Cost per mile per vehicle for the first half
The request
The task request supplied to the adapter, alongside the workspace below. Runtime instructions, authorized tools and skills also affect the system. Containerized runs withhold host-side references; local mode is not an isolation boundary.
Can you work out what each of our vehicles cost per mile to run from January to June? Save it as fleet_costs.xlsx with live formulas. Carla's note says where everything comes from.
Workspace
- fleet_list_and_odometers.xlsx
- fuel_card_transactions_2026H1.csv
- note_from_carla.txt
- shop_invoices_2026H1.csv
Expected deliverables
- fleet_costs.xlsx
Recorded acceptance checks
The table uses check names and outcomes from the frozen campaign ledger. Acceptance requires all required predicates to pass; a predicate may itself contain a tolerance or coverage threshold. Its scope is part of the measurement contract.
| Recorded check | Type | Required | Failed / observed |
|---|---|---|---|
| fleet_costs.xlsx exists | file_exists | yes | 0 / 6 |
| live formulas | xlsx_has_formulas | yes | 0 / 6 |
| no error cells | xlsx_no_errors | yes | 0 / 6 |
| Van 26 gallons (litres converted) | xlsx_value_present | yes | 0 / 6 |
| Van 24 miles (to the bill of sale) | xlsx_value_present | yes | 0 / 6 |
| Van 24 cost per mile | xlsx_value_present | yes | 0 / 6 |
| Van 22 total cost (plate-named invoices included) | xlsx_value_present | yes | 0 / 6 |
| fleet total cost (no car washes or snacks) | xlsx_value_present | yes | 0 / 6 |
| fleet miles | xlsx_value_present | yes | 0 / 6 |
Current task declaration (9 checks)
From task.yaml. Original runner grading and frozen retrospective scoring are separate. Read the frozen source for the rules behind the reported results.
- fleet_costs.xlsx exists —
file_exists, required - live formulas —
xlsx_has_formulas, required - no error cells —
xlsx_no_errors, required - Van 26 gallons (litres converted) —
xlsx_value_present, required - Van 24 miles (to the bill of sale) —
xlsx_value_present, required - Van 24 cost per mile —
xlsx_value_present, required - Van 22 total cost (plate-named invoices included) —
xlsx_value_present, required - fleet total cost (no car washes or snacks) —
xlsx_value_present, required - fleet miles —
xlsx_value_present, required
Recorded results, desk release
| System | Repetitions | Frozen / 3 | Original / 3 | Median minutes | Mean model estimate (USD) |
|---|---|---|---|---|---|
| Proto / DeepSeek V4.1 Flash | 3 | 3 | 3.8 | 0.0456 | |
| Codex / GPT-5.6 sol | 3 | 3 | 6.1 | 0.5247 |
These are repeated executions on the released fixture, not independent new business cases. Cost is captured model usage priced under the recorded assumptions, excluding infrastructure and human review. Per-attempt check verdicts, raw and frozen verdicts, usage, receipt and artifact hashes for this task are in results/latest/attempts.jsonl under "task": "fleet-cost-per-mile".
Recorded results, desk comparison, 28 September 2026
| System | Repetitions | Frozen / 3 | Original / 3 | Median minutes |
|---|---|---|---|---|
| Proto / GPT-6 sol | 3 | 3 | 2.5 | |
| Codex / GPT-6 sol | 3 | 3 | 3.9 | |
| Proto / DeepSeek V4.1 Flash | 3 | 3 | 2.3 |
A separate campaign under the same frozen scorer, not pooled with the release. Conditions and data.