generated: '2026-08-12'
method: probed
source: >-
  Observed responses from https://api.tvscientific.app/v1/ and /v2/ on 2026-08-12
  (see discovery/tvscientific-api-discovery.yml). tvScientific publishes no public API
  reference, so every convention below is read from a real HTTP response, not from docs.
api: tvScientific Campaign Management API
base_url: https://api.tvscientific.app
docs: null
docs_note: >-
  No public developer portal or API reference exists. help.tvscientific.com 301s to an
  Atlassian Service Desk login; support.tvscientific.com redirects into a Salesforce
  Community SAML flow. Conventions here are PROBED, not documented by the provider.

auth:
  style: unknown-credentialed
  observed: >-
    Every collection except /v1/terms_and_conditions/ returns HTTP 403 with
    {"detail":"Authentication credentials were not provided."}
  challenge_header: none
  note: >-
    Django REST Framework answers 403 rather than 401 and sends no WWW-Authenticate header,
    so the credential type (session cookie, token, key) cannot be determined anonymously.
    `Vary: Accept, Cookie, origin` implies at least session-cookie authentication is one of
    the accepted schemes.
  see: authentication/tvscientific-authentication.yml

versioning:
  style: uri-path
  observed_versions: [v1, v2]
  roots:
    - https://api.tvscientific.app/v1/
    - https://api.tvscientific.app/v2/
  note: >-
    Two live major versions run side by side. v2 is a narrower campaign-buying surface
    (campaigns, lineitems, creatives, targeting, bidder-config, quickstart); v1 carries the
    full account/billing/user estate. No deprecation header, Sunset header or published
    version policy was observed on either root.
  media_type_versioning: false
  header_versioning: false

pagination:
  style: page-number
  params:
    - name: page
      observed: true
      detail: '?page=2 on a 2-record collection returned 404 {"detail":"Invalid page."}'
    - name: page_size
      observed: true
      detail: '?page_size=1 returned next=".../?page=2&page_size=1"'
  response_fields:
    count: total record count
    next: absolute URL of the next page, or null
    previous: absolute URL of the previous page, or null
    results: the array of records
  envelope_example_source: https://api.tvscientific.app/v1/terms_and_conditions/
  note: >-
    Standard Django REST Framework PageNumberPagination. `limit`/`offset` were accepted
    without error but had no effect on the returned window, so they are ignored rather than
    supported.

filtering_and_sorting:
  observed_params: [ordering, search]
  note: >-
    ?ordering= and ?search= were accepted on the one public collection without error, which
    is consistent with DRF OrderingFilter/SearchFilter being installed, but their effect
    could not be confirmed on a 2-record collection. Recorded as accepted, not as supported.

content_negotiation:
  request_formats: [application/json, application/x-www-form-urlencoded, multipart/form-data]
  response_formats: [application/json, text/html]
  format_override_param: format
  format_values: [json, api]
  note: >-
    ?format=api renders the DRF browsable HTML API. Read from the DRF OPTIONS metadata
    document at /v1/terms_and_conditions/ (renders/parses arrays) and confirmed by request.

resource_identity:
  style: hyperlinked
  detail: >-
    Records carry an absolute `url` field that is the canonical self link, and relationships
    are expressed as absolute URLs rather than bare ids (e.g. a terms_and_conditions record's
    `tenant` is "https://api.tvscientific.app/v1/tenants/1/"). Integer surrogate `id` is also
    present. This is a DRF HyperlinkedModelSerializer.
  trailing_slash: required
  trailing_slash_detail: >-
    Collection and detail paths end in "/". Requests without the trailing slash 404 rather
    than redirect.

schema_introspection:
  mechanism: HTTP OPTIONS (Django REST Framework metadata)
  detail: >-
    OPTIONS on a permitted collection returns a JSON document with the view name, its
    description, the renders/parses media types, and an `actions.POST` map giving each
    writable field's type, required flag, read_only flag, label and max_length. This is the
    closest thing to a machine-readable request schema tvScientific exposes.
  publicly_reachable: false
  publicly_reachable_note: >-
    OPTIONS is 403 on every collection except /v1/terms_and_conditions/, so the field schemas
    for campaigns, creatives, lineitems, audiences and the rest require credentials. A
    customer with a token can introspect the whole surface this way.
  example: discovery/tvscientific-options-terms-and-conditions.json

idempotency:
  supported: false
  header: null
  note: >-
    No Idempotency-Key or equivalent header was advertised on any response, and no public
    documentation describes one. Recorded as ABSENT — no Idempotency pointer is emitted in
    apis.yml.

request_tracing:
  request_id_header: null
  note: No X-Request-Id / X-Correlation-Id / traceparent header was returned on any probe.

rate_limit_signalling:
  headers_observed: []
  note: >-
    No X-RateLimit-*, RateLimit-* or Retry-After header appeared on any anonymous response,
    including repeated calls. See rate-limits/tvscientific-rate-limits.yml.

errors:
  envelope: '{"detail": "<human readable message>"}'
  rfc9457: false
  content_type: application/json
  note: >-
    Django REST Framework's default error envelope. No `type`, `title`, `status` or
    `instance` member, no application/problem+json media type, and no machine-readable error
    code. See errors/tvscientific-problem-types.yml.

security_headers:
  x-frame-options: DENY
  x-content-type-options: nosniff
  referrer-policy: same-origin
  cross-origin-opener-policy: same-origin
  server: gunicorn

cross_links:
  errors: errors/tvscientific-problem-types.yml
  authentication: authentication/tvscientific-authentication.yml
  lifecycle: lifecycle/tvscientific-lifecycle.yml
  rate_limits: rate-limits/tvscientific-rate-limits.yml
  discovery: discovery/tvscientific-api-discovery.yml