the expected outcomes? Ability to query, create, update, and retrieve records from any table

Prerequisites

• MCP ServiceNow server configured and running
• Appropriate ServiceNow roles for target tables (varies by table)
• Basic understanding of ServiceNow table structure (sys_id, display values)

Procedure

Step 1: Discover Table Schema

Before working with any table, understand its structure using schema discovery.

Get Basic Schema:

Tool: SN-Get-Table-Schema
Parameters:
  table_name: incident

Response includes:

• Field names and types
• Field lengths and mandatory flags
• Reference table relationships
• Choice values for choice fields

Get Extended Schema:

Tool: SN-Discover-Table-Schema
Parameters:
  table_name: incident
  include_relationships: true

Response includes:

• All base schema information
• Parent/child table relationships
• Referenced tables and foreign keys

Step 2: Query Records (Read)

Query any table using filters, field selection, and pagination.

Basic Query:

Tool: SN-Query-Table
Parameters:
  table_name: incident
  query: active=true
  limit: 10

Query with Field Selection:

Tool: SN-Query-Table
Parameters:
  table_name: incident
  query: active=true^priority=1
  fields: sys_id,number,short_description,state,assigned_to
  limit: 20

Query with Pagination:

Tool: SN-Query-Table
Parameters:
  table_name: incident
  query: active=true
  fields: number,short_description
  limit: 50
  offset: 100

Query with Ordering:

Tool: SN-Query-Table
Parameters:
  table_name: incident
  query: active=true^ORDERBYDESCpriority
  fields: number,short_description,priority
  limit: 25

Step 3: Get Single Record (Read)

Retrieve a specific record by sys_id.

Get Record:

Tool: SN-Get-Record
Parameters:
  table_name: incident
  sys_id: abc123def456...

Get Record with Field Selection:

Tool: SN-Get-Record
Parameters:
  table_name: incident
  sys_id: abc123def456...
  fields: number,short_description,state,priority,assigned_to

Step 4: Create Records (Create)

Create new records in any table with proper field values.

Create Record:

Tool: SN-Create-Record
Parameters:
  table_name: incident
  data:
    short_description: Network connectivity issue in Building A
    description: Users in Building A floor 3 unable to access network resources
    category: network
    subcategory: connectivity
    impact: 2
    urgency: 2
    caller_id: <user_sys_id>
    assignment_group: <group_sys_id>

Response Returns:

• sys_id of the created record
• number (if auto-generated)
• All created field values

Create Record with Reference Fields:

Tool: SN-Create-Record
Parameters:
  table_name: task
  data:
    short_description: Follow up with customer
    parent: <parent_task_sys_id>
    assigned_to: <user_sys_id>
    assignment_group: <group_sys_id>
    due_date: 2024-12-31

Step 5: Update Records (Update)

Update existing records by sys_id.

Update Single Field:

Tool: SN-Update-Record
Parameters:
  table_name: incident
  sys_id: abc123def456...
  data:
    state: 2

Update Multiple Fields:

Tool: SN-Update-Record
Parameters:
  table_name: incident
  sys_id: abc123def456...
  data:
    state: 2
    assigned_to: <user_sys_id>
    work_notes: Investigation in progress
    priority: 2

Update Reference Fields:

Tool: SN-Update-Record
Parameters:
  table_name: incident
  sys_id: abc123def456...
  data:
    assignment_group: <new_group_sys_id>
    assigned_to: <new_user_sys_id>

Step 6: Batch Operations

Perform multiple operations efficiently.

Batch Create:

Tool: SN-Batch-Create
Parameters:
  records:
    - table: sys_user_group
      data:
        name: Network Support L1
        description: Level 1 network support
    - table: sys_user_group
      data:
        name: Network Support L2
        description: Level 2 network support
    - table: sys_user_group
      data:
        name: Network Support L3
        description: Level 3 network support

Batch Update:

Tool: SN-Batch-Update
Parameters:
  updates:
    - table: incident
      sys_id: abc123...
      data:
        state: 6
        resolution_code: Solved
    - table: incident
      sys_id: def456...
      data:
        state: 6
        resolution_code: Solved
    - table: incident
      sys_id: ghi789...
      data:
        state: 6
        resolution_code: Solved

Common Field Patterns

System Fields (All Tables)

Reference Fields

Reference fields link to records in other tables.

Setting a Reference:

{
  "assigned_to": "5137153cc611227c000bbd1bd8cd2007"
}

Getting Display Values:

Tool: SN-Query-Table
Parameters:
  table_name: incident
  query: active=true
  fields: number,assigned_to,assigned_to.name,assigned_to.email
  limit: 10

Response includes:

{
  "assigned_to": {
    "value": "5137153cc611227c000bbd1bd8cd2007",
    "display_value": "John Smith"
  },
  "assigned_to.name": "John Smith",
  "assigned_to.email": "john.smith@example.com"
}

Choice Fields

Choice fields have predefined values.

Get Available Choices:

Tool: SN-Query-Table
Parameters:
  table_name: sys_choice
  query: name=incident^element=state
  fields: value,label

Common Incident States:

Date/Time Fields

Date Format: YYYY-MM-DD DateTime Format: YYYY-MM-DD HH:MM:SS

Example:

{
  "due_date": "2024-12-31",
  "work_start": "2024-12-01 09:00:00"
}

Tool Usage

MCP Tools (Claude Code/Desktop)

REST API (ChatGPT/Other)

REST Query Parameters:

Native Tools (Claude Code)

Use the MCP tools above as the primary path. Raw REST is a last resort only when an operation is unavailable through MCP. In that case, rely on an organization-approved local credential helper that is already configured outside the conversation. Never ask the user to paste authentication material, embed it in a command, or generate a request containing authentication headers. Preserve the same schema checks, field selection, pagination, ACLs, and explicit instance targeting used by the MCP workflow.

Best Practices

• Always Check Schema First: Use SN-Get-Table-Schema before working with unfamiliar tables
• Specify Fields: Include only needed fields to reduce response payload and improve performance
• Use Pagination: For large datasets, use limit and offset to page through results
• Validate sys_ids: Reference fields require valid sys_ids from the referenced table
• Handle Errors: Check response for error messages, especially on create/update operations
• Batch Operations: Use batch tools for multiple records (5-10+ operations) for efficiency

Troubleshooting

Record Not Found

Symptom: Query returns empty results or get returns 404 Cause: Invalid sys_id, record deleted, or insufficient permissions Solution: Verify sys_id exists and user has read access to the table

Permission Denied (403)

Symptom: 403 Forbidden error on create/update Cause: User lacks write permissions to the table or specific fields Solution: Check ACLs and required roles for the table operation

Invalid Field Value

Symptom: Create/update fails with field validation error Cause: Value doesn't match field type or choice list Solution: Check field schema for valid values, especially for choice and reference fields

Reference Field Rejected

Symptom: Reference field value not saved Cause: Invalid sys_id or referenced record doesn't exist Solution: Query the reference table to get valid sys_ids

Mandatory Field Missing

Symptom: Create fails with mandatory field error Cause: Required field not provided in data Solution: Check schema for mandatory fields and include all required values

Examples

Example 1: Query Active Users

Tool: SN-Query-Table
Parameters:
  table_name: sys_user
  query: active=true
  fields: sys_id,user_name,name,email,department
  limit: 50

Example 2: Create Configuration Item

Tool: SN-Create-Record
Parameters:
  table_name: cmdb_ci_server
  data:
    name: web-server-01
    ip_address: 192.168.1.100
    os: Linux Red Hat
    cpu_count: 4
    ram: 16384
    classification: Production
    environment: Production

Example 3: Update Multiple Incidents

Tool: SN-Batch-Update
Parameters:
  updates:
    - table: incident
      sys_id: inc001...
      data:
        assignment_group: <network_team_id>
        work_notes: Routing to network team per triage
    - table: incident
      sys_id: inc002...
      data:
        assignment_group: <network_team_id>
        work_notes: Routing to network team per triage

Example 4: Get Record with Dot-Walking

Tool: SN-Query-Table
Parameters:
  table_name: incident
  query: number=INC0010001
  fields: number,short_description,caller_id.name,caller_id.email,caller_id.department.name

Example 5: Paginated Query

# Page 1
Tool: SN-Query-Table
Parameters:
  table_name: change_request
  query: state=-1
  fields: number,short_description,start_date,end_date
  limit: 100
  offset: 0

# Page 2
Tool: SN-Query-Table
Parameters:
  table_name: change_request
  query: state=-1
  fields: number,short_description,start_date,end_date
  limit: 100
  offset: 100

Related Skills

• itsm/natural-language-queries - Natural language search mastery
• itsm/quick-reference - ITSM quick reference card
• admin/batch-operations - Advanced batch operations

References

• ServiceNow Table API
• Encoded Query Strings
• REST API Best Practices