Skip to content

Mail authentication

The reader displays SPF, DKIM, and DMARC outcomes parsed from a received message's trusted Authentication-Results header. These indicators report what a receiving mail service evaluated; mail does not perform DNS-based SPF, DKIM, or DMARC validation itself.

Trust model

An Authentication-Results header is trusted only when its authserv-id matches the configured allowlist. Exact IDs and wildcard subdomains are supported. For example, *.google.com matches a Google subdomain but not an unrelated domain ending in the same text.

Without a matching trusted ID, the reader can still show parsed statuses, but the indicator is visually marked as not reported by a trusted receiving service. Configure only IDs emitted by the mail system that actually received the message; trusting an arbitrary sender-controlled ID defeats the boundary.

Configuration

ItemSetting nameEnvironment variableRequirement
Trusted receiving servicesNoneMAIL_AUTH_TRUSTED_AUTHSERV_IDSOptional but required for results to be marked trusted. Use a comma-separated list of exact IDs or *. wildcard suffixes.
Authentication parsing and backfillNoneNoneAutomatic for synchronized mail. The worker must be running.

Set MAIL_AUTH_TRUSTED_AUTHSERV_IDS identically on the web and worker processes so stored results and reader trust indicators use the same policy.