Client compute note / synthetic
Under the strict contract, B clears the configured sample gate. A misses the quality floor; its lower bill does not rescue the result.
Authored record state: draft. Pinned inputs: compute-result@1.
SECOND RUN / RESEARCH DELIVERY
BEP preparation / independent Second Run edition
SYNTHETIC FROZEN EXAMPLE / 2026-09-23T05:02:57.802Z
Synthetic example reviewer · 2026-09-23T05:02:57.798Z
Reviewed the strict-JSON synthetic case only; no real hardware or customer claim.
Review is bound to this exact target and dependency set. The name is operator-supplied, not identity-authenticated.
Under the strict contract, B clears the configured sample gate. A misses the quality floor; its lower bill does not rescue the result.
Authored record state: draft. Pinned inputs: compute-result@1.
Configuration B has the lowest cost among configurations clearing the chosen floor in this synthetic sample.
| Configuration | Accepted / original | Complete supplied cost | Cost / accepted | Eligible |
|---|---|---|---|---|
| Configuration A | 18 / 24 | $0.24 (actual) | $0.013333 | No |
| Configuration B | 22 / 24 | $0.48 (actual) | $0.021818 | Yes |
Acceptance joins the exact answer contract to the cumulative deadline. Sample eligibility does not establish population reliability or production readiness.
{
"blockers": [],
"recipe": "task-cost@1",
"results": [
{
"accepted": 18,
"actual_total_usd": 0.24,
"attempt_cost_usd": 0.24000000000000007,
"attempts": 24,
"blockers": [
"Joint quality/deadline floor missed."
],
"censored": 0,
"completed_latency_p95_ms": 2840,
"cost_basis": "actual",
"cost_per_success": 0.013333333333333332,
"eligible": false,
"evidence": "synthetic",
"label": "Configuration A",
"quality": 18,
"quote_total_usd": 0.36,
"rows": [
{
"accepted": true,
"attempts": 1,
"id": "task-001",
"quality": true
},
{
"accepted": true,
"attempts": 1,
"id": "task-002",
"quality": true
},
{
"accepted": true,
"attempts": 1,
"id": "task-003",
"quality": true
},
{
"accepted": true,
"attempts": 1,
"id": "task-004",
"quality": true
},
{
"accepted": true,
"attempts": 1,
"id": "task-005",
"quality": true
},
{
"accepted": true,
"attempts": 1,
"id": "task-006",
"quality": true
},
{
"accepted": true,
"attempts": 1,
"id": "task-007",
"quality": true
},
{
"accepted": true,
"attempts": 1,
"id": "task-008",
"quality": true
},
{
"accepted": true,
"attempts": 1,
"id": "task-009",
"quality": true
},
{
"accepted": true,
"attempts": 1,
"id": "task-010",
"quality": true
},
{
"accepted": true,
"attempts": 1,
"id": "task-011",
"quality": true
},
{
"accepted": true,
"attempts": 1,
"id": "task-012",
"quality": true
},
{
"accepted": true,
"attempts": 1,
"id": "task-013",
"quality": true
},
{
"accepted": true,
"attempts": 1,
"id": "task-014",
"quality": true
},
{
"accepted": true,
"attempts": 1,
"id": "task-015",
"quality": true
},
{
"accepted": true,
"attempts": 1,
"id": "task-016",
"quality": true
},
{
"accepted": true,
"attempts": 1,
"id": "task-017",
"quality": true
},
{
"accepted": true,
"attempts": 1,
"id": "task-018",
"quality": true
},
{
"accepted": false,
"attempts": 1,
"id": "task-019",
"quality": false
},
{
"accepted": false,
"attempts": 1,
"id": "task-020",
"quality": false
},
{
"accepted": false,
"attempts": 1,
"id": "task-021",
"quality": false
},
{
"accepted": false,
"attempts": 1,
"id": "task-022",
"quality": false
},
{
"accepted": false,
"attempts": 1,
"id": "task-023",
"quality": false
},
{
"accepted": false,
"attempts": 1,
"id": "task-024",
"quality": false
}
],
"run_id": "fixture-a",
"scope": "Observed fixed-task result. No population reliability estimate or production qualification.",
"success_rate": 0.75,
"tasks": 24,
"total_usd": 0.24
},
{
"accepted": 22,
"actual_total_usd": 0.48,
"attempt_cost_usd": 0.48000000000000015,
"attempts": 24,
"blockers": [],
"censored": 0,
"completed_latency_p95_ms": 2760,
"cost_basis": "actual",
"cost_per_success": 0.021818181818181816,
"eligible": true,
"evidence": "synthetic",
"label": "Configuration B",
"quality": 22,
"quote_total_usd": 0.6,
"rows": [
{
"accepted": true,
"attempts": 1,
"id": "task-001",
"quality": true
},
{
"accepted": true,
"attempts": 1,
"id": "task-002",
"quality": true
},
{
"accepted": true,
"attempts": 1,
"id": "task-003",
"quality": true
},
{
"accepted": true,
"attempts": 1,
"id": "task-004",
"quality": true
},
{
"accepted": true,
"attempts": 1,
"id": "task-005",
"quality": true
},
{
"accepted": true,
"attempts": 1,
"id": "task-006",
"quality": true
},
{
"accepted": true,
"attempts": 1,
"id": "task-007",
"quality": true
},
{
"accepted": true,
"attempts": 1,
"id": "task-008",
"quality": true
},
{
"accepted": true,
"attempts": 1,
"id": "task-009",
"quality": true
},
{
"accepted": true,
"attempts": 1,
"id": "task-010",
"quality": true
},
{
"accepted": true,
"attempts": 1,
"id": "task-011",
"quality": true
},
{
"accepted": true,
"attempts": 1,
"id": "task-012",
"quality": true
},
{
"accepted": true,
"attempts": 1,
"id": "task-013",
"quality": true
},
{
"accepted": true,
"attempts": 1,
"id": "task-014",
"quality": true
},
{
"accepted": true,
"attempts": 1,
"id": "task-015",
"quality": true
},
{
"accepted": true,
"attempts": 1,
"id": "task-016",
"quality": true
},
{
"accepted": true,
"attempts": 1,
"id": "task-017",
"quality": true
},
{
"accepted": true,
"attempts": 1,
"id": "task-018",
"quality": true
},
{
"accepted": true,
"attempts": 1,
"id": "task-019",
"quality": true
},
{
"accepted": true,
"attempts": 1,
"id": "task-020",
"quality": true
},
{
"accepted": true,
"attempts": 1,
"id": "task-021",
"quality": true
},
{
"accepted": true,
"attempts": 1,
"id": "task-022",
"quality": true
},
{
"accepted": false,
"attempts": 1,
"id": "task-023",
"quality": false
},
{
"accepted": false,
"attempts": 1,
"id": "task-024",
"quality": false
}
],
"run_id": "fixture-b",
"scope": "Observed fixed-task result. No population reliability estimate or production qualification.",
"success_rate": 0.9166666666666666,
"tasks": 24,
"total_usd": 0.48
}
],
"scope": "SYNTHETIC ACCOUNTING EXERCISE",
"selectedRecordIds": [
"run-a",
"run-b"
],
"tied": false,
"winner": "fixture-b"
}Authored record state: supported. Pinned inputs: policy@1, run-a@1, run-b@1.
Strict JSON; at least 90% correct and on time on 24 distinct fictional tasks. Whole-batch actual costs include failed attempts.
Authored record state: supported. Pinned inputs: source record.
Synthetic output and billing fixture; no provider was called.
Authored record state: supported. Pinned inputs: source record.
Synthetic output and billing fixture; no provider was called.
Authored record state: supported. Pinned inputs: source record.