generated: '2026-09-06'
method: searched
source: https://acadio.com/llms.txt and https://acadio.com/agents.md
docs: https://acadio.com/agents.md
description: >-
  Acadio publishes exactly one rate-limit statement, in its agent instructions, and it is
  qualitative. No numeric limit, window, burst or quota is published for either the storefront MCP
  endpoint or the api.acadio.com LMS gateway, and no RateLimit-* or Retry-After header was returned
  on a live unauthenticated request.
limit_count: 0
limits: []
published_guidance:
- surface: https://acadio.com/api/ucp/mcp
  scope: per-ip
  window: null
  limit: null
  burst: null
  statement: >-
    "Respect rate limits. The MCP endpoint is rate-limited per IP. Back off on 429 responses."
  exhaustion_status: 429
  source: https://acadio.com/agents.md
headers_observed:
  probed: '2026-09-06'
  method: POST tools/list to https://acadio.com/api/ucp/mcp (HTTP 200)
  ratelimit_headers: []
  retry_after: false
  note: >-
    Response headers carried server, server-timing and x-request-id only. No X-RateLimit-*,
    RateLimit-* (RFC 9238 draft) or Retry-After header was present on a successful response; whether
    one appears on a 429 could not be established without deliberately exhausting the limit, which
    this pipeline does not do.
lms_surface:
  base: https://api.acadio.com
  limits_published: false
  note: >-
    No rate limits are documented for the LMS gateway. The webhook documentation states a delivery
    retry budget (up to 6 attempts on a non-2xx) which governs Acadio's outbound calls to a
    subscriber, not a subscriber's inbound calls to Acadio.