Enhance contract search queries with contextual understanding, mapping natural language to contract fields, obligation types, clause categories, and CLM-specific terminology
Standardmäßig ist der Prompt ausgewählt, der zuerst die Quelle prüft. Sie können zu einem direkten Befehl wechseln oder eine lokale Kopie herunterladen.
Quelldateien prüfen
Lesen Sie SKILL.md und alle von SkillsMP angezeigten Begleitdateien, bevor Sie sich für eine Installation entscheiden.
Mit Codex oder Claude installieren Kopieren Sie diesen Prompt, fügen Sie ihn in Codex, Claude oder einen anderen Assistant ein und lassen Sie die Skill-Seite prüfen und installieren.
Ein direkter Befehl überspringt den Prüf-Prompt. Prüfen Sie die Quelle, bevor Sie ihn ausführen.
Enhance contract search queries with contextual understanding, mapping natural language to contract fields, obligation types, clause categories, and CLM-specific terminology
This skill transforms natural language contract search queries into precise ServiceNow queries, enabling users to find contracts without knowing table structures or encoded query syntax. It covers:
Mapping conversational search terms to contract table fields and encoded query operators
Resolving vendor names, contract types, and status values to their ServiceNow equivalents
Translating date expressions ("expiring next quarter", "signed last year") to encoded date queries
Understanding obligation-specific terminology and mapping to CLM obligation types
Combining multiple search criteria into optimized compound queries
Suggesting follow-up queries to refine broad result sets
When to use:
When users search for contracts using business language rather than field names
During contract discovery for audits, renewals, or due diligence projects
When building contract reports that require specific filtering criteria
When non-technical stakeholders need to find contracts without ServiceNow expertise
For automated contract monitoring that triggers on natural language conditions
Prerequisites
Roles:contract_reader, contract_manager, sn_clm.user, or admin
Plugins:com.snc.contract_management (Contract Management) or com.sn_clm (Contract Lifecycle Management)
Access: Read access to ast_contract, clm_obligation, core_company, sys_choice tables
Knowledge: Basic understanding of contract terminology and your organization's contract taxonomy
Run the query and verify results match the user's intent.
Using REST API:
GET /api/now/table/ast_contract?sysparm_query=contract_type=saas^vendor=[acme_sys_id]^ends<=javascript:gs.daysAgoEnd(-180)^ends>=javascript:gs.daysAgoEnd(0)^total_cost>100000^active=true&sysparm_fields=sys_id,number,short_description,vendor,contract_type,starts,ends,total_cost,auto_renew,notice_period&sysparm_limit=50&sysparm_display_value=true
Alternatively, use Natural Language Search for fuzzy matching:
Tool: SN-Natural-Language-Search
Parameters:
table_name: ast_contract
query: "SaaS contracts with Acme expiring in the next 6 months worth over 100 thousand dollars"
limit: 25
Step 5: Search Obligations by Context
Translate obligation-related queries to CLM table searches.
Example: "Find all vendor data protection obligations due this quarter"
Tool: SN-Execute-Background-Script
Parameters:
script: |
var suggestions = {
original_query: 'SaaS contracts with Acme expiring next 6 months over $100K',
results_found: 0,
refinement_suggestions: [],
broadening_suggestions: []
};
// Count results
var ga = new GlideAggregate('ast_contract');
ga.addQuery('contract_type', 'saas');
ga.addQuery('active', true);
ga.addQuery('ends', '<=', gs.daysAgoEnd(-180));
ga.addQuery('ends', '>=', gs.daysAgoEnd(0));
ga.addQuery('total_cost', '>', 100000);
ga.addAggregate('COUNT');
ga.query();
if (ga.next()) suggestions.results_found = parseInt(ga.getAggregate('COUNT'));
if (suggestions.results_found > 20) {
suggestions.refinement_suggestions = [
'Add department filter: "...in the Engineering department"',
'Narrow timeline: "...expiring in the next 30 days"',
'Add renewal filter: "...with auto-renewal enabled"'
];
} else if (suggestions.results_found == 0) {
suggestions.broadening_suggestions = [
'Remove value filter: "SaaS contracts with Acme expiring in 6 months"',
'Expand vendor: "SaaS contracts expiring in 6 months over $100K"',
'Expand type: "All contracts with Acme expiring in 6 months"'
];
}
gs.info('QUERY SUGGESTIONS:\n' + JSON.stringify(suggestions, null, 2));
description: "CLM: Generate query refinement suggestions based on result count"
Step 7: Build Saved Search Queries
Save frequently used query patterns for reuse:
Tool: SN-Create-Record
Parameters:
table_name: sys_filter
fields:
title: "Contracts Expiring Next 90 Days - Auto Renew"
table: ast_contract
filter: active=true^auto_renew=true^ends<=javascript:gs.daysAgoEnd(-90)^ends>=javascript:gs.daysAgoEnd(0)
Tool Usage
Operation
MCP Tool
REST Endpoint
Query Contracts
SN-Query-Table
GET /api/now/table/ast_contract
Fuzzy Search
SN-Natural-Language-Search
N/A
Resolve Choice Values
SN-Query-Table
GET /api/now/table/sys_choice
Resolve Vendors
SN-Query-Table
GET /api/now/table/core_company
Query Obligations
SN-Query-Table
GET /api/now/table/clm_obligation
Aggregate Counts
SN-Execute-Background-Script
POST /api/now/table/sys_trigger
Discover Fields
SN-Discover-Table-Schema
GET /api/now/table/sys_dictionary
Best Practices
Always Resolve References: Never pass raw text into reference fields; resolve vendor names and department names to sys_ids first
Date Safety: Use JavaScript date functions rather than hardcoded dates so queries remain valid over time
Progressive Refinement: Start with broader queries and offer refinements rather than returning zero results
Field Validation: Check sys_dictionary before querying non-standard fields to avoid "invalid field" errors
Display Values: Always request sysparm_display_value=true for user-facing results so reference fields show names
Compound Queries: For OR conditions, use ^OR or ^NQ (new query) operators; test complex boolean logic carefully
Performance: Add sysparm_limit to all queries; use indexed fields (number, sys_id, state) as primary filters
Encoding: URL-encode special characters in REST query parameters, especially ^, =, and date expressions
Troubleshooting
Query Returns No Results When Contracts Exist
Symptom: Natural language query translates correctly but returns zero results
Cause: Choice value mismatch (e.g., "saas" vs "SaaS" vs "software_as_a_service")
Solution: Always resolve choice values from sys_choice first:
Symptom: "Expiring next quarter" returns contracts from the wrong time period
Cause: JavaScript date functions use server timezone; gs.daysAgoEnd() counts from today
Solution: Verify server timezone with gs.getProperty('glide.sys.date.format'). Use BETWEEN with explicit quarter boundaries for precision.
Vendor Name Does Not Resolve
Symptom: Vendor lookup returns no results for a known vendor
Cause: The vendor name in core_company may differ from common usage (e.g., "International Business Machines" vs "IBM")
Solution: Use nameLIKE with partial match, or search across name, stock_symbol, and u_dba_name fields.
Examples
Example 1: Renewal Dashboard Query
User Query: "Show me all contracts that auto-renew and need notice within 60 days"
