Prohibited Practices Screening API

The Prohibited Practices Screening API manages product-scoped Article 5 screenings for AI products. It creates a default prohibited-practices questionnaire, saves draft answers, evaluates screenings, and records governance overrides for flagged outcomes.

All paths below are under /api/v1.0. Every endpoint is scoped to a product inside a space and requires space access.

Endpoints Overview

Method	Endpoint	Description
POST	/catalog/spaces/{space_slug}/products/{product_id}/prohibited-practices-screenings/	Create a draft screening.
GET	/catalog/spaces/{space_slug}/products/{product_id}/prohibited-practices-screenings/	List screenings for one product.
GET	/catalog/spaces/{space_slug}/products/{product_id}/prohibited-practices-screenings/{screening_id}	Retrieve one screening.
PUT	/catalog/spaces/{space_slug}/products/{product_id}/prohibited-practices-screenings/{screening_id}	Update draft responses.
POST	/catalog/spaces/{space_slug}/products/{product_id}/prohibited-practices-screenings/{screening_id}/submit	Submit and evaluate saved responses.
POST	/catalog/spaces/{space_slug}/products/{product_id}/prohibited-practices-screenings/{screening_id}/override	Apply a governance override to a flagged screening.

Status And Outcome Model

Field	Values
Question answer	yes, no, na, unanswered
status	draft, cleared, flagged
outcome	not_assessed, cleared, prohibited
override_outcome	exception_granted when an override is applied

Submitting rejects screenings with unanswered questions. Any yes answer produces outcome: "prohibited", sets status: "flagged", and records flagged_articles. If all answers are no or na, the screening is cleared.

Create Screening

POST /api/v1.0/catalog/spaces/{space_slug}/products/{product_id}/prohibited-practices-screenings/

{
  "responses": [
    {
      "article_ref": "5(1)(a)",
      "question": "Does the system use subliminal techniques beyond a person's consciousness to materially distort behaviour?",
      "answer": "unanswered",
      "notes": ""
    }
  ]
}

responses is optional. If omitted, Qarion creates the default Article 5 questionnaire with one question per prohibited practice.

List Screenings

GET /api/v1.0/catalog/spaces/{space_slug}/products/{product_id}/prohibited-practices-screenings/?limit=20&offset=0

The list is ordered newest first. limit is 1-100.

Update Screening

PUT /api/v1.0/catalog/spaces/{space_slug}/products/{product_id}/prohibited-practices-screenings/{screening_id}

{
  "responses": [
    {
      "article_ref": "5(1)(a)",
      "question": "Does the system use subliminal techniques beyond a person's consciousness to materially distort behaviour?",
      "answer": "no",
      "notes": "The system does not use subliminal techniques."
    }
  ]
}

Only draft screenings can be updated.

Submit for Evaluation

POST /api/v1.0/catalog/spaces/{space_slug}/products/{product_id}/prohibited-practices-screenings/{screening_id}/submit

The request body is empty. The current frontend saves the draft with PUT before calling submit. The API evaluates the stored responses and returns the updated screening.

Apply Governance Override

POST /api/v1.0/catalog/spaces/{space_slug}/products/{product_id}/prohibited-practices-screenings/{screening_id}/override

{
  "justification": "Reviewed supporting evidence and determined this is a documented false positive."
}

Overrides are accepted only for flagged screenings. The justification must be at least 10 characters. The response records reviewer_id, reviewed_at, override_justification, and override_outcome: "exception_granted".

Response Fields

Field	Description
id, product_id, space_id	Screening identity and scope.
status, outcome	Lifecycle and evaluation result.
responses	Questionnaire answers with article refs, questions, answers, and notes.
flagged_articles	Article references where the answer was yes.
screened_by_id, screened_at	Creator and evaluation timestamp.
reviewer_id, reviewed_at	Override reviewer metadata.
override_justification, override_outcome	Governance override details.
created_at, updated_at	Audit timestamps.

Create, update, submit, and override actions emit audit events for prohibited_practices_screening.

Error Responses

Status	Description
400	Updating or submitting a non-draft screening, unanswered questions, overriding a non-flagged screening, or invalid payload.
401	Authentication required for create and override operations.
404	Product or screening not found in the space.