Install
$ agentstack add skill-happy-technologies-llc-happy-platform-skills-approval-summarization ✓ scanned · ✓ verified, works with Claude Code, Cursor, and more.
Security review
✓ PassedNo issues found. Passed automated security review. · v0.1.0 How review works →
- ✓ Prompt-injection patterns
- ✓ Secret / credential exfiltration
- ✓ Dangerous shell & filesystem operations
- ✓ Untrusted network calls
- ✓ Known-malicious package signatures
What it can access
- ✓ Network access No
- ✓ Filesystem access No
- ✓ Shell / process execution No
- ✓ Environment & secrets No
- ✓ Dynamic code execution No
From automated source analysis of v0.1.0. “Used” means the capability is present in the source — more access means more to trust, not that it’s unsafe.
Verified badge
Passed review? Show it. Paste this badge into your README, it links to the public security report.
Reliability & compatibility
Declared compatibility
Compatibility is declared by the source manifest. End-to-end runtime verification is coming, see below.
We're building live execution health for every listing: tool-call success rate, median latency, uptime, and last-checked timestamps, measured, not self-reported. It isn't live yet, so we don't show numbers we can't stand behind.
How agent discovery & health will work →About
Approval Summarization
Overview
This skill generates comprehensive approval summaries that give approvers the context they need to make informed decisions. It covers:
- Retrieving approval requests from
sysapproval_approverwith full request and item context - Enriching approvals with requester profile data from
sys_userincluding role, department, and manager - Extracting variable values (form responses) from
sc_req_itemto show exactly what was requested - Analyzing historical approval patterns from
sc_requestfor similar requests - Generating structured approval summaries with recommendation signals
- Identifying cost, risk, and compliance considerations for each approval
When to use: When approvers need a quick, comprehensive summary of pending approvals, when building approval digest notifications, or when creating approval dashboards with contextual data.
Value proposition: Reduces approval decision time by presenting all relevant context in a single view, decreases approval bottlenecks, and improves decision quality by surfacing historical precedent and risk factors.
Prerequisites
- Roles:
approver_user,itil, oradmin - Access: Read access to
sysapproval_approver,sc_request,sc_req_item,sc_cat_item,item_option_new, andsys_user - Knowledge: Understanding of organizational approval policies and delegation rules
Procedure
Step 1: Retrieve Pending Approvals
Query for all approvals awaiting action by a specific approver or group.
Using MCP (Claude Code/Desktop):
Tool: SN-Query-Table
Parameters:
table_name: sysapproval_approver
query: approver=[approver_sys_id]^state=requested
fields: sys_id,sysapproval,sysapproval.number,sysapproval.short_description,sysapproval.sys_class_name,state,approver,group,sys_created_on,due_date,expected_start,comments
limit: 25
order_by: sys_created_on
Using REST API:
GET /api/now/table/sysapproval_approver?sysparm_query=approver=[approver_sys_id]^state=requested&sysparm_fields=sys_id,sysapproval,sysapproval.number,sysapproval.short_description,sysapproval.sys_class_name,state,approver,group,sys_created_on,due_date,comments&sysparm_limit=25&sysparm_display_value=true
Get group approvals:
Tool: SN-Query-Table
Parameters:
table_name: sysapproval_approver
query: group=[group_sys_id]^state=requested
fields: sys_id,sysapproval.number,sysapproval.short_description,state,group,sys_created_on,due_date
limit: 25
Step 2: Retrieve Request and Requested Item Details
For each approval, pull the full request and item context.
Get the requested item details:
Tool: SN-Query-Table
Parameters:
table_name: sc_req_item
query: request=[request_sys_id]
fields: sys_id,number,cat_item,cat_item.name,cat_item.short_description,cat_item.price,quantity,requested_for,requested_for.name,requested_for.department,requested_for.manager,opened_by,opened_by.name,stage,state,price,description
limit: 10
Using REST API:
GET /api/now/table/sc_req_item?sysparm_query=request=[request_sys_id]&sysparm_fields=sys_id,number,cat_item,cat_item.name,cat_item.short_description,cat_item.price,quantity,requested_for,requested_for.name,requested_for.department,requested_for.manager,opened_by,opened_by.name,stage,state,price,description&sysparm_limit=10&sysparm_display_value=true
Step 3: Extract Variable Values (Form Responses)
Retrieve the specific answers the requester provided on the catalog item form.
Using MCP:
Tool: SN-Query-Table
Parameters:
table_name: sc_item_option_mtom
query: request_item=[ritm_sys_id]
fields: sc_item_option.item_option_new.question_text,sc_item_option.value,sc_item_option.item_option_new.name,sc_item_option.item_option_new.type
limit: 30
Using background script for clean variable extraction:
Tool: SN-Execute-Background-Script
Parameters:
description: Extract catalog variable values for RITM
script: |
var ritmId = '[ritm_sys_id]';
var ritm = new GlideRecord('sc_req_item');
ritm.get(ritmId);
var variables = [];
var opts = new GlideRecord('sc_item_option_mtom');
opts.addQuery('request_item', ritmId);
opts.query();
while (opts.next()) {
var option = opts.sc_item_option;
var varDef = option.item_option_new;
if (varDef) {
variables.push({
question: varDef.question_text.toString(),
answer: option.value.getDisplayValue(),
variable_name: varDef.name.toString(),
mandatory: varDef.mandatory.toString() === 'true'
});
}
}
var result = {
ritm_number: ritm.number.toString(),
item_name: ritm.cat_item.getDisplayValue(),
requested_for: ritm.requested_for.getDisplayValue(),
department: ritm.requested_for.department.getDisplayValue(),
variables: variables
};
gs.info(JSON.stringify(result, null, 2));
Step 4: Retrieve Requester Context
Gather relevant information about the person making the request.
Using MCP:
Tool: SN-Query-Table
Parameters:
table_name: sys_user
query: sys_id=[requested_for_sys_id]
fields: sys_id,name,user_name,email,title,department,manager,location,cost_center,company,vip,active
limit: 1
Using REST API:
GET /api/now/table/sys_user/[requested_for_sys_id]?sysparm_fields=sys_id,name,user_name,email,title,department,manager,location,cost_center,company,vip,active&sysparm_display_value=true
Step 5: Analyze Historical Approval Patterns
Find similar past approvals to provide precedent for the approver's decision.
Using MCP:
Tool: SN-Execute-Background-Script
Parameters:
description: Analyze historical approval patterns for similar requests
script: |
var catItemId = '[cat_item_sys_id]';
var departmentId = '[department_sys_id]';
var history = {
same_item_total: 0,
same_item_approved: 0,
same_item_rejected: 0,
same_dept_total: 0,
avg_approval_time_hours: 0,
recent_approvals: []
};
// Same item across all departments (last 6 months)
var ga = new GlideAggregate('sysapproval_approver');
ga.addQuery('sysapproval.sys_class_name', 'sc_req_item');
ga.addQuery('sys_created_on', '>=', gs.monthsAgoStart(6));
ga.addJoinQuery('sc_req_item', 'sysapproval', 'sys_id')
.addCondition('cat_item', catItemId);
ga.addAggregate('COUNT');
ga.addAggregate('COUNT', 'state');
ga.groupBy('state');
ga.query();
while (ga.next()) {
var state = ga.state.toString();
var count = parseInt(ga.getAggregate('COUNT'));
history.same_item_total += count;
if (state === 'approved') history.same_item_approved = count;
if (state === 'rejected') history.same_item_rejected = count;
}
// Recent approvals for same item
var recent = new GlideRecord('sysapproval_approver');
recent.addQuery('sysapproval.sys_class_name', 'sc_req_item');
recent.addQuery('state', 'approved');
recent.addQuery('sys_created_on', '>=', gs.monthsAgoStart(3));
recent.addJoinQuery('sc_req_item', 'sysapproval', 'sys_id')
.addCondition('cat_item', catItemId);
recent.orderByDesc('sys_updated_on');
recent.setLimit(5);
recent.query();
while (recent.next()) {
history.recent_approvals.push({
approved_by: recent.approver.getDisplayValue(),
date: recent.sys_updated_on.getDisplayValue(),
request: recent.sysapproval.getDisplayValue()
});
}
history.approval_rate = history.same_item_total > 0
? Math.round((history.same_item_approved / history.same_item_total) * 100) + '%'
: 'N/A';
gs.info(JSON.stringify(history, null, 2));
Step 6: Generate the Approval Summary
Compile all gathered context into a structured summary.
Using MCP:
Tool: SN-Add-Work-Notes
Parameters:
table_name: sysapproval_approver
sys_id: [approval_sys_id]
work_notes: |
=== APPROVAL SUMMARY ===
REQUEST: REQ0045123 / RITM0067890
ITEM: New Laptop Request - Performance Model (Dell Latitude 7640)
PRICE: $1,800.00
REQUESTER:
- Name: John Smith (john.smith@company.com)
- Title: Senior Software Engineer
- Department: Engineering
- Manager: Jane Doe
- Location: Building A, Floor 3
- VIP: No
WHAT'S BEING REQUESTED:
- Laptop Model: Performance (Dell Latitude 7640)
- RAM: 32GB
- Storage: 1TB SSD
- Additional Software: Docker Desktop, IntelliJ IDEA
- Business Justification: "Current laptop is 4 years old and unable to run containerized development environments. Build times exceed 30 minutes affecting productivity."
- Replacing Existing: Yes (Asset Tag: LAP-2022-0456)
HISTORICAL CONTEXT:
- Same item ordered 47 times in last 6 months
- Approval rate: 94% (44 approved, 3 rejected)
- Engineering department: 18 orders (all approved)
- Recent approvals: 3 in last 30 days by IT Director
RISK ASSESSMENT:
- Cost: Within standard budget threshold ($2,500)
- Justification: Strong (productivity impact documented)
- Precedent: Consistent with approval history
- Compliance: Asset replacement documented
RECOMMENDATION SIGNAL: Approve
Step 7: Batch Summarize Multiple Approvals
Generate digest summaries for approvers with multiple pending items.
Using MCP:
Tool: SN-Execute-Background-Script
Parameters:
description: Generate approval digest for approver
script: |
var approverId = '[approver_sys_id]';
var digest = { approver: '', pending_count: 0, total_value: 0, approvals: [] };
var approver = new GlideRecord('sys_user');
approver.get(approverId);
digest.approver = approver.name.toString();
var ga = new GlideRecord('sysapproval_approver');
ga.addQuery('approver', approverId);
ga.addQuery('state', 'requested');
ga.orderBy('sys_created_on');
ga.query();
while (ga.next()) {
digest.pending_count++;
var ritm = new GlideRecord('sc_req_item');
if (ritm.get(ga.sysapproval.toString())) {
var price = parseFloat(ritm.price) || 0;
digest.total_value += price;
digest.approvals.push({
approval_id: ga.sys_id.toString(),
request_number: ritm.request.getDisplayValue(),
ritm_number: ritm.number.toString(),
item: ritm.cat_item.getDisplayValue(),
requested_for: ritm.requested_for.getDisplayValue(),
department: ritm.requested_for.department.getDisplayValue(),
price: price,
waiting_since: ga.sys_created_on.getDisplayValue()
});
}
}
digest.total_value = '$' + digest.total_value.toFixed(2);
gs.info(JSON.stringify(digest, null, 2));
Tool Usage
MCP Tools Reference
| Tool | When to Use | |------|-------------| | SN-Query-Table | Query approvals, requests, items, variables, and user profiles | | SN-Get-Record | Retrieve individual approval or request records | | SN-NL-Search | Find requests or items by description keywords | | SN-Execute-Background-Script | Batch analysis, historical patterns, digest generation | | SN-Add-Work-Notes | Post approval summaries to approval or request records |
REST API Reference
| Endpoint | Method | Purpose | |----------|--------|---------| | /api/now/table/sysapproval_approver | GET | Query pending approvals | | /api/now/table/sc_request | GET | Retrieve parent request details | | /api/now/table/sc_req_item | GET | Get requested item details and variables | | /api/now/table/sc_cat_item | GET | Look up catalog item definitions | | /api/now/table/sc_item_option_mtom | GET | Extract variable values from requests | | /api/now/table/sys_user | GET | Get requester profile information |
Best Practices
- Include business justification prominently: The justification variable is the most important field for approvers; always surface it first
- Show total cost clearly: Include item price, quantity, recurring costs, and any variable-based pricing
- Provide historical context: Approval rate and precedent data significantly reduce decision uncertainty
- Flag anomalies: Highlight when a request is unusual (first-time item, exceeds typical cost, VIP requester)
- Calculate wait time: Show how long the approval has been pending to create urgency awareness
- Group by priority: Sort digest approvals by age, cost, or business criticality rather than creation date
- Respect data access: Only include information the approver has permission to view
- Automate digests: Schedule daily or twice-daily approval digest generation for managers with high approval volumes
Troubleshooting
Variable Values Not Found
Cause: The sc_item_option_mtom join is not returning expected variable values Solution: Some items use sc_item_option directly. Try querying sc_item_option with request_item=[ritm_sys_id] instead. For older requests, variables may be stored differently.
Approval Record Shows Wrong Request
Cause: The sysapproval field on the approval record may reference a request (REQ) rather than a requested item (RITM) Solution: Check sysapproval.sys_class_name to determine whether the approval is linked to sc_request or sc_req_item, and adjust your joins accordingly.
Historical Data Incomplete
Cause: Old approval records may have been archived or purged Solution: Adjust the date range in historical queries. Check if the instance uses table rotation or archiving on sysapproval_approver.
Requester Department Not Populating
Cause: The department field on the user record is empty Solution: Fall back to company or cost_center for organizational context. Check if the user's manager chain can provide department information.
Examples
Example 1: Single Approval Summary
Input: "Summarize approval SYSAPPR0012345"
Output: A structured summary including request details, requester profile, variable values, historical approval rate for the same item, and a recommendation signal based on precedent and policy compliance.
Example 2: Daily Approval Digest
Input: "Generate an approval digest for manager Jane Doe"
Output:
APPROVAL DIGEST - Jane Doe
Date: 2026-03-19
Pending Approvals: 4
Total Value: $7,450.00
1. REQ0045123 / RITM0067890 - New Laptop ($1,800) - John Smith, Engineering - 2 days pending
2. REQ0045130 / RITM0067897 - Software License x5 ($2,500) - Lisa Park, Marketing - 1 day pending
3. REQ0045135 / RITM0067902 - Conference Travel ($2,400) - Mike Johnson, Sales - 4 hours pending
4. REQ0045138 / RITM0067905 - Desk Phone ($750) - Sarah Kim, HR - 1 hour pending
Example 3: High-Value Approval with Risk Flags
Tool: SN-Get-Record
Parameters:
table_name: sysapproval_approver
sys_id: [approval_sys_id]
# Result analysis flags:
# - Cost exceeds $5,000 threshold (requires VP approval)
# - First time this item has been ordered by this department
# - Requester is a contractor (non-FTE)
# - No business justification provided (mandatory field bypass)
Related Skills
catalog/approval-workflows- Configure approval routing rulescatalog/request-fulfillment- Post-approval fulfillment processingcatalog/item-creation- Catalog item setup and configurationcatalog/catalog-item-generation- Generate items from descriptionsadmin/workflow-creation- Build approval workflow logic
Source & license
This open-source skill is cataloged on AgentStack and links to its original source — we do not rehost the code.
- Author: Happy-Technologies-LLC
- Source: Happy-Technologies-LLC/happy-platform-skills
- License: Apache-2.0
Install and usage instructions live in the source repository linked above.
Reviews
No reviews yet, be the first.
Write a review
Versions
- v0.1.0 Imported from the upstream source.