Skip to content

Invoice an auto repair shop's repair orders with labour rates, a parts markup matrix, core charges, warranty claims, shop supplies and tax

Exact prompt

# Invoice an auto repair shop's repair orders with labour rates, a parts markup matrix, core charges, warranty claims, shop supplies and tax Inputs: `ro-lines.csv` (line_id, ro_id, kind, job_category, book_hours, part_no, qty, unit_cost, core_charge, core_returned, warranty) and `shop-rates.json` (labour_rates per job_category, warranty_rate, markup_matrix, supplies_pct, supplies_cap, tax_pct). All customers and vehicles are synthetic; amounts are in dollars. These rules are the shop's policy; where `front-desk-note.txt` disagrees, the rules win. Rules: - Labour: a labour line bills book_hours × the labour rate of its job_category, rounded half-up to the cent. - Parts: a part's unit price is unit_cost plus the pct of the first markup_matrix tier whose max_cost is at least unit_cost (max_cost null is the last tier), rounded half-up to the cent; the part line bills unit price × qty. - Cores: a part line with core_charge above 0 also bills core_charge × qty, unless core_returned is yes. - Warranty: a line with warranty yes bills nothing to the customer (no labour, part or core charge). It goes on the manufacturer claim instead: labour at book_hours × warranty_rate, parts at unit_cost × qty, each rounded half-up to the cent. - Shop supplies: supplies_pct percent of the repair order's customer labour, rounded half-up to the cent, but never more than supplies_cap. - Tax: tax_pct percent of the repair order's customer parts + cores + shop supplies, rounded half-up to the cent. Labour is not taxed. - Write `out/invoices.json`: `{"orders": {"<ro_id>": {"labour": <dollars>, "parts": <dollars>, "cores": <dollars>, "supplies": <dollars>, "tax": <dollars>, "total": <labour + parts + cores + supplies + tax>, "warranty_claim": <dollars>}, ...}, "warranty_claim_total": <dollars>}`. ## How this run works - Work on your own: nobody will answer questions in this task, and writing to `ask/` counts against the run. - Write only under `out/`. Create `out/` if it is missing. Inputs are synthetic; don't fetch anything from the network. - Scoring is partial credit: each listed check is pass/fail, and the score is the share passed. Numbers are compared to within 0.005 unless the task says otherwise.

Invoice a synthetic repair shop's repair orders. Labour is billed by book hours at a rate per job category, parts are marked up by a cost-tier matrix with costs sitting exactly on the tier limits, unreturned cores are charged, warranty lines move to a manufacturer claim at different rates, shop supplies are a capped percentage of labour, and tax applies to parts, cores and supplies but not labour. A front-desk note gets the tier, tax, cap and warranty rules wrong. Original benchmark task written for usecaserank.com (not based on a specific public post). Headless, offline run on a reproducible fixture. Each run's notes say how its cost was measured (estimates are marked).

Success criteria: labour parts_markup cores warranty supplies_and_tax invoice_totals

Best model for this task so far: Claude Haiku 5.5 (effort default (no --effort flag) · thinking on) (success, $0.0053, 24s).

Run this on another model

All runs (3)

100% pass rate across 3 eligible runs: 3 pass, 0 partial, 0 fail. Partial and failed runs are listed below with everything else.ⓘ

Runs for this use case
ModelCostTimeStep-insⓘAutonomyOutcomeTrustⓘDateCompare
AClaude Opus 5.5default · thinking on$0.1324s0Fully autonomouspassTrust 80%2026-10-25
AClaude Haiku 5.5default · thinking on$0.005324s0Fully autonomouspassTrust 80%2026-10-25
AClaude Sonnet 5.5default · thinking on$0.0614s0Fully autonomouspassTrust 80%2026-10-24

Embed this result

A live badge for this use case. It links back here and updates as new runs are published.

UseCaseRank badge for Invoice an auto repair shop's repair orders with labour rates, a parts markup matrix, core charges, warranty claims, shop supplies and tax

Get new runs and ranking changes weekly

Which models got better or cheaper on real tasks like this one.