generated: '2026-09-19'
method: searched
source: https://github.com/luminarylane/lane-plugin
docs: https://github.com/luminarylane/docs
limit_count: 0
summary: >-
  No request rate limit is published for the remote MCP server at https://mcp.luminarylane.app/mcp, and
  no RateLimit-*, X-RateLimit-* or Retry-After header appeared on any of the anonymous responses observed
  (401 on initialize/tools/list, 400 on GET /mcp, 200 on /health). The commercial ceiling is the weekly
  Lane Cycle quota per plan (see plans/), which governs how often the Lane agent runs, not how often an
  MCP client may call a tool. An honest zero: the limits may exist server-side but are not documented.
rate_limits: []
headers_observed: []
exhaustion_status: unknown
quota_limits:
  - name: Lane Cycles per week
    scope: per-account (allocated across active spaces)
    window: Monday 00:00 to next Monday 00:00 in the account timezone
    limit: 1 / 5 / 15 / 35 by plan (Discover / Growth / Scale / Agency)
    on_exhaustion: '"Lane pauses scheduled runs until the fresh budget opens Monday. Anything already queued picks up in the new week."'
    source: https://github.com/luminarylane/docs/blob/main/account-and-billing.md
    note: A product quota on agent runs, not an API request limit.
third_party_package_note: >-
  The provider's open-source stdio MCP servers document client-side token-bucket limiters and 429 retry
  for the third-party APIs they wrap (e.g. GA4's 10 concurrent requests, LinkedIn auto-retry on 429);
  those are the upstream platforms' limits, not Luminary Lane's.