August operations scorecard page
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.
Ruth wants the August scorecard as a web page for Thursday's branch meeting. The numbers are in the tracker export and the targets workbook, and her note explains the rest. Save it as index.html.
Workspace
- note_from_ruth.txt
- ops_tracker_export.csv
- scorecard_targets_2026.xlsx
Expected deliverables
- index.html
This task also has a task-specific grader (check.py).
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.
The frozen scorer's check set differs from this task's source declaration. These are different evaluator versions. The counts below report only checks actually recorded; a missing check is never counted as a pass.
| Recorded check | Type | Required | Failed / observed |
|---|---|---|---|
| index.html exists | file_exists | yes | 0 / 6 |
| every metric listed | text_contains_all | yes | 0 / 6 |
| retired carpet metric left off | custom | yes | 0 / 6 |
| rates shown as percents | text_numbers_present | yes | 0 / 6 |
| page structure | custom | yes | 2 / 6 |
Current task declaration (5 checks)
From task.yaml. Original runner grading and frozen retrospective scoring are separate. Read the frozen source for the rules behind the reported results.
- index.html exists —
file_exists, required - every metric listed —
text_contains_all, required - retired carpet metric left off —
text_not_contains, required - rates shown as percents —
text_numbers_present, required - page structure —
custom, required
Recorded results, desk release
| System | Repetitions | Frozen / 3 | Original / 3 | Median minutes | Mean model estimate (USD) |
|---|---|---|---|---|---|
| Proto / DeepSeek V4.1 Flash | 2 | 0 | 1.5 | 0.0232 | |
| Codex / GPT-5.6 sol | 2 | 1 | 2.5 | 0.2760 |
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": "kpi-scorecard-page".
Recorded results, desk comparison, 28 September 2026
| System | Repetitions | Frozen / 3 | Original / 3 | Median minutes |
|---|---|---|---|---|
| Proto / GPT-6 sol | 3 | 0 | 2.1 | |
| Codex / GPT-6 sol | 3 | 2 | 1.8 | |
| Proto / DeepSeek V4.1 Flash | 2 | 0 | 0.7 |
A separate campaign under the same frozen scorer, not pooled with the release. Conditions and data.