Production hardening: security, resilience, observability, and compliance

Password complexity: custom validator requiring uppercase, lowercase, digit (min 8 chars)
Token expiry: 90-day token lifetime with refresh endpoint (60-90 day renewal window)
Health check: /api/health/ now pings Postgres + Redis, returns 503 on failure
Audit logging: async audit_log table for auth events (login, register, delete, etc.)
Circuit breaker: APNs/FCM push sends wrapped with 5-failure threshold, 30s recovery
FK indexes: 27 missing foreign key indexes across all tables (migration 017)
CSP header: default-src 'none'; frame-ancestors 'none'
Gzip compression: level 5 with media endpoint skipper
Prometheus metrics: /metrics endpoint using existing monitoring service
External timeouts: 15s push, 30s SMTP, context timeouts on all external calls

Migrations: 016 (token created_at), 017 (FK indexes), 018 (audit_log)
Tests: circuit breaker (15), audit service (8), token refresh (7), health (4),
       middleware expiry (5), validator (new)

internal/push/client.go

@@ -2,6 +2,7 @@ package push
 
 import (
 	"context"
+	"time"
 
 	"github.com/rs/zerolog/log"
 
@@ -14,16 +15,25 @@ const (
 	PlatformAndroid = "android"
 )
 
+// Timeout for individual push notification send operations.
+const pushSendTimeout = 15 * time.Second
+
 // Client provides a unified interface for sending push notifications
 type Client struct {
-	apns    *APNsClient
-	fcm     *FCMClient
-	enabled bool
+	apns       *APNsClient
+	fcm        *FCMClient
+	enabled    bool
+	apnsBreaker *CircuitBreaker
+	fcmBreaker  *CircuitBreaker
 }
 
 // NewClient creates a new unified push notification client
 func NewClient(cfg *config.PushConfig, enabled bool) (*Client, error) {
-	client := &Client{enabled: enabled}
+	client := &Client{
+		enabled:     enabled,
+		apnsBreaker: NewCircuitBreaker("apns"),
+		fcmBreaker:  NewCircuitBreaker("fcm"),
+	}
 
 	// Initialize APNs client (iOS)
 	if cfg.APNSKeyPath != "" && cfg.APNSKeyID != "" && cfg.APNSTeamID != "" {
@@ -54,7 +64,8 @@ func NewClient(cfg *config.PushConfig, enabled bool) (*Client, error) {
 	return client, nil
 }
 
-// SendToIOS sends a push notification to iOS devices
+// SendToIOS sends a push notification to iOS devices.
+// The call is guarded by a circuit breaker and uses a context timeout.
 func (c *Client) SendToIOS(ctx context.Context, tokens []string, title, message string, data map[string]string) error {
 	if !c.enabled {
 		log.Debug().Msg("Push notifications disabled by feature flag")
@@ -64,10 +75,26 @@ func (c *Client) SendToIOS(ctx context.Context, tokens []string, title, message
 		log.Warn().Msg("APNs client not initialized, skipping iOS push")
 		return nil
 	}
-	return c.apns.Send(ctx, tokens, title, message, data)
+	if !c.apnsBreaker.Allow() {
+		log.Warn().Str("breaker", c.apnsBreaker.Name()).Msg("APNs circuit breaker is open, skipping iOS push")
+		return ErrCircuitOpen
+	}
+
+	sendCtx, cancel := context.WithTimeout(ctx, pushSendTimeout)
+	defer cancel()
+
+	err := c.apns.Send(sendCtx, tokens, title, message, data)
+	if err != nil {
+		c.apnsBreaker.RecordFailure()
+		log.Warn().Err(err).Str("breaker_state", c.apnsBreaker.State()).Msg("APNs send failed, recorded circuit breaker failure")
+		return err
+	}
+	c.apnsBreaker.RecordSuccess()
+	return nil
 }
 
-// SendToAndroid sends a push notification to Android devices
+// SendToAndroid sends a push notification to Android devices.
+// The call is guarded by a circuit breaker and uses a context timeout.
 func (c *Client) SendToAndroid(ctx context.Context, tokens []string, title, message string, data map[string]string) error {
 	if !c.enabled {
 		log.Debug().Msg("Push notifications disabled by feature flag")
@@ -77,7 +104,22 @@ func (c *Client) SendToAndroid(ctx context.Context, tokens []string, title, mess
 		log.Warn().Msg("FCM client not initialized, skipping Android push")
 		return nil
 	}
-	return c.fcm.Send(ctx, tokens, title, message, data)
+	if !c.fcmBreaker.Allow() {
+		log.Warn().Str("breaker", c.fcmBreaker.Name()).Msg("FCM circuit breaker is open, skipping Android push")
+		return ErrCircuitOpen
+	}
+
+	sendCtx, cancel := context.WithTimeout(ctx, pushSendTimeout)
+	defer cancel()
+
+	err := c.fcm.Send(sendCtx, tokens, title, message, data)
+	if err != nil {
+		c.fcmBreaker.RecordFailure()
+		log.Warn().Err(err).Str("breaker_state", c.fcmBreaker.State()).Msg("FCM send failed, recorded circuit breaker failure")
+		return err
+	}
+	c.fcmBreaker.RecordSuccess()
+	return nil
 }
 
 // SendToAll sends a push notification to both iOS and Android devices
@@ -115,8 +157,9 @@ func (c *Client) IsAndroidEnabled() bool {
 	return c.fcm != nil
 }
 
-// SendActionableNotification sends notifications with action button support
-// iOS receives a category for actionable notifications, Android handles actions via data payload
+// SendActionableNotification sends notifications with action button support.
+// iOS receives a category for actionable notifications, Android handles actions via data payload.
+// Both platforms are guarded by their respective circuit breakers.
 func (c *Client) SendActionableNotification(ctx context.Context, iosTokens, androidTokens []string, title, message string, data map[string]string, iosCategoryID string) error {
 	if !c.enabled {
 		log.Debug().Msg("Push notifications disabled by feature flag")
@@ -127,10 +170,19 @@ func (c *Client) SendActionableNotification(ctx context.Context, iosTokens, andr
 	if len(iosTokens) > 0 {
 		if c.apns == nil {
 			log.Warn().Msg("APNs client not initialized, skipping iOS actionable push")
+		} else if !c.apnsBreaker.Allow() {
+			log.Warn().Str("breaker", c.apnsBreaker.Name()).Msg("APNs circuit breaker is open, skipping iOS actionable push")
+			lastErr = ErrCircuitOpen
 		} else {
-			if err := c.apns.SendWithCategory(ctx, iosTokens, title, message, data, iosCategoryID); err != nil {
+			sendCtx, cancel := context.WithTimeout(ctx, pushSendTimeout)
+			err := c.apns.SendWithCategory(sendCtx, iosTokens, title, message, data, iosCategoryID)
+			cancel()
+			if err != nil {
+				c.apnsBreaker.RecordFailure()
 				log.Error().Err(err).Msg("Failed to send iOS actionable notifications")
 				lastErr = err
+			} else {
+				c.apnsBreaker.RecordSuccess()
 			}
 		}
 	}