Review a quality check
Rules on the products a production partner photographed. Find tickets waiting for you with list_qc_tickets?needs_your_action=true or the qc_ticket.needs_your_action webhook, and look at each product's photos before deciding. Accepting every product approves the ticket and lets the order ship; a refusal goes to 3DAPI, which rules on it.
What you need
- ticket_idstringRequired
The quality check to rule on (
qct_...). - verdictsarray of objectRequired
One verdict per product in review:
{ "item": "qci_...", "verdict": "accept" }, or{ "item": "qci_...", "verdict": "refuse", "reason": "...", "photos": ["qcp_..."] }pointing at the photos that show the problem.
Step 1: Get a QC ticket
Read the ticket: it is waiting for you when needs_your_action is true. The products to rule on are the items with status: in_review, each with its photos.
{
"path id": "$inputs.ticket_id"
}Succeeds when:
$statusCode == 200$response.body#/needs_your_action == true
Keeps for the next steps:
items:$response.body#/itemsstatus:$response.body#/status
Step 2: Review a quality check
POST/v2/qc-tickets/{id}/review
Send every verdict at once: all are recorded or none is. not_awaiting_review (409) means the ticket or a product is no longer in review; photo_not_on_item (422) means a refusal points at a photo of another product or round.
{
"path id": "$inputs.ticket_id"
}{
"items": "$inputs.verdicts"
}Succeeds when:
$statusCode == 200
Otherwise:
not_awaiting_review: stop the workflow when$statusCode == 409.photo_not_on_item: stop the workflow when$statusCode == 422.
Keeps for the next steps:
status:$response.body#/statusblocks_shipping:$response.body#/blocks_shipping
As a workflow file
The workflow returns:
status:$steps.submit_review.outputs.status
Values starting with $ are filled in as you go: $inputs.x is something you provide, $steps.s.outputs.o is a value an earlier step returned, and $response.body#/path a field of this response.
This recipe is the workflow review_quality_check of arazzo.json, an Arazzo 1.0 document that tools and AI agents can follow step by step.