Skip to content
Business BenchA benchmark for AI agents on business work

Task inventory·187 desk·20 build·9 process

Tasks and acceptance contracts

Start with the kind of work an agent must hand back. Each task page exposes the request, inputs, declared checks and available results. Task availability, structural validation and measured agent performance are separate facts.

Desk track

Recorded acceptance counts are shown separately for each configured system, over 3 executions on the same fixture. They are neither difficulty estimates from a representative population nor results on unseen cases. Check counts below describe the current task declaration; detail pages distinguish it from the frozen campaign scorer.

Spreadsheet · 41 tasks · reconciled, calculated, or reshaped workbook

TaskDeliverablesRequired checksProto
accepted / 3
Codex
accepted / 3
Spring no-show rates by tutor under the 24-hour policy
appointment-no-shows
no_show_rates.csv433
Material cost per product from the bill of materials
bom-cost-rollup
assembly_costs.xlsx831
First-half budget vs actual, line by line
budget-vs-actual
variance.xlsx933
Normalize the catalog sizes for the new POS
catalog-normalize-units
catalog_clean.csv633
Fall to spring enrollment, adds and drops by course
class-roster-changes
roster_changes.csv633
Q3 hours per client project, billable and not
client-hours-by-project
client_hours.csv733
Weekly hours audit from time-clock punches
clock-punch-audit
hours_audit.csv433
August sales commissions with refund recoveries
commission-calculation
commissions.xlsx732
Lifetime value per customer from the shop export
customer-value-summary
customer_value.csv633
Per-donor giving totals for 2024 and 2025
donor-totals-by-year
donor_totals.csv533
Find bills that were entered twice before the check run
duplicate-invoice-finder
possible_duplicates.csv733
Door list for the author evening
event-attendee-merge
attendees.csv723
Monthly miles per van for the insurer
fleet-mileage-monthly
mileage.csv433
Final weighted grades for the bookkeeping course
gradebook-weighted
grades.xlsx432
Reconcile the physical count against the system stock
inventory-count-reconcile
variances.csv513
Stock valued at cost by department for the accountant
inventory-valuation
valuation.xlsx833
Q2 leads and won revenue by marketing source
lead-source-rollup
lead_sources.csv533
October renewal letters and the lapsed list
membership-renewals
renewals.csv533
Cost out the fall menu, plate by plate
menu-cost-sheet
menu_costs.xlsx633
First-half sales in dollars across three currencies
multi-currency-sales
sales_usd.xlsx733
What is still coming on our purchase orders
open-purchase-orders
open_po_lines.csv533
Apply the supplier's price letter to our master price list
price-list-update
price_list_2026.xlsx633
How the August promo codes did
promo-code-analysis
promo_summary.csv623
October rent roll for the bank
rent-roll-build
rent_roll.xlsx532
Days out per machine per month for the insurer
rental-utilization
utilization.csv533
This week's reorder list for the bike shop
reorder-list
reorder.csv433
Rep by month sales table from the invoice export
sales-pivot-by-rep
sales_by_rep.xlsx731
Sales tax on each September order
sales-tax-by-address
orders_tax.csv533
How short the published week is against minimum coverage
shift-coverage-gaps
coverage_gaps.csv433
Work out what the carrier should charge per order
shipping-rate-lookup
orders_shipping.csv623
Move the shop inventory onto the new SKUs
sku-crosswalk-migration
catalog_new.csv633
One staff directory from HR and IT
staff-directory-merge
directory.csv633
September billable utilization per person
staff-utilization
utilization.xlsx902
Active subscribers and cancellations by month
subscription-status-monthly
subscriptions.xlsx530
Metric supplier price list into our inch-and-pound catalog
supplier-catalog-imperial
catalog_us.csv733
Member survey scores per question
survey-tally
survey_summary.csv533
Find double-booked tables in the reservations
table-double-bookings
conflicts.csv533
Consolidate the crew timesheets for payroll
timesheet-consolidate
hours.csv533
Merge the purchasing and accounting vendor lists
vendor-contact-merge
vendors.csv633
Year-to-date spend by vendor and category
vendor-spend-categories
vendor_spend.xlsx830
Equipment warranties running out in the next 90 days
warranty-expiry-list
expiring_warranties.csv433

Bookkeeping · 36 tasks · reconciliation, schedule, transaction classifications

TaskDeliverablesRequired checksProto
accepted / 3
Codex
accepted / 3
Payables aging by vendor for the line of credit review
ap-aging
ap_aging.xlsx, memo.md1211
August bank reconciliation for the checking account
bank-reconciliation
reconciliation.xlsx623
Code August company-card charges to the ledger by item type
card-expense-coding
coded.csv, review.csv733
August cash drawer variance by day
cash-drawer-variance
drawer_variance.csv633
Chargeback tracker with evidence deadlines
chargeback-tracker
chargebacks.csv731
Commission clawbacks on refunded deals since June
commission-clawbacks
clawbacks.csv633
Turn September returns into credit notes and apply them
credit-notes-apply
credit_notes.csv, invoice_balances.csv633
2026 deferred revenue schedule for the annual plans
deferred-revenue-schedule
deferred_revenue.xlsx931
2026 depreciation schedule from the fixed asset register
depreciation-schedule
depreciation.xlsx933
Find supplier payments we made twice
duplicate-payments
duplicate_payments.csv633
Categorize card transactions per policy, with thread overrides
expense-categorize
categorized.csv332
Audit August expense reports against the travel policy
expense-report-audit
violations.csv533
First-half FIFO cost of goods sold
fifo-cogs
cogs.xlsx922
Realized FX gain or loss on the euro invoices
fx-invoice-gains
fx_results.csv531
Gift card liability per card at year end
gift-card-liability
gift_cards.csv633
Grant budget tracking for the board packet
grant-budget-tracking
grant_tracking.xlsx833
Shrinkage by department from the Sunday floor count
inventory-shrinkage
shrinkage.xlsx933
Check the gym loan's interest and principal split
loan-schedule-check
loan_check.xlsx633
Apply today's lockbox checks to open invoices
lockbox-apply
applied.csv, unapplied.csv633
August mileage reimbursement for the field staff
mileage-reimbursement
reimbursement.csv533
August accrual entry for the theatre
month-end-accruals
accruals.csv533
Match bank payments to open invoices
payments-match
unpaid.csv433
Match bank payments to open invoices, with card fees, short wires, and a wrong reference
payments-match-v2
unapplied.csv, unpaid.csv633
Journal entry for the 4 September payroll
payroll-journal-entry
payroll_je.csv533
Federal payroll tax deposits and due dates since the provider stopped
payroll-tax-deposits
deposits.csv533
Reconcile the front desk petty cash and code the top-up
petty-cash-reconcile
petty_cash.xlsx833
Catch-up prorations for February to April plan changes
plan-change-proration
prorations.csv533
2026 prepaid expense amortization schedule
prepaid-amortization
prepaids.xlsx833
West Gallery budget burn for the capital committee
project-budget-burn
burn.xlsx933
Reconcile August refunds to the card processor
refund-reconciliation
refund_recon.csv623
Second-quarter sales tax due by local jurisdiction
sales-tax-liability
tax_liability.csv633
Allocate August overhead to departments
shared-cost-allocation
allocation.xlsx933
August resident balances with late fees
tenant-ledger-balances
tenant_balances.csv533
Split last week's tip pool by shift
tip-pooling
tip_distribution.csv423
2025 reportable totals per vendor, cards and corporations out
vendor-1099-totals
vendor_totals.csv503
Tie out the Keystone supply statement to our books
vendor-statement-reconcile
vendor_recon.csv633

Reports · 30 tasks · data summary and source-grounded memo

TaskDeliverablesRequired checksProto
accepted / 3
Codex
accepted / 3
Receivables aging at month end for the bank
ar-aging-report
aging.xlsx, memo.md931
Cost per lead by channel for the quarter
campaign-cpl-report
cpl.xlsx933
Money in and out by month for the bank meeting
cash-flow-monthly
cashflow.xlsx, memo.md1022
Visits per provider and the no-show rate for Q1
clinic-visits-summary
visits.xlsx830
First-half complaints by category and month
complaints-summary
complaints.xlsx, memo.md833
On-time delivery by hub, and why the late ones were late
delivery-performance
delivery.xlsx831
Q2 expenses by department against budget
dept-expense-report
dept_expenses.xlsx, memo.md1022
FY2026 fundraising figures for the board
donor-annual-figures
annual_figures.xlsx, memo.md1011
Monthly electricity by building from meter reads
energy-usage-sites
energy.xlsx, memo.md1022
Wrap-up numbers for the summit
event-postmortem
event_summary.xlsx, memo.md1033
Cost per mile per vehicle for the first half
fleet-cost-per-mile
fleet_costs.xlsx933
May helpdesk tickets by category and SLA breaches
helpdesk-tickets-report
memo.md, tickets_report.xlsx830
Half-year safety incidents by site and type
incident-summary
incidents.xlsx, memo.md932
August menu items ranked by sales and margin
menu-item-performance
menu_performance.xlsx723
Monthly revenue report by region from a messy sales export
monthly-report
memo.md, report.xlsx1132
Monthly revenue report by region from a messy sales export
monthly-report-v9
memo.md, report.xlsx1131
MRR by month and churn for the first half
mrr-report
memo.md, mrr.xlsx911
Q1 occupancy per cabin per month
occupancy-report
occupancy.xlsx833
2025 payroll by quarter for the accountant
payroll-quarterly-summary
payroll_q.xlsx933
First-half purchasing savings by category
procurement-savings
savings.xlsx833
August yield and scrap by production line
production-yield
yield.xlsx930
Margin by job, with the loss-makers flagged
project-margin
margins.xlsx, memo.md1033
Q3 wholesale sales by product, with the odd things called out
quarterly-sales-report
memo.md, q_report.xlsx1032
Q2 sales by territory from the monthly exports
regional-sales-monthly
memo.md, regional.xlsx932
Customer cohort retention table for the first half
retention-cohorts
cohorts.xlsx831
Q2 return rate by product and reason
returns-analysis
returns.xlsx1022
Spring tuition collected and outstanding by program
tuition-collections
collections.xlsx831
Volunteer hours per program for the county
volunteer-hours
volunteer_hours.csv733
Spring website traffic by month, crawler taken out
web-traffic-summary
memo.md, traffic.xlsx833
One-page weekly KPIs from five system exports
weekly-kpi-dashboard
kpi.xlsx830

Reformatting · 26 tasks · destination-compatible import file

TaskDeliverablesRequired checksProto
accepted / 3
Codex
accepted / 3
Clean the catalog mailing list for the mail house
address-standardize
addresses_clean.csv533
Build the Amazon inventory upload from the product sheet
amazon-inventory-loader
amazon_inventory.csv833
Restate this year's GL lines on the new chart of accounts
chart-of-accounts-remap
remapped.csv633
2025 1099-NEC recipients into the e-file service's import file
contractor-1099-list
1099_vendors.csv633
Clean a legacy customer export for HubSpot import
customer-dedupe
hubspot_import.csv533
Load the concert sign-ups into Eventbrite
eventbrite-attendee-import
eventbrite_attendees.csv633
Move the office phone list into Google Contacts
google-contacts-import
google_contacts.csv733
Build the Gusto employee upload from the HR roster
gusto-employee-import
gusto_employees.csv833
Import the roofing pipeline into HubSpot deals
hubspot-deals-import
hubspot_deals.csv733
Fall attendance grid for the district grant
long-to-wide-attendance
attendance_grid.csv533
Summer signups into one Mailchimp audience import
mailchimp-audience-import
mailchimp_import.csv633
Split the tracker notes into Pipedrive activities
notes-to-activities
activities.csv513
Convert the CRM contact phones to the new phone system's import file
phones-to-e164
contacts_e164.csv733
Load the fall menu into the new POS with modifiers
pos-menu-import
pos_menu.csv633
CRM accounts into a QuickBooks Online customer import
quickbooks-customer-import
qbo_customers.csv933
Turn the Q3 leads sheet into a Salesforce lead import
salesforce-lead-import
sf_leads.csv833
Fall class schedule as a calendar file for the website
schedule-to-ics
schedule.ics533
Fall catalog into Shopify's product import
shopify-product-import
shopify_products.csv1023
Load the stock count into Square's item import
square-inventory-import
square_items.csv733
Build the Stripe customer import from the old billing export
stripe-customer-import
stripe_customers.csv733
Flatten the nested fall catalog into the inventory platform's variant import
variants-flatten
variants_flat.csv733
Reshape the by-month sales report for the planning tool upload
wide-to-long-sales
sales_long.csv533
Turn August web store orders into journal entries
woocommerce-orders-to-journal
journal.csv533
August job invoices into Xero's sales invoice import
xero-sales-invoices-import
xero_invoices.csv623
September checking transactions into YNAB
ynab-transactions-import
ynab.csv633
Import client contacts as Zendesk users
zendesk-users-import
zendesk_users.csv633

Extraction · 26 tasks · structured records extracted from documents

TaskDeliverablesRequired checksProto
accepted / 3
Codex
accepted / 3
Key the July and August bank statements into the transactions file
bank-statement-pdf
transactions.csv733
CRM contacts from the scanned expo business cards
business-cards-scanned
contacts.csv823
August transactions from the company card statements
card-statements-pdf
card_transactions.csv733
Side-by-side table of the anniversary dinner catering quotes
catering-quotes-table
quote_comparison.csv933
Pull key terms from our service contracts into a register
contract-key-terms
contracts.csv833
Applicant list from the operations coordinator resumes
cv-batch-extraction
candidates.csv733
Log this year's grant award letters into the tracker
grant-award-letters
grants.csv833
Open findings from this season's inspection reports
inspection-findings
findings.csv733
Log vendor insurance certificates and flag lapsed policies
insurance-certificates
coi.csv833
Abstract the plaza leases for the lender
lease-abstracts
lease_abstracts.csv1023
Match this week's packing slips to the open POs
packing-slips-vs-po
receipts.csv733
Key the August pay stubs for the CPA
pay-stubs-extraction
pay_stubs.csv933
Match the scanned field trip slips to the class roster
permission-forms-scanned
consents.csv913
Log this year's property tax bills
property-tax-bills
tax_bills.csv833
Key this week's customer POs into order entry
purchase-orders-pdf
po_lines.csv833
Compare three laptop quotes (USD/EUR, hidden fees) and recommend one
quote-comparison
comparison.xlsx, recommendation.md733
Expense table from scanned trip receipts
receipts-scanned
expenses.csv733
Shipment weights and costs from last week's manifests
shipment-manifests
shipments.csv733
Key this week's superbills for charge entry
superbill-codes
superbill_lines.csv833
Key August supplier bills into the AP import file
supplier-invoices-to-csv
invoice_lines.csv920
Third-quarter ledger from the landlord statements
tenant-statements
tenant_ledger.csv700
Daily hours from the scanned caregiver time sheets
timesheets-scanned
timesheet_hours.csv633
Monthly usage and charges from the utility bills
utility-bills-extraction
utility_usage.csv833
Load the fall supplier price sheets for the ordering guide
vendor-price-sheets
prices.csv831
Log new vendors' tax forms with masked TINs
vendor-tax-forms
vendor_tax.csv1032
Key the September venue invoices into job costing
venue-invoices
venue_lines.csv1033

Drafting · 19 tasks · business text preserving required facts and rules

TaskDeliverablesRequired checksProto
accepted / 3
Codex
accepted / 3
Renewal summary for the Wi-Fi services contract
contract-renewal-summary
renewal_summary.md523
Announcement for the wine club harvest dinner
event-announcement
announcement.md633
Website FAQ from the summer support tickets
faq-from-tickets
faq.md833
Fall appeal letter for the Hatch Farm campaign
fundraising-appeal
appeal.md433
Formal incident report for the Line 3 injury
incident-report
incident_report.md723
Q2 investor update letter from the KPI tracker
investor-update
update.md730
Write the front desk coordinator job posting
job-posting-from-notes
posting.md833
Water shutoff notices for the riser valve work
maintenance-notices
notices/*[Aa][Ss][Pp][Ee][Nn]*, notices/*[Bb][Ii][Rr][Cc][Hh]*, notices/*[Cc][Yy][Pp][Rr][Ee][Ss][Ss]*, notices/*[Dd][Oo][Gg][Ww][Oo][Oo][Dd]*533
Minutes from the leadership meeting recording
minutes-from-transcript
minutes.md430
Welcome email for the new project engineer
onboarding-welcome-email
welcome.md523
Write this month's past-due reminders
overdue-reminders
reminders/*.md, reminders/*[Dd][Oo][Rr][Ss][Ee][Yy]*, reminders/*[Rr][Ee][Dd][Ww][Oo][Oo][Dd]*, reminders/*[Yy][Ee][Ll][Ll][Oo][Ww][Tt][Aa][Ii][Ll]*530
Memo to the crews on the 2026 policy changes
policy-update-memo
memo.md731
Draft the standing order price increase notice
price-increase-notice
notice.md920
Write up the warehouse lighting proposal
proposal-from-notes
proposal.md823
Reply to a customer about a cracked skillet
refund-apology-letter
response.md511
Request for quotation for the rolling bench parts
rfq-email
rfq.md433
Schedule change notices for next week's revised rota
schedule-change-notices
notices/*[eE][dD][wW][aA][rR][dD]*[wW][aA][rR][dD]*, notices/*[jJ][oO][sS][hH][uU][aA]*[sS][cC][oO][tT][tT]*, notices/*[kK][aA][rR][eE][nN]*[aA][lL][lL][eE][nN]*, notices/*[rR][yY][aA][nN]*[yY][oO][uU][nN][gG]*, notices/*[sS][tT][eE][pP][hH][aA][nN][iI][eE]*[rR][aA][mM][oO][sS]*633
Turn the refunds email thread into a counter procedure
sop-from-thread
procedure.md622
Dispute letter for a supplier invoice that does not match the PO
supplier-dispute-letter
dispute.md630

Tooling · 9 tasks · small file-based tool or static page

TaskDeliverablesRequired checksProto
accepted / 3
Codex
accepted / 3
Conference schedule page for the lobby screen
event-schedule-page
index.html423
Boat campaign progress page with top donors
fundraiser-progress-page
index.html523
Stock page for the workshop tablet
inventory-tracker-page
index.html533
August operations scorecard page
kpi-scorecard-page
index.html522
Fall menu page with allergens for the QR code
menu-page-allergens
index.html533
This week's orders page for the packing bench
orders-status-page
index.html523
Fall wholesale price list page for shop buyers
price-list-page
index.html513
Kitchen job task board for the crew wall
project-status-board
index.html523
Staff directory page for the backstage screen
team-directory-page
index.html423

Build track

Each task specifies an application, seed records and three changes. A tester must verify the running application against the enterprise baseline and task checklist. No completed acceptance scores are published. Checklist counts measure specification size.

ApplicationSeed filesChecklist itemsChange items
IT asset register with padded tags, assignment history, warranty alerts, disposal rules, and site-scoped access
asset-tracking
assets.csv, assignments.csv269
Physiotherapy clinic scheduling app with double-booking rules, no-show tracking, recall list, and utilization
clinic-scheduling
patients.csv, roster.csv, appointments.csv269
B2B sales CRM with accounts, contacts, a staged deal pipeline, weighted forecast, and activity log
crm-sales-pipeline
accounts.csv, contacts.csv, deals.csv269
Donor management app with household dedupe, pledges and balances, acknowledgement letters, and program-scoped access
donor-management
donors.csv, gifts.csv, pledges.csv269
Conference workshop registration with room capacity, waitlist promotion, badges, and chair-scoped access
event-registration
rooms.csv, sessions.csv, registrations.csv269
Expense claims app with receipt and policy-limit rules, manager approval, and reimbursement batches
expense-claims
employees.csv, expense_policy.csv, claims.csv269
Managed-service helpdesk with client contacts, first-reply promises, escalation, and agent and client scoping
helpdesk-tickets
clients.csv, contacts.csv, tickets.csv269
HVAC field-service app with customers, jobs, technicians, parts, and invoicing
hvac-field-service
customers.csv, parts.csv319
Multi-warehouse inventory app with transfers, reasoned adjustments, reorder alerts, and stock valuation
inventory-multi-location
items.csv, stock.csv269
Leave tracking with team approvals, working-day balances, holidays, and overlap warnings
leave-tracking
staff.csv, leave_requests.csv, holidays.csv269
Fitness club memberships with renewals, freezes, door check-ins, and trainer-scoped access
memberships-club
members.csv, checkins.csv269
Retail and wholesale order management with tax and shipping rules, stock on ship, and backorders
order-management
customers.csv, products.csv, order_lines.csv269
Back office for two cafes with menu items, modifiers, nightly cash-ups, item performance, price-change audit, and a dashboard
point-of-sale-backoffice
items.csv, modifiers.csv, daily_takings.csv, item_sales.csv269
Maintenance app for six apartment buildings with work requests, work orders, vendors, approvals, and a dashboard
property-maintenance
units.csv, tenants.csv, vendors.csv, work_requests.csv269
School district purchase requests with amount-based two-step approval, budget lines, and PO numbers
purchase-requests-approvals
budgets.csv, vendors.csv, requests.csv269
Quoting app for a custom furniture maker with margin floor, discount approval, and quote-to-order conversion
quoting-tool
products.csv, materials.csv, quotes.csv269
Hiring app for a growing company with roles, candidates, stages, interviews, offer approvals, and a dashboard
recruiting-pipeline
roles.csv, candidates.csv269
Consultancy time tracking with rate cards, rounding, monthly invoicing, and locked entries
time-tracking-invoicing
clients.csv, rates.csv, time_entries.csv269
Vendor onboarding app with document checklists, approval stages, and insurance alerts
vendor-onboarding
vendors.csv, documents.csv269
Work orders against bills of materials for a small manufacturer, with availability checks, stock moves, WIP, and a dashboard
work-orders-bom
parts.csv, boms.csv, work_orders.csv269

Process track

A role in an ERP over several business dates, evaluated on final state, audit rules and requested notes. The published pilot covers 6 of these 9 tasks, at one seed with 5 executions per system.

TaskRole and familyBandTurns / checksPublished evidence
Clearing three weeks of vendor invoices as AP supervisor
ap-invoice-backlog
AP supervisor
Receiving, matching and paying
analyst3 / 17No published attempts
Book September's freight accruals, then revise them when the facts behind them change
freight-accrual-revision
Staff accountant
Record to report
clerical2 / 9No published attempts
Explain why gross margin fell from August to September
margin-bridge
Financial analyst
Finance questions
clerical1 / 10Provisional pilot
Close September from the checklist and the documents in the accounting inbox
month-end-close
Staff accountant
Record to report
clerical2 / 7Provisional pilot
The weekly MRP cycle, with a lead-time change from a vendor
mrp-planner-week
Production planner
Plan to produce
clerical1 / 7Provisional pilot
Verify bank-detail changes and run the weekly payments inside the cash floor
payment-run
AP supervisor
Receiving, matching and paying
clerical2 / 10Provisional pilot
Verify bank-detail changes, run the weekly payments inside the cash floor, and keep bank details need-to-know
payment-run-need-to-know
AP supervisor
Receiving, matching and paying
clerical2 / 11No published attempts
A week of purchase orders, receiving and invoice matching
procure-to-pay-week
Buyer and AP clerk
Receiving, matching and paying
clerical3 / 18Provisional pilot
Clear the requisition approval queue under the delegation of authority
requisition-approval-queue
Operations manager
Requisitions and purchasing
clerical1 / 4Provisional pilot

Authoring and exposure

The repository includes generators, fixtures, checks and reference policies. Rerolling names and amounts tests new instances of known templates; it does not make exposed task structure unseen or eliminate development contamination. Generalization needs separately held-out templates and a stated sampling frame. See the exposure analysis.

To contribute or dispute a contract, start with the task format, desk authoring guide, build authoring guide, or process specification. Explain which recipient requirement a check represents and include a valid output and a plausible invalid counterexample.