Write the front desk coordinator job posting
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 write the job posting for the Northside front desk coordinator from Joshua's kickoff notes and the email thread? Save it as posting.md so I can put it up this week.
Workspace
- email_thread_front_desk_posting.txt
- hiring_kickoff_notes.txt
- posting_2024_front_desk.md
Expected deliverables
- posting.md
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 |
|---|---|---|---|
| pay range | text_sentence_matches | yes | 0 / 6 |
| schedule days | text_sentence_matches | yes | 0 / 6 |
| schedule hours | custom | yes | 0 / 6 |
| clinic address | text_sentence_matches | yes | 0 / 6 |
| old addresses left out | text_not_contains | yes | 0 / 6 |
| language requirement withdrawn | text_not_contains | yes | 0 / 6 |
| CPR timing | text_sentence_matches | yes | 0 / 6 |
| how to apply | text_matches_all | 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.
- pay range —
text_sentence_matches, required - schedule days —
text_sentence_matches, required - schedule hours —
custom, required - clinic address —
text_sentence_matches, required - old addresses left out —
text_not_contains, required - language requirement withdrawn —
text_not_contains, required - CPR timing —
text_sentence_matches, required - how to apply —
text_sentence_matches, required
Recorded results, desk release
| System | Repetitions | Frozen / 3 | Original / 3 | Median minutes | Mean model estimate (USD) |
|---|---|---|---|---|---|
| Proto / DeepSeek V4.1 Flash | 3 | 0 | 0.4 | 0.0059 | |
| Codex / GPT-5.6 sol | 3 | 2 | 0.7 | 0.1850 |
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": "job-posting-from-notes".
Recorded results, desk comparison, 28 September 2026
| System | Repetitions | Frozen / 3 | Original / 3 | Median minutes |
|---|---|---|---|---|
| Proto / GPT-6 sol | 3 | 3 | 0.6 | |
| Codex / GPT-6 sol | 3 | 3 | 0.5 | |
| Proto / DeepSeek V4.1 Flash | 3 | 0 | 0.3 |
A separate campaign under the same frozen scorer, not pooled with the release. Conditions and data.