Skip to content

Charge a pottery studio's members for a week of kiln firings and flag overloaded kilns

Exact prompt

# Charge a pottery studio's members for a week of kiln firings and flag overloaded kilns Inputs: `kilns.csv` (kiln_id, name, shelf_cm2), `members.csv` (member_id, name), `glazes.csv` (glaze), `firings.csv` (firing_id, kiln_id, date, kind, status), `pieces.csv` (piece_id, member_id, firing_id, footprint_cm2, glaze) and `rates.json` (bisque_per_cm2, glaze_per_cm2, minimum_charge). All members, kilns, glazes and pieces are synthetic; prices are in dollars. These rules are the studio's; where `studio-notes.txt` disagrees, the rules win. Rules: - Firings: a firing with status `fired` happened. A firing with status `aborted` was stopped by a kiln fault; its pieces are refired later at the studio's cost, so they are not charged and count nowhere in this week's figures. - Charges: each piece in a fired firing is charged footprint_cm2 × the rate for the firing's kind (bisque_per_cm2 for `bisque`, glaze_per_cm2 for `glaze`), but never less than minimum_charge: a piece whose rate charge is below minimum_charge is charged minimum_charge instead. Don't round until the output. - Overloads: a fired firing is overloaded when the total footprint_cm2 of its pieces is more than its kiln's shelf_cm2. A firing that fills the shelf exactly is not overloaded. - Glazes: a piece counts under its glaze only when it was in a fired `glaze` firing (bisque pieces have no glaze yet). - Kilns: a kiln's firing count is the number of its fired firings. - Write `out/kiln-ledger.json`: `{"overloaded_firings": [<sorted firing_ids>], "charged_total": <dollars>, "charges_by_member": {"<member_id>": <dollars>, ...}, "glazed_pieces_by_glaze": {"<glaze>": <count>, ...}, "fired_count_by_kiln": {"<kiln_id>": <count>, ...}}`. - Roll-ups: In charges_by_member, members in members.csv with no charged pieces are left out. In glazed_pieces_by_glaze, glazes in glazes.csv with no pieces in a fired glaze firing are left out. In fired_count_by_kiln, kilns in kilns.csv with no fired firings are left out. ## 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.

Work out a synthetic pottery studio's week of kiln firings. Each piece in a completed firing is charged its shelf footprint at the bisque or glaze rate, never less than a stated minimum per piece, while pieces in an aborted firing count nowhere. A firing whose pieces take up more shelf than its kiln has is overloaded. Report the overloaded firings, the charged total, charges per member, glazed pieces per glaze and completed firings per kiln. The prompt says members, glazes and kilns with nothing are left out. A firing that fills its shelf exactly, a piece charged exactly the minimum and a kiln, member and glaze seen only in an aborted firing are decoys. Studio notes get the aborted-firing, overload, minimum and roll-up 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: overloaded_firings charged_total charges_by_member glazed_pieces_by_glaze fired_count_by_kiln

Best model for this task so far: Claude Haiku 5.5 (effort default (no --effort flag) · thinking on) (success, $0.0042, 14s).

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$0.1217s0Fully autonomouspassTrust 80%2026-10-26
AClaude Haiku 5.5default · thinking on$0.004214s0Fully autonomouspassTrust 80%2026-10-26
AClaude Sonnet 5.5default · thinking on$0.0510s0Fully autonomouspassTrust 80%2026-10-26

Embed this result

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

UseCaseRank badge for Charge a pottery studio's members for a week of kiln firings and flag overloaded kilns

Get new runs and ranking changes weekly

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