ProvelyProve that every request of the batch job succeeded. A batch that reaches the status completed is not a batch in which every request succeeded. This level reads request_counts and refuses a batch with a failed count above zero.
| Fact | Value |
|---|---|
| Contract id | openai.batch.all_requests_succeeded |
| Version | 1.0.0 |
| Hash | f7efcd09d699f148 |
| Completion level | all_requests_succeeded |
| Publisher | Built and signed by Provely. |
| Certification | Provisional |
| Skill | OpenAI Platform API 0.1.0 |
| Valid for provider API versions | spec_2_3_0 |
What is the intent?
Run every request of the input file as one batch job, and get a result for each one.
What is the subject and the action?
| Member | Value |
|---|---|
| Subject type | openai.batch |
| Canonical effect | ai.batch_job_start |
| Provider operation | POST /batches |
| Idempotency | none, retry is not safe |
Which false successes give CONTRADICTED?
- partial completion
- pre existing state
- wrong subject
The signed contract states the exact conditions, the evidence and the timing. The runtime reads them. The page states the promise.
Where do these rules come from?
openai.openapi#/components/schemas/Batch/properties/request_counts: openai.openapi, retrieved 2026-09-09openai.openapi#/components/schemas/BatchRequestCounts/properties/failed: openai.openapi, retrieved 2026-09-09openai.docs.batch#batch-jobs/the-request-counts-of-a-batch: openai.docs.batch, retrieved 2026-09-09openai.docs.batch#batch-jobs/the-output-file-and-the-error-file: openai.docs.batch, retrieved 2026-09-09openai.openapi#/components/schemas/Batch/properties/id: openai.openapi, retrieved 2026-09-09openai.openapi#/paths/~1batches~1{batch_id}/get/responses/200: openai.openapi, retrieved 2026-09-09openai.openapi#/components/schemas/Batch/properties/status: openai.openapi, retrieved 2026-09-09openai.docs.batch#batch-jobs/the-status-of-a-batch: openai.docs.batch, retrieved 2026-09-09openai.openapi#/components/schemas/Batch/properties/input_file_id: openai.openapi, retrieved 2026-09-09openai.openapi#/components/schemas/CreateBatchRequest/properties/input_file_id: openai.openapi, retrieved 2026-09-09openai.openapi#/components/schemas/Batch/properties/created_at: openai.openapi, retrieved 2026-09-09openai.events#/events/batch.completed: openai.events, retrieved 2026-09-09openai.openapi#/components/schemas/WebhookBatchCompleted/properties/data: openai.openapi, retrieved 2026-09-09openai.docs.webhooks#webhooks/the-event-types: openai.docs.webhooks, retrieved 2026-09-09openai.openapi#/components/schemas/BatchRequestCounts/properties/total: openai.openapi, retrieved 2026-09-09openai.openapi#/components/schemas/BatchRequestCounts/properties/completed: openai.openapi, retrieved 2026-09-09
Can openai.batch.all_requests_succeeded return VERIFIED from the action response alone?
No. The action response is E1, and an acknowledgement is never terminal success. The contract needs evidence of level E2 or higher from OpenAI Platform API.
What happens after the timeout?
The verdict is UNVERIFIABLE. The operation goes to a person for review.