August menu items ranked by sales and margin
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.
I'm redoing the menu. Using the August POS export and Rosa's cost sheet, show me every item ranked by how many we sold and by margin - my note has how to read the export. Save it as menu_performance.xlsx.
Workspace
- note_from_marcus.txt
- pos_item_details_2026-07.csv
- pos_item_details_2026-08.csv
- recipe_costs.xlsx
Expected deliverables
- menu_performance.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 |
|---|---|---|---|
| menu_performance.xlsx exists | file_exists | yes | 0 / 6 |
| no formula errors | xlsx_no_errors | yes | 0 / 6 |
| Smashed Avocado Toast quantity | xlsx_value_present | yes | 1 / 6 |
| Cold Brew quantity | xlsx_value_present | yes | 1 / 6 |
| Latte margin | xlsx_value_present | yes | 1 / 6 |
| Breakfast Sandwich margin | xlsx_value_present | yes | 1 / 6 |
| total margin | xlsx_value_present | yes | 1 / 6 |
Current task declaration (7 checks)
From task.yaml. Original runner grading and frozen retrospective scoring are separate. Read the frozen source for the rules behind the reported results.
- menu_performance.xlsx exists —
file_exists, required - no formula errors —
xlsx_no_errors, required - Smashed Avocado Toast quantity —
xlsx_value_present, required - Cold Brew quantity —
xlsx_value_present, required - Latte margin —
xlsx_value_present, required - Breakfast Sandwich margin —
xlsx_value_present, required - total margin —
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 | 2 | 2 | 3.6 | 0.0473 | |
| Codex / GPT-5.6 sol | 3 | 3 | 5.5 | 0.6322 |
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": "menu-item-performance".
Recorded results, desk comparison, 28 September 2026
| System | Repetitions | Frozen / 3 | Original / 3 | Median minutes |
|---|---|---|---|---|
| Proto / GPT-6 sol | 1 | 1 | 4.6 | |
| Codex / GPT-6 sol | 3 | 3 | 2.7 | |
| Proto / DeepSeek V4.1 Flash | 3 | 3 | 1.7 |
A separate campaign under the same frozen scorer, not pooled with the release. Conditions and data.