Tasks · desk · Spreadsheet
Find bills that were entered twice before the check run
duplicate-invoice-finder · timeout 1200 s · source
The ask
Sent verbatim as the prompt. The agent works in a copy of the workspace below and sees nothing else.
Go through this quarter's AP register and list the bills that look like they were entered twice. Marcus's note has how we decide. Save the list as possible_duplicates.csv.
Workspace
- ap_bills_register_2026-04-01_to_2026-06-30.csv
- ap_bills_register_2026-07-01_to_2026-09-10.csv
- note_from_marcus.txt
- vendor_list.xlsx
Expected deliverables
- possible_duplicates.csv
Checks
The task passes only when every required check passes. Reference answers, tolerances, and the planted traps are in the repository, not here.
| Check | Type | Required | Failed (frozen) / 6 |
|---|---|---|---|
| requested columns | csv_columns | yes | 0 |
| flagged bills | csv_set_equal | yes | 0 |
| row count | csv_row_count | yes | 0 |
| paired with the original | csv_values_match | yes | 0 |
| amount of the duplicate | csv_values_match | yes | 0 |
| days apart | csv_values_match | yes | 0 |
| credits and recurring bills not flagged | text_not_contains | yes | 0 |
Recorded results, complete desk comparison
pass pass, frozen scorer only fail timed out
| System | Repetitions | Frozen / 3 | Raw / 3 | Median minutes | Mean cost (USD) |
|---|---|---|---|---|---|
| Proto / DeepSeek V4.1 Flash | 3 | 3 | 1.7 | 0.0292 | |
| Codex / GPT-5.6 sol | 3 | 3 | 2.4 | 0.4005 |
Per-attempt check verdicts, raw and frozen verdicts, usage, receipt and artifact hashes for this task are in results/latest/attempts.jsonl under "task": "duplicate-invoice-finder".