#18: implement SMTP Submission (587/465), SASL PLAIN/LOGIN auth and preflight diagnostics

services/protocols/internal/infrastructure/postgres/auth_repository.go

@@ -21,7 +21,7 @@ func NewAuthRepository(pool *pgxpool.Pool) *AuthRepository {
 
 func (r *AuthRepository) FindByAddress(ctx context.Context, address string) (*port.MailboxAuth, error) {
 	row := r.pool.QueryRow(ctx, `
-		SELECT m.id, m.password_hash
+		SELECT m.id, m.password_hash, m.email_address, d.name, m.max_recipients_per_hour
 		FROM mailboxes m
 		JOIN domains d ON d.id = m.domain_id
 		WHERE m.email_address = $1
@@ -31,7 +31,7 @@ func (r *AuthRepository) FindByAddress(ctx context.Context, address string) (*po
 
 	var rec port.MailboxAuth
 	var id uuid.UUID
-	if err := row.Scan(&id, &rec.PasswordHash); err != nil {
+	if err := row.Scan(&id, &rec.PasswordHash, &rec.EmailAddress, &rec.DomainName, &rec.MaxRecipientsPerHour); err != nil {
 		if err == pgx.ErrNoRows {
 			return nil, nil
 		}

services/protocols/internal/infrastructure/postgres/outbound_repository.go

@@ -3,6 +3,9 @@ package postgres
 import (
 	"context"
 	"fmt"
+	"time"
+
+	"github.com/google/uuid"
 
 	"github.com/jackc/pgx/v5/pgxpool"
 
@@ -37,8 +40,13 @@ func (r *OutboundRepository) FetchNextReady(ctx context.Context, workerID string
 	rows, err := r.pool.Query(ctx, `
 		WITH ready_jobs AS (
 			SELECT id FROM outbound_jobs
-			WHERE status IN ('QUEUED', 'DEFERRED')
-			  AND next_attempt_at <= NOW()
+			WHERE (
+			        (status IN ('QUEUED', 'DEFERRED') AND next_attempt_at <= NOW())
+			        -- Reclaim jobs whose worker died mid-delivery. Without this
+			        -- they stay SENDING forever and the message is silently
+			        -- lost, which contradicts ADR-002's durability guarantee.
+			     OR (status = 'SENDING' AND locked_until IS NOT NULL AND locked_until <= NOW())
+			      )
 			ORDER BY next_attempt_at ASC
 			LIMIT $2
 			FOR UPDATE SKIP LOCKED
@@ -90,6 +98,22 @@ func (r *OutboundRepository) FetchNextReady(ctx context.Context, workerID string
 	return jobs, nil
 }
 
+// CountRecipientsSince counts queued recipients, not messages: one message to
+// fifty addresses is fifty recipients against the limit, which is the number
+// that matters for abuse.
+func (r *OutboundRepository) CountRecipientsSince(ctx context.Context, mailboxID uuid.UUID, since time.Time) (int, error) {
+	var count int
+	err := r.pool.QueryRow(ctx, `
+		SELECT COUNT(*) FROM outbound_jobs
+		WHERE mailbox_id = $1 AND created_at >= $2
+	`, mailboxID, since).Scan(&count)
+
+	if err != nil {
+		return 0, fmt.Errorf("count recent recipients for mailbox %s: %w", mailboxID, err)
+	}
+	return count, nil
+}
+
 func (r *OutboundRepository) UpdateJob(ctx context.Context, job *entity.OutboundJob) error {
 	_, err := r.pool.Exec(ctx, `
 		UPDATE outbound_jobs