generated: '2026-08-13'
method: probed
source: live responses from app.brand.ai
limit_count: 0
rate_limits: []
note: >-
  Brand.ai documents no rate limits — it publishes no developer documentation at all
  (no docs.brand.ai, no developers.brand.ai, no developer page in the sitemap). Nor
  does it SIGNAL any at runtime: the full response headers of the MCP endpoint's 401
  challenge were captured and contain no X-RateLimit-*, no RateLimit-* (RFC 9331 draft),
  and no Retry-After. An agent driving the Brand.ai MCP server therefore has no way to
  learn its budget except by being throttled. A zero here is a measured absence, not an
  unchecked field.
response_headers:
  observed: []
  ratelimit_family: none
  retry_after: false
  exhaustion_status: unknown
headers_present_on_401:
- date
- content-type
- cache-control
- content-security-policy
- content-security-policy-report-only
- referrer-policy
- report-to
- server
- strict-transport-security
- www-authenticate
- x-clerk-auth-reason
- x-clerk-auth-status
- x-content-type-options
- x-matched-path
- x-vercel-cache
- x-vercel-id
- cf-cache-status
- cf-ray
- alt-svc
caveat: >-
  Only unauthenticated responses could be observed. Brand.ai may emit rate-limit
  headers on authenticated MCP traffic; that cannot be established without a token
  and is not asserted either way.
x-evidence:
- url: https://app.brand.ai/api/mcp
  status: 401
  fetched: '2026-08-13'
  note: full header dump inspected; no rate-limit family present