generated: '2026-09-02'
method: searched
source: >-
  https://medblocks.com/docs/webhooks/redelivery, https://medblocks.com/docs/reference/errors,
  https://medblocks.com/docs/billing, https://medblocks.com/docs/webhooks/retries-and-disable
applies_to: medblocks-platform-api
limit_count: 1
note: >-
  Medblocks documents exactly ONE numeric rate limit — on webhook redelivery. There is no
  published per-key, per-account or per-endpoint request ceiling for the rest of the Platform
  API. What IS published, and what matters more to an agent, is the runtime signal: a 429 with
  a Retry-After header and a typed rate_limit_error envelope, plus a separate 402 billing_error
  path for plan quota. An unauthenticated probe of GET https://app.medblocks.com/health on
  2026-09-02 returned no X-RateLimit-* or RateLimit-* headers, so no standard-header budget
  disclosure was observed on an anonymous response; whether authenticated responses carry them
  is not documented and was not tested.
rate_limits:
- scope: per-event
  surface: POST /events/{id}/redeliver (api.redeliverWebhookEvent)
  limit: 1
  window: 60s
  burst: null
  on_exhaustion:
    status: 429
    error_type: rate_limit_error
    header: Retry-After
    sdk_field: err.retryAfter
    sdk_default_backoff_seconds: 60
  docs: https://medblocks.com/docs/webhooks/redelivery
  note: >-
    "Redelivery is rate-limited to one call per minute per event. A second call within 60
    seconds throws a MedblocksError with type rate_limit_error and the Retry-After header
    parsed onto err.retryAfter."
response_headers:
- name: Retry-After
  emitted_on: 429
  documented: true
  observed: false
  note: Documented as set on rate_limit_error and null otherwise. Not observed on an anonymous 200.
- name: X-Request-Id
  emitted_on: every response
  documented: true
  observed: true
  note: Not a rate-limit header, but the correlation id an agent should carry into any limit escalation.
- name: X-RateLimit-Limit / X-RateLimit-Remaining / RateLimit-*
  documented: false
  observed: false
  note: >-
    Not documented and not present on the anonymous GET /health probe of 2026-09-02. Absence
    of a budget header is the finding: a caller cannot see how close it is to a limit until
    the 429 arrives.
exhaustion_codes:
- {status: 429, type: rate_limit_error, code: throttled, meaning: Too many requests.}
- {status: 429, type: rate_limit_error, code: quota_exceeded, meaning: API key quota is exceeded.}
- {status: 429, type: rate_limit_error, code: api_key_limit_exceeded, meaning: 'Organization has reached its active API key cap. Revoke an existing key first.'}
- {status: 402, type: billing_error, code: quota_exceeded, meaning: Free-tier active-patient limit has been reached.}
- {status: 402, type: billing_error, code: feature_not_enabled, meaning: Workspace plan does not include the requested feature.}
client_retry_behaviour:
  sdk: medblocks (npm)
  retries_on: [429, 502, 503, 504, network failures]
  does_not_retry: [other 4xx, 500]
  max_network_retries_default: 3
  honors_retry_after: true
  docs: https://medblocks.com/docs/reference/errors
adjacent_quotas:
  patient_quota:
    plan: Free
    limit: 1000
    unit: active patients
    definition: >-
      A patient counts as active the first time Medblocks successfully pulls FHIR data for
      them; creating a patient record alone does not draw against the quota.
    on_exhaustion: {status: 402, type: billing_error, code: quota_exceeded}
    docs: https://medblocks.com/docs/billing
  webhook_delivery_retries:
    attempts: 9
    schedule: [immediate, 1m, 5m, 30m, 2h, 12h, 24h, 24h, 24h]
    worst_case: ~3.6 days from first failure to exhaustion
    on_exhaustion: endpoint status flips to "disabled" with no event emitted
    receiver_timeout: 5s
    docs: https://medblocks.com/docs/webhooks/retries-and-disable
    note: This is Medblocks' outbound retry budget against your receiver, not a limit on your calls.