Log this year's grant award letters into the tracker
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.
Go through the funder letters in the folder and put the key details of every grant into grants.csv, one row per grant. Dana's note has the columns I need.
Workspace
- funder_letters/Alder_Family_Foundation_award.pdf
- funder_letters/Granite_Peak_Foundation_response.pdf
- funder_letters/HarborTrust_grant_letter.pdf
- funder_letters/Meridian_Health_Fdn_notice_of_award.pdf
- funder_letters/Sycamore-Fund-2026-award.pdf
- funder_letters/WillowCreek_amendment_letter.pdf
- funder_letters/WillowCreek_original_award_2026-05.pdf
- funder_letters/scan_northgate_letter.pdf
- note_from_dana.txt
Expected deliverables
- grants.csv
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 |
|---|---|---|---|
| requested columns | csv_columns | yes | 0 / 6 |
| one row per awarded grant | csv_set_equal | yes | 0 / 6 |
| row count | csv_row_count | yes | 0 / 6 |
| funder names | csv_values_match | yes | 0 / 6 |
| award amounts | csv_values_match | yes | 0 / 6 |
| grant period | csv_values_match | yes | 0 / 6 |
| report due dates | csv_values_match | yes | 0 / 6 |
| declined request absent | text_not_contains | yes | 0 / 6 |
Current task declaration (8 checks)
From task.yaml. Original runner grading and frozen retrospective scoring are separate. Read the frozen source for the rules behind the reported results.
- requested columns —
csv_columns, required - one row per awarded grant —
csv_set_equal, required - row count —
csv_row_count, required - funder names —
csv_values_match, required - award amounts —
csv_values_match, required - grant period —
csv_values_match, required - report due dates —
csv_values_match, required - declined request absent —
text_not_contains, required
Recorded results, desk release
| System | Repetitions | Frozen / 3 | Original / 3 | Median minutes | Mean model estimate (USD) |
|---|---|---|---|---|---|
| Proto / DeepSeek V4.1 Flash | 3 | 1 | 1.1 | 0.0111 | |
| Codex / GPT-5.6 sol | 3 | 0 | 1.4 | 0.1911 |
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": "grant-award-letters".
Recorded results, desk comparison, 28 September 2026
| System | Repetitions | Frozen / 3 | Original / 3 | Median minutes |
|---|---|---|---|---|
| Proto / GPT-6 sol | 3 | 2 | 1.6 | |
| Codex / GPT-6 sol | 3 | 2 | 1.2 | |
| Proto / DeepSeek V4.1 Flash | 3 | 2 | 0.7 |
A separate campaign under the same frozen scorer, not pooled with the release. Conditions and data.