The interface is localized. Skill descriptions and documents are machine-translated by AI into your language; code, commands, paths and technical terms stay in the original, and the author's document governs.
Explore skills
Development · SKILL.md

workhuman-cost-tuning

workhuman-cost-tuning

Govern Workhuman recognition budget, award spend, redemption, licensing, services, and integration operations using current customer evidence. Use when reviewing cost and value. Trigger with "optimize Workhuman cost".

Independent GitHub snapshot · Automated classification · Original documentation governs

Skill documentation

GitHub

name: workhuman-cost-tuning description: 'Govern Workhuman recognition budget, award spend, redemption, licensing, services, and integration operations using current customer evidence. Use when reviewing cost and value. Trigger with "optimize Workhuman cost".' argument-hint: "[program] [review-window]" allowed-tools: Read, Glob, Grep, WebFetch, Write, Edit version: 1.4.0 author: Jeremy Longshore jeremy@intentsolutions.io license: MIT tags: [saas, workhuman, cost, budget, governance] model: inherit effort: high compatibility: Designed for Claude Code; budget, award, eligibility, license, service, and program changes require finance, HR, procurement, and program-owner approval

Workhuman Cost and Recognition-Budget Governance

Overview

Build a dated cost and value model that separates contract charges, recognition awards, redemption effects, services, and customer-operated integration costs.

Prerequisites

  • Current order form, invoices, program budget, award ledger, redemption reporting, and implementation or support statements
  • Eligible population, participation and outcome definitions, finance policy, and program owner
  • Data classification and approved aggregation thresholds for workforce analysis

Tool Discipline

Use Read, Glob, and Grep to inspect approved financial and operational evidence, WebFetch for current public context, and Write or Edit for models, scenarios, and redacted decision records.

Current Contract

Workhuman publicly describes active spend monitoring and control in Admin Hub and a points-based global Store. Public pages do not provide universal customer pricing or prove that one API call, award, point, or redemption has a fixed cost; use customer contract and ledger evidence.

Authentication

Use read-only authorized exports or reports for analysis. Separate financial, program, and technical access and never expose individual compensation, award, balance, or redemption details in broad reports.

Instructions

  1. Freeze the review window, population, currencies, budget basis, contract date, and owners.
  2. Reconcile invoices and commitments with recognition awards, active spend, redemption or liability reporting, services, taxes, and customer infrastructure.
  3. Normalize currencies and time periods while preserving source, effective date, and accounting treatment.
  4. Calculate participation, recognition reach, manager and peer distribution, award mix, redemption experience, operational effort, and approved outcome measures.
  5. Detect overspend, underspend, concentration, inequity, dormant entitlements, duplicate services, failed jobs, and support-heavy workflows without exposing individuals.
  6. Model bounded scenarios with assumptions, confidence, financial impact, cultural or equity risk, and reversibility.
  7. Present policy, budget, eligibility, contract, or architecture changes to finance, HR, procurement, privacy, and program owners.
  8. After approval, canary reversible changes and compare spend, participation, equity, support, and business outcomes.

Approval Boundaries

Do not change budgets, award values, eligibility, redemption policy, licenses, contracts, or workforce communications without accountable owners.

Output

Return the reconciled baseline, dated sources, allocation and value metrics, anomalies, scenarios, assumptions, approvals, canary result, and review date.

Error Handling

ConditionResponse
Invoice and ledger totals disagreePreserve both, stop optimization, and route reconciliation to finance and the program owner.
A scenario harms equity or recognition qualityReject it even if nominal spend falls.
Pricing evidence is staleMark the model provisional and obtain current customer terms.

Example

A redacted completion receipt might look like this:

window=FY2026-Q3; sources=invoice+award-ledger+ops; reconciliation=exact; scenarios=3; recommendation=canary-schedule-change; savings=not-guaranteed

Resources

Next Steps

Review approved changes after a representative cycle and reverse them if participation, equity, quality, or outcomes regress.

Saved skills