generated: '2026-09-03'
method: searched
source: https://mystars.tg/docs
note: >-
  Rate limits are documented in the developer portal overview (added v1.5.0, 2026-06-21; the
  order-lifecycle bucket split off in v1.10.0). Limits and headers below are quoted from the docs
  and the published OpenAPI RateLimited response component.
limit_count: 4
limits:
  - scope: per-tenant, /v1 read endpoints
    window: 1 minute
    limit: 60
    note: Default per-minute request budget; read the live ceiling from RateLimit-Limit rather than hard-coding.
  - scope: per-tenant, order lifecycle (createOrder, getOrder, cancelOrder)
    window: 1 minute
    limit: 60
    note: >-
      A separate bucket of the same per-minute size, isolated from the read budget - a burst of
      pricing/catalog reads can never throttle order create, poll, or cancel.
  - scope: per-tenant probe cap (getPricing, getPricingBatch, checkRecipient)
    window: 1 minute
    limit: 60
    note: >-
      Tighter cap in addition to the general budget because each call drives an upstream
      price/eligibility lookup. A batch pricing call counts as one unit. (Raised 10 -> 30/min in
      v1.5.0; the current docs and spec state 60/min.)
  - scope: per-tenant daily order cap + per-recipient concurrent-order guard (createOrder)
    window: 1 day / short-term
    limit: null
    note: >-
      Optional daily order cap and a flood guard on concurrent orders to the same @username; numbers
      are not published. These 429s carry the error envelope only, without RateLimit-* headers.
headers:
  - RateLimit-Limit
  - RateLimit-Remaining
  - RateLimit-Reset
  - Retry-After
exhaustion_status: 429
exhaustion_envelope: '{ "error": { "code": "rate_limited", "message": "..." } }'