MERCHANT REVIEW EVIDENCE

Creem compliance verification summary.

Read-only evidence for Moonwish’s public pricing, policies, support contact, and server-side prompt moderation gate.

Repository verification: September 18, 2026

Public pages and support

Moderation verification

MODERATION_ENABLED
YES — server-side and fail-closed
MODERATION_REJECTS_NSFW
PASS — mocked contract tests
PROVIDER_CALLS_AFTER_MODERATION_REJECT
0
REAL_ARK_CALLS
0 during verification tests
REAL_SEEDREAM_CALLS
0 during verification tests
CREEM_REAL_MODERATION_CALLS
Not run without explicit approval

Covered entry points: Custom Story and Template submissions, edited outlines before illustration, direct Pollinations prompts, and client-supplied legacy illustration recovery prompts.

Verified Creem API contract

  • Endpoint: POST https://api.creem.io/v1/moderation/prompt
  • Authentication: x-api-key
  • Request: required prompt, optional external_id
  • Decisions: allow, flag, deny
  • Moonwish action: allow only allow; block flag, deny, timeouts, network errors, non-2xx responses, and invalid responses.
  • Documented timeout guidance: approximately 5 seconds.
  • Rate limits: not documented on the referenced Creem page.

Known image-moderation boundary

Creem’s published contract documents text prompt screening only. Moonwish therefore retains separate file validation and vision-based photo safety checks for uploaded child and family photos. This page does not claim that Creem performs image moderation.

Customer Support: marson.mhs@gmail.com