generated: '2026-08-12'
method: probed
source: >-
  Response headers observed live on https://www.ethyrealbio.com/wp-json/wp/v2/* on 2026-08-12.
  Ethyreal Bio publishes no API documentation, so there is no docs page to cite.
limit_count: 0
published: false
note: >-
  Ethyreal Bio documents NO rate limits, and no rate-limit response headers are returned. This
  is an honest zero, not an unchecked field: every probed response was inspected and carried no
  X-RateLimit-*, no RateLimit-* (RFC 9331 style) and no Retry-After. An agent gets no runtime
  budget signal from this API whatsoever.
limits: []
headers_observed:
  rate_limit_headers: []
  retry_after: false
  note: >-
    The complete set of non-standard headers returned was X-WP-Total, X-WP-TotalPages, Link,
    Allow, X-Powered-By (WP Engine), X-Cacheable, X-Cache, X-Cache-Group, X-Robots-Tag and
    Cloudflare's cf-ray / cf-cache-status. None of them carry a quota.
de_facto_throttle:
  mechanism: cloudflare-bot-management
  status_code: 403
  content_type: text/html
  body: Cloudflare "Attention Required!" interstitial
  observed: true
  evidence:
  - url: https://www.ethyrealbio.com/wp-json/
    status: 403
    note: Returned mid-session on a path that had returned 200 moments earlier.
  - url: https://www.ethyrealbio.com/feed/
    status: 403
    note: Returned 200 on retry with a browser User-Agent.
  description: >-
    The effective limiter is Cloudflare bot management in front of WP Engine, not an API quota.
    It is triggered by client fingerprint rather than request volume — a non-browser User-Agent
    is challenged where a browser User-Agent on the same path succeeds. There is no header that
    announces it, no Retry-After telling a client how long to wait, and the 403 body is HTML,
    so a JSON client will fail to parse the error as well as fail the request.
  consequence_for_agents: >-
    An automated consumer cannot distinguish "you are over quota" from "you are blocked as a
    bot" from "you lack permission" — all three are 403. Back off on any 403 whose content-type
    is not application/json, and treat it as a client-identity problem rather than a quota one.
crawl_policy:
  robots_txt: https://www.ethyrealbio.com/robots.txt
  crawl_delay: 10
  disallow: []
  note: >-
    robots.txt allows all user agents (an empty Disallow) but requests a 10-second crawl delay.
    That is the only throttling guidance Ethyreal Bio publishes anywhere, and it governs
    crawlers rather than API clients.
caching_as_mitigation:
  cache_control: max-age=600, must-revalidate
  note: >-
    Responses are cacheable for 10 minutes and Last-Modified is returned, so conditional
    requests are the practical way to stay under the radar. There is nothing to poll faster
    for: the site publishes roughly one press release a month.