generated: '2026-09-19'
method: searched
source: https://agentcheck.care/terms
docs:
- https://agentcheck.care/terms
- https://agentcheck.care/privacy
- https://agentcheck.care/api/free-scans
limit_count: 2
summary: >-
  Two published limits, both on the free tier and both quota-shaped rather than requests-per-second: a
  shared pool of 30 free scans per week across all users, and 2 free scans per week per individual user
  (terms section 3.1). The pool is exposed live and anonymously at GET /api/free-scans as {used, max,
  remaining, reset_in_seconds, reset_label} — observed 2026-09-19 as max 30, remaining 30, reset in ~26h.
  The privacy policy (1.4) states IP addresses are used for "server-side rate limiting" and that Cloudflare
  bot management fronts the origin, but no numeric request limit, window, or exhaustion status is
  published for the REST API, and no RateLimit-*/X-RateLimit-*/Retry-After headers were observed on
  anonymous 200 responses from /api/tiers, /api/free-scans or the agent card.
rate_limits:
- name: Free scan global pool
  scope: global (shared across all users)
  limit: 30
  window: 1 week
  metric: free checkups
  burst: null
  applies_to:
  - POST /api/checkup (tier "free")
  - A2A skill free-scan
  signal: >-
    GET /api/free-scans returns used/max/remaining/reset_in_seconds; the home page shows the same counter
    and an "exhausted" timer
  source: https://agentcheck.care/terms
- name: Free scans per user
  scope: per-user
  limit: 2
  window: 1 week
  metric: free checkups
  burst: null
  applies_to:
  - POST /api/checkup (tier "free")
  - A2A skill free-scan
  source: https://agentcheck.care/terms
  note: >-
    How a "user" is identified without login is not published; the privacy policy names IP address as
    the rate-limiting key.
undocumented:
- name: IP-based request rate limiting
  scope: per-ip
  limit: null
  window: null
  source: https://agentcheck.care/privacy
  note: >-
    Stated to exist ("IP address - used for rate limiting", privacy 1.4); no numbers published.
- name: Bot connection window during a checkup
  limit: 5 minutes
  source: https://agentcheck.care/terms
  note: Not a request limit — the time AgentCheck will keep trying to reach the bot under test before the checkup fails (terms 3.4).
headers:
  observed: []
  exhaustion_status: undocumented
  note: >-
    Observed response headers on anonymous JSON responses were content-type, server: cloudflare,
    x-frame-options, x-content-type-options, x-xss-protection and (on the card) cache-control. No rate-limit
    signaling header of any family. An agent has to poll /api/free-scans to learn whether a free scan will
    be accepted.