generated: '2026-08-14'
method: probed
source: >-
  live probe of https://golden.com/api/v2/public/schema/entityTypes/ (2026-08-14)
  plus https://docs.golden.com/openapi.json and https://golden.com/pricing
docs: https://docs.golden.com/
notes: >-
  Golden API v2 publishes NO numeric request-rate limit and returns NO
  rate-limit response headers. Two live probes of
  https://golden.com/api/v2/public/schema/entityTypes/ (anonymous, and with an
  invalid key) both returned HTTP 403 with an application/json DRF error and a
  header set carrying no X-RateLimit-*, no RateLimit-*, and no Retry-After —
  only allow, cache-control, vary and the Cloudflare edge headers. The only
  throttling signal the provider documents is the 429 `throttled` response
  declared on every operation in the OpenAPI, with no window, quota, or
  Retry-After contract attached. Consumption is instead metered commercially in
  monthly "data credits" per plan — see
  plans/golden-recursion-plans-pricing.yml. The one tier-dependent limit the
  spec states is the `pageSize` page cap.
limit_count: 3
limits:
  - id: throttle-429
    scope: per-key
    window: undocumented
    limit: null
    burst: null
    status_on_exhaustion: 429
    error_code: throttled
    error_envelope: '{ "type": "client_error", "errors": [ { "code": "throttled", "detail": "Request was throttled.", "attr": null } ] }'
    applies_to: all operations
    source: docs.golden.com/openapi.json (429 response on every operation)
    note: >-
      Documented as a response, not as a published rate. No numeric ceiling and
      no window are stated anywhere on the developer surface.
  - id: page-size-cap
    scope: per-request
    window: n/a
    limit: 100
    unit: results per page
    parameter: pageSize
    defaults:
      free_plan: 2
      paid_plan: 10
      schema_endpoints: 100
    applies_to: all list operations
    source: docs.golden.com/openapi.json (pageSize parameter description)
  - id: data-credit-quota
    scope: per-account
    window: month
    limit: null
    unit: data credits
    by_plan:
      Free: 10
      Team: 125
      Business: 2500
      Enterprise: custom
    applies_to: all metered data retrieval (API and web app share the pool)
    source: https://golden.com/pricing
    note: >-
      A commercial quota, not a request-rate limit. No documented API response
      signals remaining credit.
response_headers:
  observed: []
  documented: []
  retry_after: false
  note: >-
    Probed live 2026-08-14 on a 403 from the production API host. Response
    headers were: date, content-type, allow, cache-control, report-to,
    cross-origin-opener-policy, gld-build-number, referrer-policy, server, vary,
    x-content-type-options, x-frame-options, nel, cf-cache-status, cf-ray,
    alt-svc. No rate-limit family header of any kind is emitted, so an agent
    cannot read remaining budget or a back-off hint from the runtime.
x-evidence:
  - url: https://golden.com/api/v2/public/schema/entityTypes/
    status: 403
    fetched: '2026-08-14'
    note: anonymous; no rate-limit headers present
  - url: https://docs.golden.com/openapi.json
    status: 200
    fetched: '2026-08-14'
  - url: https://golden.com/pricing
    status: 200
    fetched: '2026-08-14'