generated: '2026-09-19'
method: searched
source: https://theloopbreaker.com/llms.txt
description: >-
  llms.txt ("Rate Limits") is the only published statement: "All API endpoints return standard rate limit
  headers" — X-RateLimit-Limit 120 (requests per minute), X-RateLimit-Remaining, X-RateLimit-Reset (Unix
  timestamp), and Retry-After in seconds on 429. Live check 2026-09-19: 200 responses from GET /api/health,
  GET /api/hub/stats and GET /api/agent/status carried NONE of these headers (only Vercel cache/HSTS/CSP
  headers), so the runtime signal the docs promise was not observed. The number is recorded from the docs;
  the header absence is recorded as a finding. No burst, per-endpoint or per-key tiers are published (there
  are no keys).
limit_count: 1
limits:
  - scope: per-client (undocumented key — no API keys exist; presumably per IP)
    window: 1 minute
    limit: 120
    burst: null
    applies_to: all /api/* endpoints (per llms.txt)
    status_on_exhaustion: 429
    headers:
      - {name: X-RateLimit-Limit, value: '120', observed_live: false}
      - {name: X-RateLimit-Remaining, value: requests left in current window, observed_live: false}
      - {name: X-RateLimit-Reset, value: Unix timestamp when the window resets, observed_live: false}
      - {name: Retry-After, value: seconds to wait (only on 429), observed_live: false}
    source: https://theloopbreaker.com/llms.txt
observed:
  date: '2026-09-19'
  requests:
  - GET /api/health
  - GET /api/hub/stats
  - 'GET /api/agent/status?address=0xfA15...813C'
  rate_limit_headers_present: false
  other_headers: [cache-control, strict-transport-security, x-content-type-options, x-frame-options, x-vercel-cache, x-vercel-id]
x402_timeout: 'Payment challenges carry maxTimeoutSeconds 60 — a payment window, not a rate limit.'