Outbound SMTP relay ignores domain adult-content/phishing protection settings

Outbound messages sent via SMTP relay are unconditionally scanned for phishing, malware, and adult-content links (via Cloudflare's Family and Security DNS resolvers) in Emails.queue(), app/models/emails.js around lines 1787 to 1889, regardless of the domain's has_adult_content_protection / has_phishing_protection settings (app/models/domains.js, lines 358 to 365, both default true). The domain object is fully loaded in that function, but neither field is ever referenced there.

Repro: uncheck "Adult-related content protection" on a domain's Advanced Settings page, then send an outbound message via SMTP relay containing a link to a hostname Cloudflare's Family DNS (family.cloudflare-dns.com) categorizes as adult/sensitive. The message still bounces with a 554 5.6.0 from the same check the toggle is supposed to control.

Expected: unchecking the setting suppresses the adult-content portion of this check for outbound sends, matching how it's described on the settings page and in the FAQ.

Actual: the setting has no effect on outbound relay, the check always runs.

Suggested fix: gate the message-building logic in Emails.queue() on domain.has_adult_content_protection / domain.has_phishing_protection, mirroring the equivalent (currently inactive, see Issue 2) logic in helpers/get-recipients.js lines 137 to 165.

Start in app/models/emails.js, reading Emails.queue() around lines 1787–1889, then compare the protection logic in helpers/get-recipients.js lines 137–165 and the settings fields in app/models/domains.js lines 358–365. Done means outbound SMTP relay checks honor the domain’s adult-content and phishing protection settings, including the reported adult-content toggle scenario.

Written by the indexing model from the issue text.