generated: '2026-08-26'
method: searched
source: >-
  Searched every published Rose Rocket developer page on 2026-08-26 — the full page
  index at https://roserocket.readme.io/llms.txt, /docs/getting-started,
  /docs/rose-rocket-api-oauth-20-authentication-guide,
  /docs/object-descriptions-and-operations, /docs/upserts-and-external-ids,
  /docs/webhooks-2, /docs/roles-and-permissions,
  /docs/application-lifecycle-management (all HTTP 200) — plus every response
  declaration in openapi/rose-rocket-platform-model-api.json and the pricing page at
  https://www.roserocket.com/pricing (HTTP 200).
limit_count: 0
description: >-
  Rose Rocket publishes NO rate limits for the Platform Model API. This is a genuine
  zero, not an unchecked one: no limits page exists, no quota or throttle is
  mentioned on any of the eight documentation pages, no 429 response is declared on
  any of the 19 operations in the spec, and no rate-limit header appears anywhere in
  the contract or the docs. The nearest thing to a statement is the Enterprise tier
  feature bullet "unlimited API access" on the pricing page, which is a commercial
  entitlement rather than a runtime limit. An agent integrating here has no published
  budget to plan against and no runtime signal to back off on — it will discover any
  throttle only by being throttled, and cannot distinguish that from a 500.
limits: []
headers:
  documented: false
  standard_ratelimit_headers: not documented
  x_ratelimit_headers: not documented
  retry_after: not documented
  note: >-
    Neither the RFC 9239-style RateLimit-* family nor the X-RateLimit-* convention is
    documented, and no Retry-After is mentioned. Not probed live: the API requires an
    OAuth bearer token issued by an account representative, so no unauthenticated
    response is observable.
status_on_exhaustion:
  documented: false
  note: >-
    429 does not appear in the spec or in either of the two error tables the docs
    publish (object CRUD, webhooks). Both tables jump from 404 to 500.
adjacent_limits_note: The only quantified operational limits Rose Rocket does publish. None of them is a request rate.
adjacent_limits:
  - {scope: webhook delivery retries, limit: 3 attempts, window: 30-second delay between attempts, source: 'https://roserocket.readme.io/docs/webhooks-2'}
  - {scope: webhook receipt history, limit: 7 days, window: rolling, source: 'https://roserocket.readme.io/docs/webhooks-2'}
  - {scope: access token lifetime, limit: deliberately unspecified and variable, window: n/a, source: 'https://roserocket.readme.io/docs/rose-rocket-api-oauth-20-authentication-guide'}
  - {scope: POST /objects/search result set, limit: caller-supplied `limit` with no documented maximum and no way to page past it, window: per request, source: openapi/rose-rocket-platform-model-api.json}
gaps:
  - No published request rate, burst, quota or concurrency limit at any scope.
  - No 429 declared and no exhaustion behaviour documented.
  - No rate-limit response headers, so a client cannot adapt at runtime.
  - '"Unlimited API access" is asserted on the Enterprise tier without a fair-use definition.'