feat(attack-paths): merge master but UI

api/src/backend/api/db_router.py

@@ -26,6 +26,7 @@ class MainRouter:
     default_db = "default"
     admin_db = "admin"
     replica_db = "replica"
+    admin_replica_db = "admin_replica"
 
     def db_for_read(self, model, **hints):  # noqa: F841
         model_table_name = model._meta.db_table
@@ -49,7 +50,12 @@ class MainRouter:
 
     def allow_relation(self, obj1, obj2, **hints):  # noqa: F841
         # Allow relations when both objects originate from allowed connectors
-        allowed_dbs = {self.default_db, self.admin_db, self.replica_db}
+        allowed_dbs = {
+            self.default_db,
+            self.admin_db,
+            self.replica_db,
+            self.admin_replica_db,
+        }
         if {obj1._state.db, obj2._state.db} <= allowed_dbs:
             return True
         return None

api/src/backend/api/decorators.py

@@ -1,10 +1,14 @@
 import uuid
 from functools import wraps
 
-from django.db import connection, transaction
+from django.core.exceptions import ObjectDoesNotExist
+from django.db import IntegrityError, connection, transaction
 from rest_framework_json_api.serializers import ValidationError
 
-from api.db_utils import POSTGRES_TENANT_VAR, SET_CONFIG_QUERY
+from api.db_router import READ_REPLICA_ALIAS
+from api.db_utils import POSTGRES_TENANT_VAR, SET_CONFIG_QUERY, rls_transaction
+from api.exceptions import ProviderDeletedException
+from api.models import Provider, Scan
 
 
 def set_tenant(func=None, *, keep_tenant=False):
@@ -66,3 +70,49 @@ def set_tenant(func=None, *, keep_tenant=False):
         return decorator
     else:
         return decorator(func)
+
+
+def handle_provider_deletion(func):
+    """
+    Decorator that raises ProviderDeletedException if provider was deleted during execution.
+
+    Catches ObjectDoesNotExist and IntegrityError, checks if provider still exists,
+    and raises ProviderDeletedException if not. Otherwise, re-raises original exception.
+
+    Requires tenant_id and provider_id in kwargs.
+
+    Example:
+        @shared_task
+        @handle_provider_deletion
+        def scan_task(scan_id, tenant_id, provider_id):
+            ...
+    """
+
+    @wraps(func)
+    def wrapper(*args, **kwargs):
+        try:
+            return func(*args, **kwargs)
+        except (ObjectDoesNotExist, IntegrityError):
+            tenant_id = kwargs.get("tenant_id")
+            provider_id = kwargs.get("provider_id")
+
+            with rls_transaction(tenant_id, using=READ_REPLICA_ALIAS):
+                if provider_id is None:
+                    scan_id = kwargs.get("scan_id")
+                    if scan_id is None:
+                        raise AssertionError(
+                            "This task does not have provider or scan in the kwargs"
+                        )
+                    scan = Scan.objects.filter(pk=scan_id).first()
+                    if scan is None:
+                        raise ProviderDeletedException(
+                            f"Provider for scan '{scan_id}' was deleted during the scan"
+                        ) from None
+                    provider_id = str(scan.provider_id)
+                if not Provider.objects.filter(pk=provider_id).exists():
+                    raise ProviderDeletedException(
+                        f"Provider '{provider_id}' was deleted during the scan"
+                    ) from None
+            raise
+
+    return wrapper

api/src/backend/api/exceptions.py

@@ -66,6 +66,10 @@ class ProviderConnectionError(Exception):
     """Base exception for provider connection errors."""
 
 
+class ProviderDeletedException(Exception):
+    """Raised when a provider has been deleted during scan/task execution."""
+
+
 def custom_exception_handler(exc, context):
     if isinstance(exc, django_validation_error):
         if hasattr(exc, "error_dict"):

api/src/backend/api/filters.py

@@ -23,7 +23,9 @@ from api.db_utils import (
     StatusEnumField,
 )
 from api.models import (
+    AttackSurfaceOverview,
     ComplianceRequirementOverview,
+    DailySeveritySummary,
     Finding,
     Integration,
     Invitation,
@@ -36,12 +38,14 @@ from api.models import (
     PermissionChoices,
     Processor,
     Provider,
+    ProviderComplianceScore,
     ProviderGroup,
     ProviderSecret,
     Resource,
     ResourceTag,
     Role,
     Scan,
+    ScanCategorySummary,
     ScanSummary,
     SeverityChoices,
     StateChoices,
@@ -90,10 +94,62 @@ class ChoiceInFilter(BaseInFilter, ChoiceFilter):
     pass
 
 
+class BaseProviderFilter(FilterSet):
+    """
+    Abstract base filter for models with direct FK to Provider.
+
+    Provides standard provider_id and provider_type filters.
+    Subclasses must define Meta.model.
+    """
+
+    provider_id = UUIDFilter(field_name="provider__id", lookup_expr="exact")
+    provider_id__in = UUIDInFilter(field_name="provider__id", lookup_expr="in")
+    provider_type = ChoiceFilter(
+        field_name="provider__provider", choices=Provider.ProviderChoices.choices
+    )
+    provider_type__in = ChoiceInFilter(
+        field_name="provider__provider",
+        choices=Provider.ProviderChoices.choices,
+        lookup_expr="in",
+    )
+
+    class Meta:
+        abstract = True
+        fields = {}
+
+
+class BaseScanProviderFilter(FilterSet):
+    """
+    Abstract base filter for models with FK to Scan (and Scan has FK to Provider).
+
+    Provides standard provider_id and provider_type filters via scan relationship.
+    Subclasses must define Meta.model.
+    """
+
+    provider_id = UUIDFilter(field_name="scan__provider__id", lookup_expr="exact")
+    provider_id__in = UUIDInFilter(field_name="scan__provider__id", lookup_expr="in")
+    provider_type = ChoiceFilter(
+        field_name="scan__provider__provider", choices=Provider.ProviderChoices.choices
+    )
+    provider_type__in = ChoiceInFilter(
+        field_name="scan__provider__provider",
+        choices=Provider.ProviderChoices.choices,
+        lookup_expr="in",
+    )
+
+    class Meta:
+        abstract = True
+        fields = {}
+
+
 class CommonFindingFilters(FilterSet):
     # We filter providers from the scan in findings
+    # Both 'provider' and 'provider_id' parameters are supported for API consistency
+    # Frontend uses 'provider_id' uniformly across all endpoints
     provider = UUIDFilter(field_name="scan__provider__id", lookup_expr="exact")
     provider__in = UUIDInFilter(field_name="scan__provider__id", lookup_expr="in")
+    provider_id = UUIDFilter(field_name="scan__provider__id", lookup_expr="exact")
+    provider_id__in = UUIDInFilter(field_name="scan__provider__id", lookup_expr="in")
     provider_type = ChoiceFilter(
         choices=Provider.ProviderChoices.choices, field_name="scan__provider__provider"
     )
@@ -156,6 +212,9 @@ class CommonFindingFilters(FilterSet):
         field_name="resources__type", lookup_expr="icontains"
     )
 
+    category = CharFilter(method="filter_category")
+    category__in = CharInFilter(field_name="categories", lookup_expr="overlap")
+
     # Temporarily disabled until we implement tag filtering in the UI
     # resource_tag_key = CharFilter(field_name="resources__tags__key")
     # resource_tag_key__in = CharInFilter(
@@ -187,6 +246,9 @@ class CommonFindingFilters(FilterSet):
     def filter_resource_type(self, queryset, name, value):
         return queryset.filter(resource_types__contains=[value])
 
+    def filter_category(self, queryset, name, value):
+        return queryset.filter(categories__contains=[value])
+
     def filter_resource_tag(self, queryset, name, value):
         overall_query = Q()
         for key_value_pair in value:
@@ -778,7 +840,7 @@ class RoleFilter(FilterSet):
 
 class ComplianceOverviewFilter(FilterSet):
     inserted_at = DateFilter(field_name="inserted_at", lookup_expr="date")
-    scan_id = UUIDFilter(field_name="scan_id")
+    scan_id = UUIDFilter(field_name="scan_id", required=True)
     region = CharFilter(field_name="region")
 
     class Meta:
@@ -812,6 +874,68 @@ class ScanSummaryFilter(FilterSet):
         }
 
 
+class DailySeveritySummaryFilter(FilterSet):
+    """Filter for findings_severity/timeseries endpoint."""
+
+    MAX_DATE_RANGE_DAYS = 365
+
+    provider_id = UUIDFilter(field_name="provider_id", lookup_expr="exact")
+    provider_id__in = UUIDInFilter(field_name="provider_id", lookup_expr="in")
+    provider_type = ChoiceFilter(
+        field_name="provider__provider", choices=Provider.ProviderChoices.choices
+    )
+    provider_type__in = ChoiceInFilter(
+        field_name="provider__provider", choices=Provider.ProviderChoices.choices
+    )
+    date_from = DateFilter(method="filter_noop")
+    date_to = DateFilter(method="filter_noop")
+
+    class Meta:
+        model = DailySeveritySummary
+        fields = ["provider_id"]
+
+    def filter_noop(self, queryset, name, value):
+        return queryset
+
+    def filter_queryset(self, queryset):
+        if not self.data.get("date_from"):
+            raise ValidationError(
+                [
+                    {
+                        "detail": "This query parameter is required.",
+                        "status": "400",
+                        "source": {"pointer": "filter[date_from]"},
+                        "code": "required",
+                    }
+                ]
+            )
+
+        today = date.today()
+        date_from = self.form.cleaned_data.get("date_from")
+        date_to = min(self.form.cleaned_data.get("date_to") or today, today)
+
+        if (date_to - date_from).days > self.MAX_DATE_RANGE_DAYS:
+            raise ValidationError(
+                [
+                    {
+                        "detail": f"Date range cannot exceed {self.MAX_DATE_RANGE_DAYS} days.",
+                        "status": "400",
+                        "source": {"pointer": "filter[date_from]"},
+                        "code": "invalid",
+                    }
+                ]
+            )
+
+        # View access
+        self.request._date_from = date_from
+        self.request._date_to = date_to
+
+        # Apply date filter (only lte for fill-forward logic)
+        queryset = queryset.filter(date__lte=date_to)
+
+        return super().filter_queryset(queryset)
+
+
 class ScanSummarySeverityFilter(ScanSummaryFilter):
     """Filter for findings_severity ScanSummary endpoint - includes status filters"""
 
@@ -1031,3 +1155,27 @@ class ThreatScoreSnapshotFilter(FilterSet):
             "inserted_at": ["date", "gte", "lte"],
             "overall_score": ["exact", "gte", "lte"],
         }
+
+
+class AttackSurfaceOverviewFilter(BaseScanProviderFilter):
+    """Filter for attack surface overview aggregations by provider."""
+
+    class Meta(BaseScanProviderFilter.Meta):
+        model = AttackSurfaceOverview
+
+
+class CategoryOverviewFilter(BaseScanProviderFilter):
+    """Filter for category overview aggregations by provider."""
+
+    category = CharFilter(field_name="category", lookup_expr="exact")
+    category__in = CharInFilter(field_name="category", lookup_expr="in")
+
+    class Meta(BaseScanProviderFilter.Meta):
+        model = ScanCategorySummary
+
+
+class ComplianceWatchlistFilter(BaseProviderFilter):
+    """Filter for compliance watchlist overview by provider."""
+
+    class Meta(BaseProviderFilter.Meta):
+        model = ProviderComplianceScore

api/src/backend/api/migrations/0060_attack_surface_overview.py

@@ -0,0 +1,89 @@
+# Generated by Django 5.1.14 on 2025-11-19 13:03
+
+import uuid
+
+import django.db.models.deletion
+from django.db import migrations, models
+
+import api.rls
+
+
+class Migration(migrations.Migration):
+    dependencies = [
+        ("api", "0059_compliance_overview_summary"),
+    ]
+
+    operations = [
+        migrations.CreateModel(
+            name="AttackSurfaceOverview",
+            fields=[
+                (
+                    "id",
+                    models.UUIDField(
+                        default=uuid.uuid4,
+                        editable=False,
+                        primary_key=True,
+                        serialize=False,
+                    ),
+                ),
+                ("inserted_at", models.DateTimeField(auto_now_add=True)),
+                (
+                    "attack_surface_type",
+                    models.CharField(
+                        choices=[
+                            ("internet-exposed", "Internet Exposed"),
+                            ("secrets", "Exposed Secrets"),
+                            ("privilege-escalation", "Privilege Escalation"),
+                            ("ec2-imdsv1", "EC2 IMDSv1 Enabled"),
+                        ],
+                        max_length=50,
+                    ),
+                ),
+                ("total_findings", models.IntegerField(default=0)),
+                ("failed_findings", models.IntegerField(default=0)),
+                ("muted_failed_findings", models.IntegerField(default=0)),
+            ],
+            options={
+                "db_table": "attack_surface_overviews",
+                "abstract": False,
+            },
+        ),
+        migrations.AddField(
+            model_name="attacksurfaceoverview",
+            name="scan",
+            field=models.ForeignKey(
+                on_delete=django.db.models.deletion.CASCADE,
+                related_name="attack_surface_overviews",
+                related_query_name="attack_surface_overview",
+                to="api.scan",
+            ),
+        ),
+        migrations.AddField(
+            model_name="attacksurfaceoverview",
+            name="tenant",
+            field=models.ForeignKey(
+                on_delete=django.db.models.deletion.CASCADE, to="api.tenant"
+            ),
+        ),
+        migrations.AddIndex(
+            model_name="attacksurfaceoverview",
+            index=models.Index(
+                fields=["tenant_id", "scan_id"], name="attack_surf_tenant_scan_idx"
+            ),
+        ),
+        migrations.AddConstraint(
+            model_name="attacksurfaceoverview",
+            constraint=models.UniqueConstraint(
+                fields=("tenant_id", "scan_id", "attack_surface_type"),
+                name="unique_attack_surface_per_scan",
+            ),
+        ),
+        migrations.AddConstraint(
+            model_name="attacksurfaceoverview",
+            constraint=api.rls.RowLevelSecurityConstraint(
+                "tenant_id",
+                name="rls_on_attacksurfaceoverview",
+                statements=["SELECT", "INSERT", "UPDATE", "DELETE"],
+            ),
+        ),
+    ]

api/src/backend/api/migrations/0061_daily_severity_summary.py

@@ -0,0 +1,96 @@
+# Generated by Django 5.1.14 on 2025-12-03 13:38
+
+import uuid
+
+import django.db.models.deletion
+from django.db import migrations, models
+
+import api.rls
+
+
+class Migration(migrations.Migration):
+    dependencies = [
+        ("api", "0060_attack_surface_overview"),
+    ]
+
+    operations = [
+        migrations.CreateModel(
+            name="DailySeveritySummary",
+            fields=[
+                (
+                    "id",
+                    models.UUIDField(
+                        default=uuid.uuid4,
+                        editable=False,
+                        primary_key=True,
+                        serialize=False,
+                    ),
+                ),
+                ("date", models.DateField()),
+                ("critical", models.IntegerField(default=0)),
+                ("high", models.IntegerField(default=0)),
+                ("medium", models.IntegerField(default=0)),
+                ("low", models.IntegerField(default=0)),
+                ("informational", models.IntegerField(default=0)),
+                ("muted", models.IntegerField(default=0)),
+                (
+                    "provider",
+                    models.ForeignKey(
+                        on_delete=django.db.models.deletion.CASCADE,
+                        related_name="daily_severity_summaries",
+                        related_query_name="daily_severity_summary",
+                        to="api.provider",
+                    ),
+                ),
+                (
+                    "scan",
+                    models.ForeignKey(
+                        on_delete=django.db.models.deletion.CASCADE,
+                        related_name="daily_severity_summaries",
+                        related_query_name="daily_severity_summary",
+                        to="api.scan",
+                    ),
+                ),
+                (
+                    "tenant",
+                    models.ForeignKey(
+                        on_delete=django.db.models.deletion.CASCADE,
+                        to="api.tenant",
+                    ),
+                ),
+            ],
+            options={
+                "db_table": "daily_severity_summaries",
+                "abstract": False,
+            },
+        ),
+        migrations.AddIndex(
+            model_name="dailyseveritysummary",
+            index=models.Index(
+                fields=["tenant_id", "id"],
+                name="dss_tenant_id_idx",
+            ),
+        ),
+        migrations.AddIndex(
+            model_name="dailyseveritysummary",
+            index=models.Index(
+                fields=["tenant_id", "provider_id"],
+                name="dss_tenant_provider_idx",
+            ),
+        ),
+        migrations.AddConstraint(
+            model_name="dailyseveritysummary",
+            constraint=models.UniqueConstraint(
+                fields=("tenant_id", "provider", "date"),
+                name="unique_daily_severity_summary",
+            ),
+        ),
+        migrations.AddConstraint(
+            model_name="dailyseveritysummary",
+            constraint=api.rls.RowLevelSecurityConstraint(
+                "tenant_id",
+                name="rls_on_dailyseveritysummary",
+                statements=["SELECT", "INSERT", "UPDATE", "DELETE"],
+            ),
+        ),
+    ]

api/src/backend/api/migrations/0062_backfill_daily_severity_summaries.py

@@ -0,0 +1,30 @@
+# Generated by Django 5.1.14 on 2025-12-10
+
+from django.db import migrations
+from tasks.tasks import backfill_daily_severity_summaries_task
+
+from api.db_router import MainRouter
+from api.rls import Tenant
+
+
+def trigger_backfill_task(apps, schema_editor):
+    """
+    Trigger the backfill task for all tenants.
+
+    This dispatches backfill_daily_severity_summaries_task for each tenant
+    in the system to populate DailySeveritySummary records from historical scans.
+    """
+    tenant_ids = Tenant.objects.using(MainRouter.admin_db).values_list("id", flat=True)
+
+    for tenant_id in tenant_ids:
+        backfill_daily_severity_summaries_task.delay(tenant_id=str(tenant_id), days=90)
+
+
+class Migration(migrations.Migration):
+    dependencies = [
+        ("api", "0061_daily_severity_summary"),
+    ]
+
+    operations = [
+        migrations.RunPython(trigger_backfill_task, migrations.RunPython.noop),
+    ]

api/src/backend/api/migrations/0063_scan_category_summary.py

@@ -0,0 +1,111 @@
+import uuid
+
+import django.db.models.deletion
+from django.db import migrations, models
+
+import api.db_utils
+import api.rls
+
+
+class Migration(migrations.Migration):
+    dependencies = [
+        ("api", "0062_backfill_daily_severity_summaries"),
+    ]
+
+    operations = [
+        migrations.CreateModel(
+            name="ScanCategorySummary",
+            fields=[
+                (
+                    "id",
+                    models.UUIDField(
+                        default=uuid.uuid4,
+                        editable=False,
+                        primary_key=True,
+                        serialize=False,
+                    ),
+                ),
+                (
+                    "tenant",
+                    models.ForeignKey(
+                        on_delete=django.db.models.deletion.CASCADE,
+                        to="api.tenant",
+                    ),
+                ),
+                (
+                    "inserted_at",
+                    models.DateTimeField(auto_now_add=True),
+                ),
+                (
+                    "scan",
+                    models.ForeignKey(
+                        on_delete=django.db.models.deletion.CASCADE,
+                        related_name="category_summaries",
+                        related_query_name="category_summary",
+                        to="api.scan",
+                    ),
+                ),
+                (
+                    "category",
+                    models.CharField(max_length=100),
+                ),
+                (
+                    "severity",
+                    api.db_utils.SeverityEnumField(
+                        choices=[
+                            ("critical", "Critical"),
+                            ("high", "High"),
+                            ("medium", "Medium"),
+                            ("low", "Low"),
+                            ("informational", "Informational"),
+                        ],
+                    ),
+                ),
+                (
+                    "total_findings",
+                    models.IntegerField(
+                        default=0, help_text="Non-muted findings (PASS + FAIL)"
+                    ),
+                ),
+                (
+                    "failed_findings",
+                    models.IntegerField(
+                        default=0,
+                        help_text="Non-muted FAIL findings (subset of total_findings)",
+                    ),
+                ),
+                (
+                    "new_failed_findings",
+                    models.IntegerField(
+                        default=0,
+                        help_text="Non-muted FAIL with delta='new' (subset of failed_findings)",
+                    ),
+                ),
+            ],
+            options={
+                "db_table": "scan_category_summaries",
+                "abstract": False,
+            },
+        ),
+        migrations.AddIndex(
+            model_name="scancategorysummary",
+            index=models.Index(
+                fields=["tenant_id", "scan"], name="scs_tenant_scan_idx"
+            ),
+        ),
+        migrations.AddConstraint(
+            model_name="scancategorysummary",
+            constraint=models.UniqueConstraint(
+                fields=("tenant_id", "scan_id", "category", "severity"),
+                name="unique_category_severity_per_scan",
+            ),
+        ),
+        migrations.AddConstraint(
+            model_name="scancategorysummary",
+            constraint=api.rls.RowLevelSecurityConstraint(
+                field="tenant_id",
+                name="rls_on_scancategorysummary",
+                statements=["SELECT", "INSERT", "UPDATE", "DELETE"],
+            ),
+        ),
+    ]

api/src/backend/api/migrations/0064_finding_categories.py

@@ -0,0 +1,22 @@
+import django.contrib.postgres.fields
+from django.db import migrations, models
+
+
+class Migration(migrations.Migration):
+    dependencies = [
+        ("api", "0063_scan_category_summary"),
+    ]
+
+    operations = [
+        migrations.AddField(
+            model_name="finding",
+            name="categories",
+            field=django.contrib.postgres.fields.ArrayField(
+                base_field=models.CharField(max_length=100),
+                blank=True,
+                null=True,
+                size=None,
+                help_text="Categories from check metadata for efficient filtering",
+            ),
+        ),
+    ]

api/src/backend/api/migrations/0065_alibabacloud_provider.py

@@ -0,0 +1,37 @@
+# Generated by Django migration for Alibaba Cloud provider support
+
+from django.db import migrations
+
+import api.db_utils
+
+
+class Migration(migrations.Migration):
+    dependencies = [
+        ("api", "0064_finding_categories"),
+    ]
+
+    operations = [
+        migrations.AlterField(
+            model_name="provider",
+            name="provider",
+            field=api.db_utils.ProviderEnumField(
+                choices=[
+                    ("aws", "AWS"),
+                    ("azure", "Azure"),
+                    ("gcp", "GCP"),
+                    ("kubernetes", "Kubernetes"),
+                    ("m365", "M365"),
+                    ("github", "GitHub"),
+                    ("mongodbatlas", "MongoDB Atlas"),
+                    ("iac", "IaC"),
+                    ("oraclecloud", "Oracle Cloud Infrastructure"),
+                    ("alibabacloud", "Alibaba Cloud"),
+                ],
+                default="aws",
+            ),
+        ),
+        migrations.RunSQL(
+            "ALTER TYPE provider ADD VALUE IF NOT EXISTS 'alibabacloud';",
+            reverse_sql=migrations.RunSQL.noop,
+        ),
+    ]

api/src/backend/api/migrations/0066_provider_compliance_score.py

@@ -0,0 +1,94 @@
+import uuid
+
+import django.db.models.deletion
+from django.db import migrations, models
+
+import api.db_utils
+import api.rls
+
+
+class Migration(migrations.Migration):
+    dependencies = [
+        ("api", "0065_alibabacloud_provider"),
+    ]
+
+    operations = [
+        migrations.CreateModel(
+            name="ProviderComplianceScore",
+            fields=[
+                (
+                    "id",
+                    models.UUIDField(
+                        default=uuid.uuid4,
+                        editable=False,
+                        primary_key=True,
+                        serialize=False,
+                    ),
+                ),
+                ("compliance_id", models.TextField()),
+                ("requirement_id", models.TextField()),
+                (
+                    "requirement_status",
+                    api.db_utils.StatusEnumField(
+                        choices=[
+                            ("FAIL", "Fail"),
+                            ("PASS", "Pass"),
+                            ("MANUAL", "Manual"),
+                        ]
+                    ),
+                ),
+                ("scan_completed_at", models.DateTimeField()),
+                (
+                    "provider",
+                    models.ForeignKey(
+                        on_delete=django.db.models.deletion.CASCADE,
+                        related_name="compliance_scores",
+                        related_query_name="compliance_score",
+                        to="api.provider",
+                    ),
+                ),
+                (
+                    "scan",
+                    models.ForeignKey(
+                        on_delete=django.db.models.deletion.CASCADE,
+                        related_name="compliance_scores",
+                        related_query_name="compliance_score",
+                        to="api.scan",
+                    ),
+                ),
+                (
+                    "tenant",
+                    models.ForeignKey(
+                        on_delete=django.db.models.deletion.CASCADE,
+                        to="api.tenant",
+                    ),
+                ),
+            ],
+            options={
+                "db_table": "provider_compliance_scores",
+                "abstract": False,
+            },
+        ),
+        migrations.AddConstraint(
+            model_name="providercompliancescore",
+            constraint=models.UniqueConstraint(
+                fields=("tenant_id", "provider_id", "compliance_id", "requirement_id"),
+                name="unique_provider_compliance_req",
+            ),
+        ),
+        migrations.AddConstraint(
+            model_name="providercompliancescore",
+            constraint=api.rls.RowLevelSecurityConstraint(
+                "tenant_id",
+                name="rls_on_providercompliancescore",
+                statements=["SELECT", "INSERT", "UPDATE", "DELETE"],
+            ),
+        ),
+        migrations.AddIndex(
+            model_name="providercompliancescore",
+            index=models.Index(
+                fields=["tenant_id", "provider_id", "compliance_id"],
+                name="pcs_tenant_prov_comp_idx",
+            ),
+        ),
+    ]

api/src/backend/api/migrations/0067_tenant_compliance_summary.py

@@ -0,0 +1,61 @@
+import uuid
+
+import django.db.models.deletion
+from django.db import migrations, models
+
+import api.rls
+
+
+class Migration(migrations.Migration):
+    dependencies = [
+        ("api", "0066_provider_compliance_score"),
+    ]
+
+    operations = [
+        migrations.CreateModel(
+            name="TenantComplianceSummary",
+            fields=[
+                (
+                    "id",
+                    models.UUIDField(
+                        default=uuid.uuid4,
+                        editable=False,
+                        primary_key=True,
+                        serialize=False,
+                    ),
+                ),
+                ("compliance_id", models.TextField()),
+                ("requirements_passed", models.IntegerField(default=0)),
+                ("requirements_failed", models.IntegerField(default=0)),
+                ("requirements_manual", models.IntegerField(default=0)),
+                ("total_requirements", models.IntegerField(default=0)),
+                ("updated_at", models.DateTimeField(auto_now=True)),
+                (
+                    "tenant",
+                    models.ForeignKey(
+                        on_delete=django.db.models.deletion.CASCADE,
+                        to="api.tenant",
+                    ),
+                ),
+            ],
+            options={
+                "db_table": "tenant_compliance_summaries",
+                "abstract": False,
+            },
+        ),
+        migrations.AddConstraint(
+            model_name="tenantcompliancesummary",
+            constraint=models.UniqueConstraint(
+                fields=("tenant_id", "compliance_id"),
+                name="unique_tenant_compliance_summary",
+            ),
+        ),
+        migrations.AddConstraint(
+            model_name="tenantcompliancesummary",
+            constraint=api.rls.RowLevelSecurityConstraint(
+                "tenant_id",
+                name="rls_on_tenantcompliancesummary",
+                statements=["SELECT", "INSERT", "UPDATE", "DELETE"],
+            ),
+        ),
+    ]

api/src/backend/api/models.py

@@ -287,6 +287,7 @@ class Provider(RowLevelSecurityProtectedModel):
         MONGODBATLAS = "mongodbatlas", _("MongoDB Atlas")
         IAC = "iac", _("IaC")
         ORACLECLOUD = "oraclecloud", _("Oracle Cloud Infrastructure")
+        ALIBABACLOUD = "alibabacloud", _("Alibaba Cloud")
 
     @staticmethod
     def validate_aws_uid(value):
@@ -391,6 +392,15 @@ class Provider(RowLevelSecurityProtectedModel):
                 pointer="/data/attributes/uid",
             )
 
+    @staticmethod
+    def validate_alibabacloud_uid(value):
+        if not re.match(r"^\d{16}$", value):
+            raise ModelValidationError(
+                detail="Alibaba Cloud account ID must be exactly 16 digits.",
+                code="alibabacloud-uid",
+                pointer="/data/attributes/uid",
+            )
+
     id = models.UUIDField(primary_key=True, default=uuid4, editable=False)
     inserted_at = models.DateTimeField(auto_now_add=True, editable=False)
     updated_at = models.DateTimeField(auto_now=True, editable=False)
@@ -811,14 +821,19 @@ class Resource(RowLevelSecurityProtectedModel):
             self.clear_tags()
             return
 
-        # Add new relationships with the tenant_id field
+        # Add new relationships with the tenant_id field; avoid touching the
+        # Resource row unless a mapping is actually created to prevent noisy
+        # updates during scans.
+        mapping_created = False
         for tag in tags:
-            ResourceTagMapping.objects.update_or_create(
+            _, created = ResourceTagMapping.objects.update_or_create(
                 tag=tag, resource=self, tenant_id=self.tenant_id
             )
+            mapping_created = mapping_created or created
 
-        # Save the instance
-        self.save()
+        if mapping_created:
+            # Only bump updated_at when the tag set truly changed
+            self.save(update_fields=["updated_at"])
 
     class Meta(RowLevelSecurityProtectedModel.Meta):
         db_table = "resources"
@@ -963,6 +978,14 @@ class Finding(PostgresPartitionedModel, RowLevelSecurityProtectedModel):
         null=True,
     )
 
+    # Check metadata denormalization
+    categories = ArrayField(
+        models.CharField(max_length=100),
+        blank=True,
+        null=True,
+        help_text="Categories from check metadata for efficient filtering",
+    )
+
     # Relationships
     scan = models.ForeignKey(to=Scan, related_name="findings", on_delete=models.CASCADE)
 
@@ -1595,6 +1618,65 @@ class ScanSummary(RowLevelSecurityProtectedModel):
         resource_name = "scan-summaries"
 
 
+class DailySeveritySummary(RowLevelSecurityProtectedModel):
+    """
+    Pre-aggregated daily severity counts per provider.
+    Used by findings_severity/timeseries endpoint for efficient queries.
+    """
+
+    objects = ActiveProviderManager()
+
+    id = models.UUIDField(primary_key=True, default=uuid4, editable=False)
+    date = models.DateField()
+
+    provider = models.ForeignKey(
+        Provider,
+        on_delete=models.CASCADE,
+        related_name="daily_severity_summaries",
+        related_query_name="daily_severity_summary",
+    )
+    scan = models.ForeignKey(
+        Scan,
+        on_delete=models.CASCADE,
+        related_name="daily_severity_summaries",
+        related_query_name="daily_severity_summary",
+    )
+
+    # Aggregated fail counts by severity
+    critical = models.IntegerField(default=0)
+    high = models.IntegerField(default=0)
+    medium = models.IntegerField(default=0)
+    low = models.IntegerField(default=0)
+    informational = models.IntegerField(default=0)
+    muted = models.IntegerField(default=0)
+
+    class Meta(RowLevelSecurityProtectedModel.Meta):
+        db_table = "daily_severity_summaries"
+
+        constraints = [
+            models.UniqueConstraint(
+                fields=("tenant_id", "provider", "date"),
+                name="unique_daily_severity_summary",
+            ),
+            RowLevelSecurityConstraint(
+                field="tenant_id",
+                name="rls_on_%(class)s",
+                statements=["SELECT", "INSERT", "UPDATE", "DELETE"],
+            ),
+        ]
+
+        indexes = [
+            models.Index(
+                fields=["tenant_id", "id"],
+                name="dss_tenant_id_idx",
+            ),
+            models.Index(
+                fields=["tenant_id", "provider_id"],
+                name="dss_tenant_provider_idx",
+            ),
+        ]
+
+
 class Integration(RowLevelSecurityProtectedModel):
     class IntegrationChoices(models.TextChoices):
         AMAZON_S3 = "amazon_s3", _("Amazon S3")
@@ -1987,6 +2069,64 @@ class ResourceScanSummary(RowLevelSecurityProtectedModel):
         ]
 
 
+class ScanCategorySummary(RowLevelSecurityProtectedModel):
+    """
+    Pre-aggregated category metrics per scan by severity.
+
+    Stores one row per (category, severity) combination per scan for efficient
+    overview queries. Categories come from check_metadata.categories.
+
+    Count relationships (each is a subset of the previous):
+        - total_findings >= failed_findings >= new_failed_findings
+    """
+
+    id = models.UUIDField(primary_key=True, default=uuid4, editable=False)
+    inserted_at = models.DateTimeField(auto_now_add=True, editable=False)
+
+    scan = models.ForeignKey(
+        Scan,
+        on_delete=models.CASCADE,
+        related_name="category_summaries",
+        related_query_name="category_summary",
+    )
+
+    category = models.CharField(max_length=100)
+    severity = SeverityEnumField(choices=SeverityChoices)
+
+    total_findings = models.IntegerField(
+        default=0, help_text="Non-muted findings (PASS + FAIL)"
+    )
+    failed_findings = models.IntegerField(
+        default=0, help_text="Non-muted FAIL findings (subset of total_findings)"
+    )
+    new_failed_findings = models.IntegerField(
+        default=0,
+        help_text="Non-muted FAIL with delta='new' (subset of failed_findings)",
+    )
+
+    class Meta(RowLevelSecurityProtectedModel.Meta):
+        db_table = "scan_category_summaries"
+
+        indexes = [
+            models.Index(fields=["tenant_id", "scan"], name="scs_tenant_scan_idx"),
+        ]
+
+        constraints = [
+            models.UniqueConstraint(
+                fields=("tenant_id", "scan_id", "category", "severity"),
+                name="unique_category_severity_per_scan",
+            ),
+            RowLevelSecurityConstraint(
+                field="tenant_id",
+                name="rls_on_%(class)s",
+                statements=["SELECT", "INSERT", "UPDATE", "DELETE"],
+            ),
+        ]
+
+    class JSONAPIMeta:
+        resource_name = "scan-category-summaries"
+
+
 class LighthouseConfiguration(RowLevelSecurityProtectedModel):
     """
     Stores configuration and API keys for LLM services.
@@ -2500,3 +2640,152 @@ class ThreatScoreSnapshot(RowLevelSecurityProtectedModel):
 
     class JSONAPIMeta:
         resource_name = "threatscore-snapshots"
+
+
+class AttackSurfaceOverview(RowLevelSecurityProtectedModel):
+    """
+    Pre-aggregated attack surface metrics per scan.
+
+    Stores counts for each attack surface type (internet-exposed, secrets,
+    privilege-escalation, ec2-imdsv1) to enable fast overview queries.
+    """
+
+    class AttackSurfaceTypeChoices(models.TextChoices):
+        INTERNET_EXPOSED = "internet-exposed", _("Internet Exposed")
+        SECRETS = "secrets", _("Exposed Secrets")
+        PRIVILEGE_ESCALATION = "privilege-escalation", _("Privilege Escalation")
+        EC2_IMDSV1 = "ec2-imdsv1", _("EC2 IMDSv1 Enabled")
+
+    id = models.UUIDField(primary_key=True, default=uuid4, editable=False)
+    inserted_at = models.DateTimeField(auto_now_add=True, editable=False)
+
+    scan = models.ForeignKey(
+        Scan,
+        on_delete=models.CASCADE,
+        related_name="attack_surface_overviews",
+        related_query_name="attack_surface_overview",
+    )
+
+    attack_surface_type = models.CharField(
+        max_length=50,
+        choices=AttackSurfaceTypeChoices.choices,
+    )
+
+    # Finding counts
+    total_findings = models.IntegerField(default=0)  # All findings (PASS + FAIL)
+    failed_findings = models.IntegerField(default=0)  # Non-muted failed findings
+    muted_failed_findings = models.IntegerField(default=0)  # Muted failed findings
+
+    class Meta(RowLevelSecurityProtectedModel.Meta):
+        db_table = "attack_surface_overviews"
+
+        constraints = [
+            models.UniqueConstraint(
+                fields=("tenant_id", "scan_id", "attack_surface_type"),
+                name="unique_attack_surface_per_scan",
+            ),
+            RowLevelSecurityConstraint(
+                field="tenant_id",
+                name="rls_on_%(class)s",
+                statements=["SELECT", "INSERT", "UPDATE", "DELETE"],
+            ),
+        ]
+
+        indexes = [
+            models.Index(
+                fields=["tenant_id", "scan_id"],
+                name="attack_surf_tenant_scan_idx",
+            ),
+        ]
+
+    class JSONAPIMeta:
+        resource_name = "attack-surface-overviews"
+
+
+class ProviderComplianceScore(RowLevelSecurityProtectedModel):
+    """
+    Compliance requirement status from latest completed scan per provider.
+
+    Used for efficient compliance watchlist queries with FAIL-dominant aggregation
+    across multiple providers. Updated via atomic upsert after each scan completion.
+    """
+
+    id = models.UUIDField(primary_key=True, default=uuid4, editable=False)
+
+    scan = models.ForeignKey(
+        Scan,
+        on_delete=models.CASCADE,
+        related_name="compliance_scores",
+        related_query_name="compliance_score",
+    )
+
+    provider = models.ForeignKey(
+        Provider,
+        on_delete=models.CASCADE,
+        related_name="compliance_scores",
+        related_query_name="compliance_score",
+    )
+
+    compliance_id = models.TextField()
+    requirement_id = models.TextField()
+    requirement_status = StatusEnumField(choices=StatusChoices)
+
+    scan_completed_at = models.DateTimeField()
+
+    class Meta(RowLevelSecurityProtectedModel.Meta):
+        db_table = "provider_compliance_scores"
+
+        constraints = [
+            models.UniqueConstraint(
+                fields=("tenant_id", "provider_id", "compliance_id", "requirement_id"),
+                name="unique_provider_compliance_req",
+            ),
+            RowLevelSecurityConstraint(
+                field="tenant_id",
+                name="rls_on_%(class)s",
+                statements=["SELECT", "INSERT", "UPDATE", "DELETE"],
+            ),
+        ]
+
+        indexes = [
+            models.Index(
+                fields=["tenant_id", "provider_id", "compliance_id"],
+                name="pcs_tenant_prov_comp_idx",
+            ),
+        ]
+
+
+class TenantComplianceSummary(RowLevelSecurityProtectedModel):
+    """
+    Pre-aggregated compliance counts per tenant with FAIL-dominant logic applied.
+
+    One row per (tenant, compliance_id). Used for fast watchlist queries when
+    no provider filter is applied. Recalculated after each scan by aggregating
+    across all providers with FAIL-dominant logic at requirement level.
+    """
+
+    id = models.UUIDField(primary_key=True, default=uuid4, editable=False)
+
+    compliance_id = models.TextField()
+
+    requirements_passed = models.IntegerField(default=0)
+    requirements_failed = models.IntegerField(default=0)
+    requirements_manual = models.IntegerField(default=0)
+    total_requirements = models.IntegerField(default=0)
+
+    updated_at = models.DateTimeField(auto_now=True)
+
+    class Meta(RowLevelSecurityProtectedModel.Meta):
+        db_table = "tenant_compliance_summaries"
+
+        constraints = [
+            models.UniqueConstraint(
+                fields=("tenant_id", "compliance_id"),
+                name="unique_tenant_compliance_summary",
+            ),
+            RowLevelSecurityConstraint(
+                field="tenant_id",
+                name="rls_on_%(class)s",
+                statements=["SELECT", "INSERT", "UPDATE", "DELETE"],
+            ),
+        ]

api/src/backend/api/rbac/permissions.py

@@ -65,11 +65,11 @@ def get_providers(role: Role) -> QuerySet[Provider]:
         A QuerySet of Provider objects filtered by the role's provider groups.
         If the role has no provider groups, returns an empty queryset.
     """
-    tenant = role.tenant
+    tenant_id = role.tenant_id
     provider_groups = role.provider_groups.all()
     if not provider_groups.exists():
         return Provider.objects.none()
 
     return Provider.objects.filter(
-        tenant=tenant, provider_groups__in=provider_groups
+        tenant_id=tenant_id, provider_groups__in=provider_groups
     ).distinct()

api/src/backend/api/tests/test_decorators.py

@@ -2,9 +2,12 @@ import uuid
 from unittest.mock import call, patch
 
 import pytest
+from django.core.exceptions import ObjectDoesNotExist
+from django.db import IntegrityError
 
 from api.db_utils import POSTGRES_TENANT_VAR, SET_CONFIG_QUERY
-from api.decorators import set_tenant
+from api.decorators import handle_provider_deletion, set_tenant
+from api.exceptions import ProviderDeletedException
 
 
 @pytest.mark.django_db
@@ -34,3 +37,142 @@ class TestSetTenantDecorator:
 
         with pytest.raises(KeyError):
             random_func("test_arg")
+
+
+@pytest.mark.django_db
+class TestHandleProviderDeletionDecorator:
+    def test_success_no_exception(self, tenants_fixture, providers_fixture):
+        """Decorated function runs normally when no exception is raised."""
+        tenant = tenants_fixture[0]
+        provider = providers_fixture[0]
+
+        @handle_provider_deletion
+        def task_func(**kwargs):
+            return "success"
+
+        result = task_func(
+            tenant_id=str(tenant.id),
+            provider_id=str(provider.id),
+        )
+        assert result == "success"
+
+    @patch("api.decorators.rls_transaction")
+    @patch("api.decorators.Provider.objects.filter")
+    def test_provider_deleted_with_provider_id(
+        self, mock_filter, mock_rls, tenants_fixture
+    ):
+        """Raises ProviderDeletedException when provider_id provided and provider deleted."""
+        tenant = tenants_fixture[0]
+        deleted_provider_id = str(uuid.uuid4())
+
+        mock_rls.return_value.__enter__ = lambda s: None
+        mock_rls.return_value.__exit__ = lambda s, *args: None
+        mock_filter.return_value.exists.return_value = False
+
+        @handle_provider_deletion
+        def task_func(**kwargs):
+            raise ObjectDoesNotExist("Some object not found")
+
+        with pytest.raises(ProviderDeletedException) as exc_info:
+            task_func(tenant_id=str(tenant.id), provider_id=deleted_provider_id)
+
+        assert deleted_provider_id in str(exc_info.value)
+
+    @patch("api.decorators.rls_transaction")
+    @patch("api.decorators.Provider.objects.filter")
+    @patch("api.decorators.Scan.objects.filter")
+    def test_provider_deleted_with_scan_id(
+        self, mock_scan_filter, mock_provider_filter, mock_rls, tenants_fixture
+    ):
+        """Raises ProviderDeletedException when scan exists but provider deleted."""
+        tenant = tenants_fixture[0]
+        scan_id = str(uuid.uuid4())
+        provider_id = str(uuid.uuid4())
+
+        mock_rls.return_value.__enter__ = lambda s: None
+        mock_rls.return_value.__exit__ = lambda s, *args: None
+
+        mock_scan = type("MockScan", (), {"provider_id": provider_id})()
+        mock_scan_filter.return_value.first.return_value = mock_scan
+        mock_provider_filter.return_value.exists.return_value = False
+
+        @handle_provider_deletion
+        def task_func(**kwargs):
+            raise ObjectDoesNotExist("Some object not found")
+
+        with pytest.raises(ProviderDeletedException) as exc_info:
+            task_func(tenant_id=str(tenant.id), scan_id=scan_id)
+
+        assert provider_id in str(exc_info.value)
+
+    @patch("api.decorators.rls_transaction")
+    @patch("api.decorators.Scan.objects.filter")
+    def test_scan_deleted_cascade(self, mock_scan_filter, mock_rls, tenants_fixture):
+        """Raises ProviderDeletedException when scan was deleted (CASCADE from provider)."""
+        tenant = tenants_fixture[0]
+        scan_id = str(uuid.uuid4())
+
+        mock_rls.return_value.__enter__ = lambda s: None
+        mock_rls.return_value.__exit__ = lambda s, *args: None
+        mock_scan_filter.return_value.first.return_value = None
+
+        @handle_provider_deletion
+        def task_func(**kwargs):
+            raise ObjectDoesNotExist("Some object not found")
+
+        with pytest.raises(ProviderDeletedException) as exc_info:
+            task_func(tenant_id=str(tenant.id), scan_id=scan_id)
+
+        assert scan_id in str(exc_info.value)
+
+    @patch("api.decorators.rls_transaction")
+    @patch("api.decorators.Provider.objects.filter")
+    def test_provider_exists_reraises_original(
+        self, mock_filter, mock_rls, tenants_fixture, providers_fixture
+    ):
+        """Re-raises original exception when provider still exists."""
+        tenant = tenants_fixture[0]
+        provider = providers_fixture[0]
+
+        mock_rls.return_value.__enter__ = lambda s: None
+        mock_rls.return_value.__exit__ = lambda s, *args: None
+        mock_filter.return_value.exists.return_value = True
+
+        @handle_provider_deletion
+        def task_func(**kwargs):
+            raise ObjectDoesNotExist("Actual object missing")
+
+        with pytest.raises(ObjectDoesNotExist):
+            task_func(tenant_id=str(tenant.id), provider_id=str(provider.id))
+
+    @patch("api.decorators.rls_transaction")
+    @patch("api.decorators.Provider.objects.filter")
+    def test_integrity_error_provider_deleted(
+        self, mock_filter, mock_rls, tenants_fixture
+    ):
+        """Raises ProviderDeletedException on IntegrityError when provider deleted."""
+        tenant = tenants_fixture[0]
+        deleted_provider_id = str(uuid.uuid4())
+
+        mock_rls.return_value.__enter__ = lambda s: None
+        mock_rls.return_value.__exit__ = lambda s, *args: None
+        mock_filter.return_value.exists.return_value = False
+
+        @handle_provider_deletion
+        def task_func(**kwargs):
+            raise IntegrityError("FK constraint violation")
+
+        with pytest.raises(ProviderDeletedException):
+            task_func(tenant_id=str(tenant.id), provider_id=deleted_provider_id)
+
+    def test_missing_provider_and_scan_raises_assertion(self, tenants_fixture):
+        """Raises AssertionError when neither provider_id nor scan_id in kwargs."""
+
+        @handle_provider_deletion
+        def task_func(**kwargs):
+            raise ObjectDoesNotExist("Some object not found")
+
+        with pytest.raises(AssertionError) as exc_info:
+            task_func(tenant_id=str(tenants_fixture[0].id))
+
+        assert "provider or scan" in str(exc_info.value)

api/src/backend/api/tests/test_models.py

@@ -1,9 +1,21 @@
+from datetime import datetime, timezone
+
 import pytest
 from allauth.socialaccount.models import SocialApp
 from django.core.exceptions import ValidationError
+from django.db import IntegrityError
 
 from api.db_router import MainRouter
-from api.models import Resource, ResourceTag, SAMLConfiguration, SAMLDomainIndex
+from api.models import (
+    ProviderComplianceScore,
+    Resource,
+    ResourceTag,
+    SAMLConfiguration,
+    SAMLDomainIndex,
+    StateChoices,
+    StatusChoices,
+    TenantComplianceSummary,
+)
 
 
 @pytest.mark.django_db
@@ -324,3 +336,159 @@ class TestSAMLConfigurationModel:
         errors = exc_info.value.message_dict
         assert "metadata_xml" in errors
         assert "There is a problem with your metadata." in errors["metadata_xml"][0]
+
+
+@pytest.mark.django_db
+class TestProviderComplianceScoreModel:
+    def test_create_provider_compliance_score(self, providers_fixture, scans_fixture):
+        provider = providers_fixture[0]
+        scan = scans_fixture[0]
+        scan.completed_at = datetime.now(timezone.utc)
+        scan.save()
+
+        score = ProviderComplianceScore.objects.create(
+            tenant_id=provider.tenant_id,
+            provider=provider,
+            scan=scan,
+            compliance_id="aws_cis_2.0",
+            requirement_id="req_1",
+            requirement_status=StatusChoices.PASS,
+            scan_completed_at=scan.completed_at,
+        )
+
+        assert score.compliance_id == "aws_cis_2.0"
+        assert score.requirement_id == "req_1"
+        assert score.requirement_status == StatusChoices.PASS
+
+    def test_unique_constraint_per_provider_compliance_requirement(
+        self, providers_fixture, scans_fixture
+    ):
+        provider = providers_fixture[0]
+        scan = scans_fixture[0]
+        scan.completed_at = datetime.now(timezone.utc)
+        scan.save()
+
+        ProviderComplianceScore.objects.create(
+            tenant_id=provider.tenant_id,
+            provider=provider,
+            scan=scan,
+            compliance_id="aws_cis_2.0",
+            requirement_id="req_1",
+            requirement_status=StatusChoices.PASS,
+            scan_completed_at=scan.completed_at,
+        )
+
+        with pytest.raises(IntegrityError):
+            ProviderComplianceScore.objects.create(
+                tenant_id=provider.tenant_id,
+                provider=provider,
+                scan=scan,
+                compliance_id="aws_cis_2.0",
+                requirement_id="req_1",
+                requirement_status=StatusChoices.FAIL,
+                scan_completed_at=scan.completed_at,
+            )
+
+    def test_different_providers_same_requirement_allowed(
+        self, providers_fixture, scans_fixture
+    ):
+        provider1, provider2, *_ = providers_fixture
+        scan1 = scans_fixture[0]
+        scan1.completed_at = datetime.now(timezone.utc)
+        scan1.save()
+
+        scan2 = scans_fixture[2]
+        scan2.state = StateChoices.COMPLETED
+        scan2.completed_at = datetime.now(timezone.utc)
+        scan2.save()
+
+        score1 = ProviderComplianceScore.objects.create(
+            tenant_id=provider1.tenant_id,
+            provider=provider1,
+            scan=scan1,
+            compliance_id="aws_cis_2.0",
+            requirement_id="req_1",
+            requirement_status=StatusChoices.PASS,
+            scan_completed_at=scan1.completed_at,
+        )
+
+        score2 = ProviderComplianceScore.objects.create(
+            tenant_id=provider2.tenant_id,
+            provider=provider2,
+            scan=scan2,
+            compliance_id="aws_cis_2.0",
+            requirement_id="req_1",
+            requirement_status=StatusChoices.FAIL,
+            scan_completed_at=scan2.completed_at,
+        )
+
+        assert score1.id != score2.id
+        assert score1.requirement_status != score2.requirement_status
+
+
+@pytest.mark.django_db
+class TestTenantComplianceSummaryModel:
+    def test_create_tenant_compliance_summary(self, tenants_fixture):
+        tenant = tenants_fixture[0]
+
+        summary = TenantComplianceSummary.objects.create(
+            tenant_id=tenant.id,
+            compliance_id="aws_cis_2.0",
+            requirements_passed=5,
+            requirements_failed=2,
+            requirements_manual=1,
+            total_requirements=8,
+        )
+
+        assert summary.compliance_id == "aws_cis_2.0"
+        assert summary.requirements_passed == 5
+        assert summary.requirements_failed == 2
+        assert summary.requirements_manual == 1
+        assert summary.total_requirements == 8
+        assert summary.updated_at is not None
+
+    def test_unique_constraint_per_tenant_compliance(self, tenants_fixture):
+        tenant = tenants_fixture[0]
+
+        TenantComplianceSummary.objects.create(
+            tenant_id=tenant.id,
+            compliance_id="aws_cis_2.0",
+            requirements_passed=5,
+            requirements_failed=2,
+            requirements_manual=1,
+            total_requirements=8,
+        )
+
+        with pytest.raises(IntegrityError):
+            TenantComplianceSummary.objects.create(
+                tenant_id=tenant.id,
+                compliance_id="aws_cis_2.0",
+                requirements_passed=3,
+                requirements_failed=4,
+                requirements_manual=1,
+                total_requirements=8,
+            )
+
+    def test_different_tenants_same_compliance_allowed(self, tenants_fixture):
+        tenant1, tenant2, *_ = tenants_fixture
+
+        summary1 = TenantComplianceSummary.objects.create(
+            tenant_id=tenant1.id,
+            compliance_id="aws_cis_2.0",
+            requirements_passed=5,
+            requirements_failed=2,
+            requirements_manual=1,
+            total_requirements=8,
+        )
+
+        summary2 = TenantComplianceSummary.objects.create(
+            tenant_id=tenant2.id,
+            compliance_id="aws_cis_2.0",
+            requirements_passed=3,
+            requirements_failed=4,
+            requirements_manual=1,
+            total_requirements=8,
+        )
+
+        assert summary1.id != summary2.id
+        assert summary1.requirements_passed != summary2.requirements_passed

api/src/backend/api/tests/test_utils.py

@@ -16,6 +16,7 @@ from api.utils import (
     return_prowler_provider,
     validate_invitation,
 )
+from prowler.providers.alibabacloud.alibabacloud_provider import AlibabacloudProvider
 from prowler.providers.aws.aws_provider import AwsProvider
 from prowler.providers.aws.lib.security_hub.security_hub import SecurityHubConnection
 from prowler.providers.azure.azure_provider import AzureProvider
@@ -116,6 +117,7 @@ class TestReturnProwlerProvider:
             (Provider.ProviderChoices.MONGODBATLAS.value, MongodbatlasProvider),
             (Provider.ProviderChoices.ORACLECLOUD.value, OraclecloudProvider),
             (Provider.ProviderChoices.IAC.value, IacProvider),
+            (Provider.ProviderChoices.ALIBABACLOUD.value, AlibabacloudProvider),
         ],
     )
     def test_return_prowler_provider(self, provider_type, expected_provider):

api/src/backend/api/utils.py

@@ -11,6 +11,7 @@ from api.exceptions import InvitationTokenExpiredException
 from api.models import Integration, Invitation, Processor, Provider, Resource
 from api.v1.serializers import FindingMetadataSerializer
 from prowler.lib.outputs.jira.jira import Jira, JiraBasicAuthError
+from prowler.providers.alibabacloud.alibabacloud_provider import AlibabacloudProvider
 from prowler.providers.aws.aws_provider import AwsProvider
 from prowler.providers.aws.lib.s3.s3 import S3
 from prowler.providers.aws.lib.security_hub.security_hub import SecurityHub
@@ -63,8 +64,9 @@ def merge_dicts(default_dict: dict, replacement_dict: dict) -> dict:
 
 def return_prowler_provider(
     provider: Provider,
-) -> [
-    AwsProvider
+) -> (
+    AlibabacloudProvider
+    | AwsProvider
     | AzureProvider
     | GcpProvider
     | GithubProvider
@@ -73,14 +75,14 @@ def return_prowler_provider(
     | M365Provider
     | MongodbatlasProvider
     | OraclecloudProvider
-]:
+):
     """Return the Prowler provider class based on the given provider type.
 
     Args:
         provider (Provider): The provider object containing the provider type and associated secrets.
 
     Returns:
-        AwsProvider | AzureProvider | GcpProvider | GithubProvider | IacProvider | KubernetesProvider | M365Provider | OraclecloudProvider | MongodbatlasProvider: The corresponding provider class.
+        AlibabacloudProvider | AwsProvider | AzureProvider | GcpProvider | GithubProvider | IacProvider | KubernetesProvider | M365Provider | MongodbatlasProvider | OraclecloudProvider: The corresponding provider class.
 
     Raises:
         ValueError: If the provider type specified in `provider.provider` is not supported.
@@ -104,6 +106,8 @@ def return_prowler_provider(
             prowler_provider = IacProvider
         case Provider.ProviderChoices.ORACLECLOUD.value:
             prowler_provider = OraclecloudProvider
+        case Provider.ProviderChoices.ALIBABACLOUD.value:
+            prowler_provider = AlibabacloudProvider
         case _:
             raise ValueError(f"Provider type {provider.provider} not supported")
     return prowler_provider
@@ -169,7 +173,8 @@ def initialize_prowler_provider(
     provider: Provider,
     mutelist_processor: Processor | None = None,
 ) -> (
-    AwsProvider
+    AlibabacloudProvider
+    | AwsProvider
     | AzureProvider
     | GcpProvider
     | GithubProvider
@@ -186,9 +191,8 @@ def initialize_prowler_provider(
         mutelist_processor (Processor): The mutelist processor object containing the mutelist configuration.
 
     Returns:
-        AwsProvider | AzureProvider | GcpProvider | GithubProvider | IacProvider | KubernetesProvider | M365Provider | OraclecloudProvider | MongodbatlasProvider: An instance of the corresponding provider class
-            (`AwsProvider`, `AzureProvider`, `GcpProvider`, `GithubProvider`, `IacProvider`, `KubernetesProvider`, `M365Provider`, `OraclecloudProvider` or `MongodbatlasProvider`) initialized with the
-            provider's secrets.
+        AlibabacloudProvider | AwsProvider | AzureProvider | GcpProvider | GithubProvider | IacProvider | KubernetesProvider | M365Provider | MongodbatlasProvider | OraclecloudProvider: An instance of the corresponding provider class
+            initialized with the provider's secrets.
     """
     prowler_provider = return_prowler_provider(provider)
     prowler_provider_kwargs = get_prowler_provider_kwargs(provider, mutelist_processor)
@@ -382,10 +386,18 @@ def get_findings_metadata_no_aggregations(tenant_id: str, filtered_queryset):
     regions = sorted({region for region in aggregation["regions"] or [] if region})
     resource_types = sorted(set(aggregation["resource_types"] or []))
 
+    # Aggregate categories from findings
+    categories_set = set()
+    for categories_list in filtered_queryset.values_list("categories", flat=True):
+        if categories_list:
+            categories_set.update(categories_list)
+    categories = sorted(categories_set)
+
     result = {
         "services": services,
         "regions": regions,
         "resource_types": resource_types,
+        "categories": categories,
     }
 
     serializer = FindingMetadataSerializer(data=result)

api/src/backend/api/v1/serializer_utils/lighthouse.py

@@ -40,11 +40,16 @@ class BedrockCredentialsSerializer(serializers.Serializer):
     """
     Serializer for AWS Bedrock credentials validation.
 
-    Validates long-term AWS credentials (AKIA) and region format.
+    Supports two authentication methods:
+    1. AWS access key + secret key
+    2. Bedrock API key (bearer token)
+
+    In both cases, region is mandatory.
     """
 
-    access_key_id = serializers.CharField()
-    secret_access_key = serializers.CharField()
+    access_key_id = serializers.CharField(required=False, allow_blank=False)
+    secret_access_key = serializers.CharField(required=False, allow_blank=False)
+    api_key = serializers.CharField(required=False, allow_blank=False)
     region = serializers.CharField()
 
     def validate_access_key_id(self, value: str) -> str:
@@ -65,6 +70,15 @@ class BedrockCredentialsSerializer(serializers.Serializer):
             )
         return value
 
+    def validate_api_key(self, value: str) -> str:
+        """
+        Validate Bedrock API key (bearer token).
+        """
+        pattern = r"^ABSKQmVkcm9ja0FQSUtleS[A-Za-z0-9+/=]{110}$"
+        if not re.match(pattern, value or ""):
+            raise serializers.ValidationError("Invalid Bedrock API key format.")
+        return value
+
     def validate_region(self, value: str) -> str:
         """Validate AWS region format."""
         pattern = r"^[a-z]{2}-[a-z]+-\d+$"
@@ -74,6 +88,50 @@ class BedrockCredentialsSerializer(serializers.Serializer):
             )
         return value
 
+    def validate(self, attrs):
+        """
+        Enforce either:
+        - access_key_id + secret_access_key + region
+        OR
+        - api_key + region
+        """
+        access_key_id = attrs.get("access_key_id")
+        secret_access_key = attrs.get("secret_access_key")
+        api_key = attrs.get("api_key")
+        region = attrs.get("region")
+
+        errors = {}
+
+        if not region:
+            errors["region"] = ["Region is required."]
+
+        using_access_keys = bool(access_key_id or secret_access_key)
+        using_api_key = api_key is not None and api_key != ""
+
+        if using_access_keys and using_api_key:
+            errors["non_field_errors"] = [
+                "Provide either access key + secret key OR api key, not both."
+            ]
+        elif not using_access_keys and not using_api_key:
+            errors["non_field_errors"] = [
+                "You must provide either access key + secret key OR api key."
+            ]
+        elif using_access_keys:
+            # Both access_key_id and secret_access_key must be present together
+            if not access_key_id:
+                errors.setdefault("access_key_id", []).append(
+                    "AWS access key ID is required when using access key authentication."
+                )
+            if not secret_access_key:
+                errors.setdefault("secret_access_key", []).append(
+                    "AWS secret access key is required when using access key authentication."
+                )
+
+        if errors:
+            raise serializers.ValidationError(errors)
+
+        return attrs
+
     def to_internal_value(self, data):
         """Check for unknown fields before DRF filters them out."""
         if not isinstance(data, dict):
@@ -111,6 +169,15 @@ class BedrockCredentialsUpdateSerializer(BedrockCredentialsSerializer):
         for field in self.fields.values():
             field.required = False
 
+    def validate(self, attrs):
+        """
+        For updates, this serializer only checks individual fields.
+        It does NOT enforce the "either access keys OR api key" rule.
+        That rule is applied later, after merging with existing stored
+        credentials, in LighthouseProviderConfigUpdateSerializer.
+        """
+        return attrs
+
 
 class OpenAICompatibleCredentialsSerializer(serializers.Serializer):
     """
@@ -168,27 +235,51 @@ class OpenAICompatibleCredentialsSerializer(serializers.Serializer):
                 "required": ["api_key"],
             },
             {
-                "type": "object",
                 "title": "AWS Bedrock Credentials",
-                "properties": {
-                    "access_key_id": {
-                        "type": "string",
-                        "description": "AWS access key ID.",
-                        "pattern": "^AKIA[0-9A-Z]{16}$",
+                "oneOf": [
+                    {
+                        "title": "IAM Access Key Pair",
+                        "type": "object",
+                        "description": "Authenticate with AWS access key and secret key. Recommended when you manage IAM users or roles.",
+                        "properties": {
+                            "access_key_id": {
+                                "type": "string",
+                                "description": "AWS access key ID.",
+                                "pattern": "^AKIA[0-9A-Z]{16}$",
+                            },
+                            "secret_access_key": {
+                                "type": "string",
+                                "description": "AWS secret access key.",
+                                "pattern": "^[A-Za-z0-9/+=]{40}$",
+                            },
+                            "region": {
+                                "type": "string",
+                                "description": "AWS region identifier where Bedrock is available. Examples: us-east-1, "
+                                "us-west-2, eu-west-1, ap-northeast-1.",
+                                "pattern": "^[a-z]{2}-[a-z]+-\\d+$",
+                            },
+                        },
+                        "required": ["access_key_id", "secret_access_key", "region"],
                     },
-                    "secret_access_key": {
-                        "type": "string",
-                        "description": "AWS secret access key.",
-                        "pattern": "^[A-Za-z0-9/+=]{40}$",
+                    {
+                        "title": "Amazon Bedrock API Key",
+                        "type": "object",
+                        "description": "Authenticate with an Amazon Bedrock API key (bearer token). Region is still required.",
+                        "properties": {
+                            "api_key": {
+                                "type": "string",
+                                "description": "Amazon Bedrock API key (bearer token).",
+                            },
+                            "region": {
+                                "type": "string",
+                                "description": "AWS region identifier where Bedrock is available. Examples: us-east-1, "
+                                "us-west-2, eu-west-1, ap-northeast-1.",
+                                "pattern": "^[a-z]{2}-[a-z]+-\\d+$",
+                            },
+                        },
+                        "required": ["api_key", "region"],
                     },
-                    "region": {
-                        "type": "string",
-                        "description": "AWS region identifier where Bedrock is available. Examples: us-east-1, "
-                        "us-west-2, eu-west-1, ap-northeast-1.",
-                        "pattern": "^[a-z]{2}-[a-z]+-\\d+$",
-                    },
-                },
-                "required": ["access_key_id", "secret_access_key", "region"],
+                ],
             },
             {
                 "type": "object",

api/src/backend/api/v1/serializer_utils/providers.py

@@ -304,6 +304,48 @@ from rest_framework_json_api import serializers
                 },
                 "required": ["atlas_public_key", "atlas_private_key"],
             },
+            {
+                "type": "object",
+                "title": "Alibaba Cloud Static Credentials",
+                "properties": {
+                    "access_key_id": {
+                        "type": "string",
+                        "description": "The Alibaba Cloud access key ID for authentication.",
+                    },
+                    "access_key_secret": {
+                        "type": "string",
+                        "description": "The Alibaba Cloud access key secret for authentication.",
+                    },
+                    "security_token": {
+                        "type": "string",
+                        "description": "The STS security token for temporary credentials (optional).",
+                    },
+                },
+                "required": ["access_key_id", "access_key_secret"],
+            },
+            {
+                "type": "object",
+                "title": "Alibaba Cloud RAM Role Assumption",
+                "properties": {
+                    "role_arn": {
+                        "type": "string",
+                        "description": "The ARN of the RAM role to assume (e.g., acs:ram::1234567890123456:role/ProwlerRole).",
+                    },
+                    "access_key_id": {
+                        "type": "string",
+                        "description": "The Alibaba Cloud access key ID of the RAM user that will assume the role.",
+                    },
+                    "access_key_secret": {
+                        "type": "string",
+                        "description": "The Alibaba Cloud access key secret of the RAM user that will assume the role.",
+                    },
+                    "role_session_name": {
+                        "type": "string",
+                        "description": "An identifier for the role session (optional, defaults to 'ProwlerSession').",
+                    },
+                },
+                "required": ["role_arn", "access_key_id", "access_key_secret"],
+            },
         ]
     }
 )

api/src/backend/api/v1/serializers.py

@@ -73,6 +73,42 @@ from api.v1.serializer_utils.processors import ProcessorConfigField
 from api.v1.serializer_utils.providers import ProviderSecretField
 from prowler.lib.mutelist.mutelist import Mutelist
 
+# Base
+
+
+class BaseModelSerializerV1(serializers.ModelSerializer):
+    def get_root_meta(self, _resource, _many):
+        return {"version": "v1"}
+
+
+class BaseSerializerV1(serializers.Serializer):
+    def get_root_meta(self, _resource, _many):
+        return {"version": "v1"}
+
+
+class BaseWriteSerializer(BaseModelSerializerV1):
+    def validate(self, data):
+        if hasattr(self, "initial_data"):
+            initial_data = set(self.initial_data.keys()) - {"id", "type"}
+            unknown_keys = initial_data - set(self.fields.keys())
+            if unknown_keys:
+                raise ValidationError(f"Invalid fields: {unknown_keys}")
+        return data
+
+
+class RLSSerializer(BaseModelSerializerV1):
+    def create(self, validated_data):
+        tenant_id = self.context.get("tenant_id")
+        validated_data["tenant_id"] = tenant_id
+        return super().create(validated_data)
+
+
+class StateEnumSerializerField(serializers.ChoiceField):
+    def __init__(self, **kwargs):
+        kwargs["choices"] = StateChoices.choices
+        super().__init__(**kwargs)
+
+
 # Tokens
 
 
@@ -180,7 +216,7 @@ class TokenSocialLoginSerializer(BaseTokenSerializer):
 
 
 # TODO: Check if we can change the parent class to TokenRefreshSerializer from rest_framework_simplejwt.serializers
-class TokenRefreshSerializer(serializers.Serializer):
+class TokenRefreshSerializer(BaseSerializerV1):
     refresh = serializers.CharField()
 
     # Output token
@@ -214,7 +250,7 @@ class TokenRefreshSerializer(serializers.Serializer):
             raise ValidationError({"refresh": "Invalid or expired token"})
 
 
-class TokenSwitchTenantSerializer(serializers.Serializer):
+class TokenSwitchTenantSerializer(BaseSerializerV1):
     tenant_id = serializers.UUIDField(
         write_only=True, help_text="The tenant ID for which to request a new token."
     )
@@ -238,41 +274,10 @@ class TokenSwitchTenantSerializer(serializers.Serializer):
         return generate_tokens(user, tenant_id)
 
 
-# Base
-
-
-class BaseSerializerV1(serializers.ModelSerializer):
-    def get_root_meta(self, _resource, _many):
-        return {"version": "v1"}
-
-
-class BaseWriteSerializer(BaseSerializerV1):
-    def validate(self, data):
-        if hasattr(self, "initial_data"):
-            initial_data = set(self.initial_data.keys()) - {"id", "type"}
-            unknown_keys = initial_data - set(self.fields.keys())
-            if unknown_keys:
-                raise ValidationError(f"Invalid fields: {unknown_keys}")
-        return data
-
-
-class RLSSerializer(BaseSerializerV1):
-    def create(self, validated_data):
-        tenant_id = self.context.get("tenant_id")
-        validated_data["tenant_id"] = tenant_id
-        return super().create(validated_data)
-
-
-class StateEnumSerializerField(serializers.ChoiceField):
-    def __init__(self, **kwargs):
-        kwargs["choices"] = StateChoices.choices
-        super().__init__(**kwargs)
-
-
 # Users
 
 
-class UserSerializer(BaseSerializerV1):
+class UserSerializer(BaseModelSerializerV1):
     """
     Serializer for the User model.
     """
@@ -403,7 +408,7 @@ class UserUpdateSerializer(BaseWriteSerializer):
         return super().update(instance, validated_data)
 
 
-class RoleResourceIdentifierSerializer(serializers.Serializer):
+class RoleResourceIdentifierSerializer(BaseSerializerV1):
     resource_type = serializers.CharField(source="type")
     id = serializers.UUIDField()
 
@@ -586,7 +591,7 @@ class TaskSerializer(RLSSerializer, TaskBase):
 # Tenants
 
 
-class TenantSerializer(BaseSerializerV1):
+class TenantSerializer(BaseModelSerializerV1):
     """
     Serializer for the Tenant model.
     """
@@ -598,7 +603,7 @@ class TenantSerializer(BaseSerializerV1):
         fields = ["id", "name", "memberships"]
 
 
-class TenantIncludeSerializer(BaseSerializerV1):
+class TenantIncludeSerializer(BaseModelSerializerV1):
     class Meta:
         model = Tenant
         fields = ["id", "name"]
@@ -774,7 +779,7 @@ class ProviderGroupUpdateSerializer(ProviderGroupSerializer):
         return super().update(instance, validated_data)
 
 
-class ProviderResourceIdentifierSerializer(serializers.Serializer):
+class ProviderResourceIdentifierSerializer(BaseSerializerV1):
     resource_type = serializers.CharField(source="type")
     id = serializers.UUIDField()
 
@@ -1111,7 +1116,7 @@ class ScanTaskSerializer(RLSSerializer):
         ]
 
 
-class ScanReportSerializer(serializers.Serializer):
+class ScanReportSerializer(BaseSerializerV1):
     id = serializers.CharField(source="scan")
 
     class Meta:
@@ -1119,7 +1124,7 @@ class ScanReportSerializer(serializers.Serializer):
         fields = ["id"]
 
 
-class ScanComplianceReportSerializer(serializers.Serializer):
+class ScanComplianceReportSerializer(BaseSerializerV1):
     id = serializers.CharField(source="scan")
     name = serializers.CharField()
 
@@ -1371,7 +1376,7 @@ class ResourceIncludeSerializer(RLSSerializer):
         return fields
 
 
-class ResourceMetadataSerializer(serializers.Serializer):
+class ResourceMetadataSerializer(BaseSerializerV1):
     services = serializers.ListField(child=serializers.CharField(), allow_empty=True)
     regions = serializers.ListField(child=serializers.CharField(), allow_empty=True)
     types = serializers.ListField(child=serializers.CharField(), allow_empty=True)
@@ -1400,6 +1405,7 @@ class FindingSerializer(RLSSerializer):
             "severity",
             "check_id",
             "check_metadata",
+            "categories",
             "raw_result",
             "inserted_at",
             "updated_at",
@@ -1441,7 +1447,7 @@ class FindingIncludeSerializer(RLSSerializer):
 
 
 # To be removed when the related endpoint is removed as well
-class FindingDynamicFilterSerializer(serializers.Serializer):
+class FindingDynamicFilterSerializer(BaseSerializerV1):
     services = serializers.ListField(child=serializers.CharField(), allow_empty=True)
     regions = serializers.ListField(child=serializers.CharField(), allow_empty=True)
 
@@ -1449,12 +1455,13 @@ class FindingDynamicFilterSerializer(serializers.Serializer):
         resource_name = "finding-dynamic-filters"
 
 
-class FindingMetadataSerializer(serializers.Serializer):
+class FindingMetadataSerializer(BaseSerializerV1):
     services = serializers.ListField(child=serializers.CharField(), allow_empty=True)
     regions = serializers.ListField(child=serializers.CharField(), allow_empty=True)
     resource_types = serializers.ListField(
         child=serializers.CharField(), allow_empty=True
     )
+    categories = serializers.ListField(child=serializers.CharField(), allow_empty=True)
     # Temporarily disabled until we implement tag filtering in the UI
     # tags = serializers.JSONField(help_text="Tags are described as key-value pairs.")
 
@@ -1487,12 +1494,23 @@ class BaseWriteProviderSecretSerializer(BaseWriteSerializer):
                 serializer = OracleCloudProviderSecret(data=secret)
             elif provider_type == Provider.ProviderChoices.MONGODBATLAS.value:
                 serializer = MongoDBAtlasProviderSecret(data=secret)
+            elif provider_type == Provider.ProviderChoices.ALIBABACLOUD.value:
+                serializer = AlibabaCloudProviderSecret(data=secret)
             else:
                 raise serializers.ValidationError(
                     {"provider": f"Provider type not supported {provider_type}"}
                 )
         elif secret_type == ProviderSecret.TypeChoices.ROLE:
-            serializer = AWSRoleAssumptionProviderSecret(data=secret)
+            if provider_type == Provider.ProviderChoices.AWS.value:
+                serializer = AWSRoleAssumptionProviderSecret(data=secret)
+            elif provider_type == Provider.ProviderChoices.ALIBABACLOUD.value:
+                serializer = AlibabaCloudRoleAssumptionProviderSecret(data=secret)
+            else:
+                raise serializers.ValidationError(
+                    {
+                        "secret_type": f"Role assumption not supported for provider type: {provider_type}"
+                    }
+                )
         elif secret_type == ProviderSecret.TypeChoices.SERVICE_ACCOUNT:
             serializer = GCPServiceAccountProviderSecret(data=secret)
         else:
@@ -1629,6 +1647,34 @@ class OracleCloudProviderSecret(serializers.Serializer):
         resource_name = "provider-secrets"
 
 
+class AlibabaCloudProviderSecret(serializers.Serializer):
+    access_key_id = serializers.CharField()
+    access_key_secret = serializers.CharField()
+    security_token = serializers.CharField(required=False)
+
+    class Meta:
+        resource_name = "provider-secrets"
+
+
+class AlibabaCloudRoleAssumptionProviderSecret(serializers.Serializer):
+    role_arn = serializers.CharField(
+        help_text="Access Key ID of the RAM user that will assume the role"
+    )
+    access_key_id = serializers.CharField(
+        help_text="Access Key ID of the RAM user that will assume the role"
+    )
+    access_key_secret = serializers.CharField(
+        help_text="Access Key Secret of the RAM user that will assume the role"
+    )
+    role_session_name = serializers.CharField(
+        required=False,
+        help_text="Session name for the assumed role session (optional, defaults to 'ProwlerSession')",
+    )
+
+    class Meta:
+        resource_name = "provider-secrets"
+
+
 class AWSRoleAssumptionProviderSecret(serializers.Serializer):
     role_arn = serializers.CharField()
     external_id = serializers.CharField()
@@ -2143,7 +2189,7 @@ class RoleProviderGroupRelationshipSerializer(RLSSerializer, BaseWriteSerializer
 # Compliance overview
 
 
-class ComplianceOverviewSerializer(serializers.Serializer):
+class ComplianceOverviewSerializer(BaseSerializerV1):
     """
     Serializer for compliance requirement status aggregated by compliance framework.
 
@@ -2165,7 +2211,7 @@ class ComplianceOverviewSerializer(serializers.Serializer):
         resource_name = "compliance-overviews"
 
 
-class ComplianceOverviewDetailSerializer(serializers.Serializer):
+class ComplianceOverviewDetailSerializer(BaseSerializerV1):
     """
     Serializer for detailed compliance requirement information.
 
@@ -2194,7 +2240,7 @@ class ComplianceOverviewDetailThreatscoreSerializer(ComplianceOverviewDetailSeri
     total_findings = serializers.IntegerField()
 
 
-class ComplianceOverviewAttributesSerializer(serializers.Serializer):
+class ComplianceOverviewAttributesSerializer(BaseSerializerV1):
     id = serializers.CharField()
     compliance_name = serializers.CharField()
     framework_description = serializers.CharField()
@@ -2208,7 +2254,7 @@ class ComplianceOverviewAttributesSerializer(serializers.Serializer):
         resource_name = "compliance-requirements-attributes"
 
 
-class ComplianceOverviewMetadataSerializer(serializers.Serializer):
+class ComplianceOverviewMetadataSerializer(BaseSerializerV1):
     regions = serializers.ListField(child=serializers.CharField(), allow_empty=True)
 
     class JSONAPIMeta:
@@ -2218,7 +2264,7 @@ class ComplianceOverviewMetadataSerializer(serializers.Serializer):
 # Overviews
 
 
-class OverviewProviderSerializer(serializers.Serializer):
+class OverviewProviderSerializer(BaseSerializerV1):
     id = serializers.CharField(source="provider")
     findings = serializers.SerializerMethodField(read_only=True)
     resources = serializers.SerializerMethodField(read_only=True)
@@ -2226,9 +2272,6 @@ class OverviewProviderSerializer(serializers.Serializer):
     class JSONAPIMeta:
         resource_name = "providers-overview"
 
-    def get_root_meta(self, _resource, _many):
-        return {"version": "v1"}
-
     @extend_schema_field(
         {
             "type": "object",
@@ -2262,18 +2305,15 @@ class OverviewProviderSerializer(serializers.Serializer):
         }
 
 
-class OverviewProviderCountSerializer(serializers.Serializer):
+class OverviewProviderCountSerializer(BaseSerializerV1):
     id = serializers.CharField(source="provider")
     count = serializers.IntegerField()
 
     class JSONAPIMeta:
         resource_name = "providers-count-overview"
 
-    def get_root_meta(self, _resource, _many):
-        return {"version": "v1"}
 
-
-class OverviewFindingSerializer(serializers.Serializer):
+class OverviewFindingSerializer(BaseSerializerV1):
     id = serializers.CharField(default="n/a")
     new = serializers.IntegerField()
     changed = serializers.IntegerField()
@@ -2292,15 +2332,12 @@ class OverviewFindingSerializer(serializers.Serializer):
     class JSONAPIMeta:
         resource_name = "findings-overview"
 
-    def get_root_meta(self, _resource, _many):
-        return {"version": "v1"}
-
     def __init__(self, *args, **kwargs):
         super().__init__(*args, **kwargs)
         self.fields["pass"] = self.fields.pop("_pass")
 
 
-class OverviewSeveritySerializer(serializers.Serializer):
+class OverviewSeveritySerializer(BaseSerializerV1):
     id = serializers.CharField(default="n/a")
     critical = serializers.IntegerField()
     high = serializers.IntegerField()
@@ -2311,11 +2348,24 @@ class OverviewSeveritySerializer(serializers.Serializer):
     class JSONAPIMeta:
         resource_name = "findings-severity-overview"
 
-    def get_root_meta(self, _resource, _many):
-        return {"version": "v1"}
+
+class FindingsSeverityOverTimeSerializer(BaseSerializerV1):
+    """Serializer for daily findings severity trend data."""
+
+    id = serializers.DateField(source="date")
+    critical = serializers.IntegerField()
+    high = serializers.IntegerField()
+    medium = serializers.IntegerField()
+    low = serializers.IntegerField()
+    informational = serializers.IntegerField()
+    muted = serializers.IntegerField()
+    scan_ids = serializers.ListField(child=serializers.UUIDField())
+
+    class JSONAPIMeta:
+        resource_name = "findings-severity-over-time"
 
 
-class OverviewServiceSerializer(serializers.Serializer):
+class OverviewServiceSerializer(BaseSerializerV1):
     id = serializers.CharField(source="service")
     total = serializers.IntegerField()
     _pass = serializers.IntegerField()
@@ -2329,8 +2379,46 @@ class OverviewServiceSerializer(serializers.Serializer):
         super().__init__(*args, **kwargs)
         self.fields["pass"] = self.fields.pop("_pass")
 
-    def get_root_meta(self, _resource, _many):
-        return {"version": "v1"}
+
+class AttackSurfaceOverviewSerializer(BaseSerializerV1):
+    """Serializer for attack surface overview aggregations."""
+
+    id = serializers.CharField(source="attack_surface_type")
+    total_findings = serializers.IntegerField()
+    failed_findings = serializers.IntegerField()
+    muted_failed_findings = serializers.IntegerField()
+
+    class JSONAPIMeta:
+        resource_name = "attack-surface-overviews"
+
+
+class CategoryOverviewSerializer(BaseSerializerV1):
+    """Serializer for category overview aggregations."""
+
+    id = serializers.CharField(source="category")
+    total_findings = serializers.IntegerField()
+    failed_findings = serializers.IntegerField()
+    new_failed_findings = serializers.IntegerField()
+    severity = serializers.JSONField(
+        help_text="Severity breakdown: {informational, low, medium, high, critical}"
+    )
+
+    class JSONAPIMeta:
+        resource_name = "category-overviews"
+
+
+class ComplianceWatchlistOverviewSerializer(BaseSerializerV1):
+    """Serializer for compliance watchlist overview with FAIL-dominant aggregation."""
+
+    id = serializers.CharField(source="compliance_id")
+    compliance_id = serializers.CharField()
+    requirements_passed = serializers.IntegerField()
+    requirements_failed = serializers.IntegerField()
+    requirements_manual = serializers.IntegerField()
+    total_requirements = serializers.IntegerField()
+
+    class JSONAPIMeta:
+        resource_name = "compliance-watchlist-overviews"
 
 
 class OverviewRegionSerializer(serializers.Serializer):
@@ -2360,7 +2448,7 @@ class OverviewRegionSerializer(serializers.Serializer):
 # Schedules
 
 
-class ScheduleDailyCreateSerializer(serializers.Serializer):
+class ScheduleDailyCreateSerializer(BaseSerializerV1):
     provider_id = serializers.UUIDField(required=True)
 
     class JSONAPIMeta:
@@ -2696,7 +2784,7 @@ class IntegrationUpdateSerializer(BaseWriteIntegrationSerializer):
         return representation
 
 
-class IntegrationJiraDispatchSerializer(serializers.Serializer):
+class IntegrationJiraDispatchSerializer(BaseSerializerV1):
     """
     Serializer for dispatching findings to JIRA integration.
     """
@@ -2859,14 +2947,14 @@ class ProcessorUpdateSerializer(BaseWriteSerializer):
 # SSO
 
 
-class SamlInitiateSerializer(serializers.Serializer):
+class SamlInitiateSerializer(BaseSerializerV1):
     email_domain = serializers.CharField()
 
     class JSONAPIMeta:
         resource_name = "saml-initiate"
 
 
-class SamlMetadataSerializer(serializers.Serializer):
+class SamlMetadataSerializer(BaseSerializerV1):
     class JSONAPIMeta:
         resource_name = "saml-meta"
 
@@ -3398,6 +3486,19 @@ class LighthouseProviderConfigUpdateSerializer(BaseWriteSerializer):
             and provider_type
             == LighthouseProviderConfiguration.LLMProviderChoices.BEDROCK
         ):
+            # For updates, enforce that the authentication method (access keys vs API key)
+            # is immutable. To switch methods, the UI must delete and recreate the provider.
+            existing_credentials = (
+                self.instance.credentials_decoded if self.instance else {}
+            ) or {}
+
+            existing_uses_api_key = "api_key" in existing_credentials
+            existing_uses_access_keys = any(
+                k in existing_credentials
+                for k in ("access_key_id", "secret_access_key")
+            )
+
+            # First run field-level validation on the partial payload
             try:
                 BedrockCredentialsUpdateSerializer(data=credentials).is_valid(
                     raise_exception=True
@@ -3408,6 +3509,31 @@ class LighthouseProviderConfigUpdateSerializer(BaseWriteSerializer):
                     e.detail[f"credentials/{key}"] = value
                     del e.detail[key]
                 raise e
+
+            # Then enforce invariants about not changing the auth method
+            # If the existing config uses an API key, forbid introducing access keys.
+            if existing_uses_api_key and any(
+                k in credentials for k in ("access_key_id", "secret_access_key")
+            ):
+                raise ValidationError(
+                    {
+                        "credentials/non_field_errors": [
+                            "Cannot change Bedrock authentication method from API key "
+                            "to access key via update. Delete and recreate the provider instead."
+                        ]
+                    }
+                )
+
+            # If the existing config uses access keys, forbid introducing an API key.
+            if existing_uses_access_keys and "api_key" in credentials:
+                raise ValidationError(
+                    {
+                        "credentials/non_field_errors": [
+                            "Cannot change Bedrock authentication method from access key "
+                            "to API key via update. Delete and recreate the provider instead."
+                        ]
+                    }
+                )
         elif (
             credentials is not None
             and provider_type

api/src/backend/api/v1/views.py

@@ -90,8 +90,12 @@ from api.db_router import MainRouter
 from api.db_utils import rls_transaction
 from api.exceptions import TaskFailedException
 from api.filters import (
+    AttackSurfaceOverviewFilter,
+    CategoryOverviewFilter,
     ComplianceOverviewFilter,
+    ComplianceWatchlistFilter,
     CustomDjangoFilterBackend,
+    DailySeveritySummaryFilter,
     FindingFilter,
     IntegrationFilter,
     IntegrationJiraFindingsFilter,
@@ -119,7 +123,10 @@ from api.filters import (
     UserFilter,
 )
 from api.models import (
+    AttackSurfaceOverview,
+    ComplianceOverviewSummary,
     ComplianceRequirementOverview,
+    DailySeveritySummary,
     Finding,
     Integration,
     Invitation,
@@ -132,6 +139,7 @@ from api.models import (
     MuteRule,
     Processor,
     Provider,
+    ProviderComplianceScore,
     ProviderGroup,
     ProviderGroupMembership,
     ProviderSecret,
@@ -145,11 +153,13 @@ from api.models import (
     SAMLDomainIndex,
     SAMLToken,
     Scan,
+    ScanCategorySummary,
     ScanSummary,
     SeverityChoices,
     StateChoices,
     Task,
     TenantAPIKey,
+    TenantComplianceSummary,
     ThreatScoreSnapshot,
     User,
     UserRoleRelationship,
@@ -169,14 +179,18 @@ from api.v1.serializers import (
     AttackPathsQuerySerializer,
     AttackPathsQueryResultSerializer,
     AttackPathsScanSerializer,
+    AttackSurfaceOverviewSerializer,
+    CategoryOverviewSerializer,
     ComplianceOverviewAttributesSerializer,
     ComplianceOverviewDetailSerializer,
     ComplianceOverviewDetailThreatscoreSerializer,
     ComplianceOverviewMetadataSerializer,
     ComplianceOverviewSerializer,
+    ComplianceWatchlistOverviewSerializer,
     FindingDynamicFilterSerializer,
     FindingMetadataSerializer,
     FindingSerializer,
+    FindingsSeverityOverTimeSerializer,
     IntegrationCreateSerializer,
     IntegrationJiraDispatchSerializer,
     IntegrationSerializer,
@@ -250,6 +264,7 @@ from tasks.beat import schedule_provider_scan
 from tasks.jobs.attack_paths import db_utils as attack_paths_db_utils
 from tasks.jobs.export import get_s3_client
 from tasks.tasks import (
+    backfill_compliance_summaries_task,
     backfill_scan_resource_summaries_task,
     check_integration_connection_task,
     check_lighthouse_connection_task,
@@ -363,7 +378,7 @@ class SchemaView(SpectacularAPIView):
 
     def get(self, request, *args, **kwargs):
         spectacular_settings.TITLE = "Prowler API"
-        spectacular_settings.VERSION = "1.15.0"
+        spectacular_settings.VERSION = "1.18.0"
         spectacular_settings.DESCRIPTION = (
             "Prowler API specification.\n\nThis file is auto-generated."
         )
@@ -2957,12 +2972,15 @@ class FindingViewSet(PaginateByPkMixin, BaseRLSViewSet):
 
         queryset = ResourceScanSummary.objects.filter(tenant_id=tenant_id)
         scan_based_filters = {}
+        category_scan_filters = {}  # Filters for ScanCategorySummary
 
         if scans := query_params.get("filter[scan__in]") or query_params.get(
             "filter[scan]"
         ):
-            queryset = queryset.filter(scan_id__in=scans.split(","))
-            scan_based_filters = {"id__in": scans.split(",")}
+            scan_ids_list = scans.split(",")
+            queryset = queryset.filter(scan_id__in=scan_ids_list)
+            scan_based_filters = {"id__in": scan_ids_list}
+            category_scan_filters = {"scan_id__in": scan_ids_list}
         else:
             exact = query_params.get("filter[inserted_at]")
             gte = query_params.get("filter[inserted_at__gte]")
@@ -3006,6 +3024,7 @@ class FindingViewSet(PaginateByPkMixin, BaseRLSViewSet):
                 scan_based_filters = {
                     key.lstrip("scan_"): value for key, value in date_filters.items()
                 }
+                category_scan_filters = date_filters
 
         # ToRemove: Temporary fallback mechanism
         if not queryset.exists():
@@ -3052,10 +3071,31 @@ class FindingViewSet(PaginateByPkMixin, BaseRLSViewSet):
             .order_by("resource_type")
         )
 
+        # Get categories from ScanCategorySummary using same scan filters
+        categories = list(
+            ScanCategorySummary.objects.filter(
+                tenant_id=tenant_id, **category_scan_filters
+            )
+            .values_list("category", flat=True)
+            .distinct()
+            .order_by("category")
+        )
+
+        # Fallback to finding aggregation if no ScanCategorySummary exists
+        if not categories:
+            categories_set = set()
+            for categories_list in filtered_queryset.values_list(
+                "categories", flat=True
+            ):
+                if categories_list:
+                    categories_set.update(categories_list)
+            categories = sorted(categories_set)
+
         result = {
             "services": services,
             "regions": regions,
             "resource_types": resource_types,
+            "categories": categories,
         }
 
         serializer = self.get_serializer(data=result)
@@ -3160,10 +3200,36 @@ class FindingViewSet(PaginateByPkMixin, BaseRLSViewSet):
             .order_by("resource_type")
         )
 
+        # Get categories from ScanCategorySummary for latest scans
+        categories = list(
+            ScanCategorySummary.objects.filter(
+                tenant_id=tenant_id,
+                scan_id__in=latest_scans_queryset.values_list("id", flat=True),
+            )
+            .values_list("category", flat=True)
+            .distinct()
+            .order_by("category")
+        )
+
+        # Fallback to finding aggregation if no ScanCategorySummary exists
+        if not categories:
+            filtered_queryset = self.filter_queryset(self.get_queryset()).filter(
+                tenant_id=tenant_id,
+                scan_id__in=latest_scans_queryset.values_list("id", flat=True),
+            )
+            categories_set = set()
+            for categories_list in filtered_queryset.values_list(
+                "categories", flat=True
+            ):
+                if categories_list:
+                    categories_set.update(categories_list)
+            categories = sorted(categories_set)
+
         result = {
             "services": services,
             "regions": regions,
             "resource_types": resource_types,
+            "categories": categories,
         }
 
         serializer = self.get_serializer(data=result)
@@ -3728,6 +3794,126 @@ class ComplianceOverviewViewSet(BaseRLSViewSet, TaskManagementMixin):
     def retrieve(self, request, *args, **kwargs):
         raise MethodNotAllowed(method="GET")
 
+    def _compliance_summaries_queryset(self, scan_id):
+        """Return pre-aggregated summaries constrained by RBAC visibility."""
+        role = get_role(self.request.user)
+        unlimited_visibility = getattr(
+            role, Permissions.UNLIMITED_VISIBILITY.value, False
+        )
+        summaries = ComplianceOverviewSummary.objects.filter(
+            tenant_id=self.request.tenant_id,
+            scan_id=scan_id,
+        )
+
+        if not unlimited_visibility:
+            providers = Provider.all_objects.filter(
+                provider_groups__in=role.provider_groups.all()
+            ).distinct()
+            summaries = summaries.filter(scan__provider__in=providers)
+
+        return summaries
+
+    def _get_compliance_template(self, *, provider=None, scan_id=None):
+        """Return the compliance template for the given provider or scan."""
+        if provider is None and scan_id is not None:
+            try:
+                scan = Scan.all_objects.select_related("provider").get(pk=scan_id)
+            except Scan.DoesNotExist:
+                raise ValidationError(
+                    [
+                        {
+                            "detail": "Scan not found",
+                            "status": 404,
+                            "source": {"pointer": "filter[scan_id]"},
+                            "code": "not_found",
+                        }
+                    ]
+                )
+            provider = scan.provider
+
+        if not provider:
+            return {}
+
+        return PROWLER_COMPLIANCE_OVERVIEW_TEMPLATE.get(provider.provider, {})
+
+    def _aggregate_compliance_overview(self, queryset, template_metadata=None):
+        """
+        Aggregate requirement rows into compliance overview dictionaries.
+
+        Args:
+            queryset: ComplianceRequirementOverview queryset already filtered.
+            template_metadata: Optional dict mapping compliance_id -> metadata.
+        """
+        template_metadata = template_metadata or {}
+        requirement_status_subquery = queryset.values(
+            "compliance_id", "requirement_id"
+        ).annotate(
+            fail_count=Count("id", filter=Q(requirement_status="FAIL")),
+            pass_count=Count("id", filter=Q(requirement_status="PASS")),
+            total_count=Count("id"),
+        )
+
+        compliance_data = {}
+        fallback_metadata = {
+            item["compliance_id"]: {
+                "framework": item["framework"],
+                "version": item["version"],
+            }
+            for item in queryset.values(
+                "compliance_id", "framework", "version"
+            ).distinct()
+        }
+
+        for item in requirement_status_subquery:
+            compliance_id = item["compliance_id"]
+
+            if item["fail_count"] > 0:
+                req_status = "FAIL"
+            elif item["pass_count"] == item["total_count"]:
+                req_status = "PASS"
+            else:
+                req_status = "MANUAL"
+
+            compliance_status = compliance_data.setdefault(
+                compliance_id,
+                {
+                    "total_requirements": 0,
+                    "requirements_passed": 0,
+                    "requirements_failed": 0,
+                    "requirements_manual": 0,
+                },
+            )
+
+            compliance_status["total_requirements"] += 1
+            if req_status == "PASS":
+                compliance_status["requirements_passed"] += 1
+            elif req_status == "FAIL":
+                compliance_status["requirements_failed"] += 1
+            else:
+                compliance_status["requirements_manual"] += 1
+
+        response_data = []
+        for compliance_id, data in compliance_data.items():
+            template = template_metadata.get(compliance_id, {})
+            fallback = fallback_metadata.get(compliance_id, {})
+
+            response_data.append(
+                {
+                    "id": compliance_id,
+                    "compliance_id": compliance_id,
+                    "framework": template.get("framework")
+                    or fallback.get("framework", ""),
+                    "version": template.get("version") or fallback.get("version", ""),
+                    "requirements_passed": data["requirements_passed"],
+                    "requirements_failed": data["requirements_failed"],
+                    "requirements_manual": data["requirements_manual"],
+                    "total_requirements": data["total_requirements"],
+                }
+            )
+
+        serializer = self.get_serializer(response_data, many=True)
+        return serializer.data
+
     def _task_response_if_running(self, scan_id):
         """Check for an in-progress task only when no compliance data exists."""
         try:
@@ -3742,90 +3928,84 @@ class ComplianceOverviewViewSet(BaseRLSViewSet, TaskManagementMixin):
                 status=status.HTTP_500_INTERNAL_SERVER_ERROR,
             )
 
-    def list(self, request, *args, **kwargs):
-        scan_id = request.query_params.get("filter[scan_id]")
-        if not scan_id:
-            raise ValidationError(
-                [
-                    {
-                        "detail": "This query parameter is required.",
-                        "status": 400,
-                        "source": {"pointer": "filter[scan_id]"},
-                        "code": "required",
-                    }
-                ]
-            )
-        try:
-            if task := self.get_task_response_if_running(
-                task_name="scan-compliance-overviews",
-                task_kwargs={"tenant_id": self.request.tenant_id, "scan_id": scan_id},
-                raise_on_not_found=False,
-            ):
-                return task
-        except TaskFailedException:
-            return Response(
-                {"detail": "Task failed to generate compliance overview data."},
-                status=status.HTTP_500_INTERNAL_SERVER_ERROR,
-            )
-        queryset = self.filter_queryset(self.filter_queryset(self.get_queryset()))
-
-        requirement_status_subquery = queryset.values(
-            "compliance_id", "requirement_id"
-        ).annotate(
-            fail_count=Count("id", filter=Q(requirement_status="FAIL")),
-            pass_count=Count("id", filter=Q(requirement_status="PASS")),
-            total_count=Count("id"),
+    def _list_with_region_filter(self, scan_id, region_filter):
+        """
+        Fall back to detailed ComplianceRequirementOverview query when region filter is applied.
+        This uses the original aggregation logic across filtered regions.
+        """
+        regions = region_filter.split(",") if "," in region_filter else [region_filter]
+        queryset = self.filter_queryset(self.get_queryset()).filter(
+            scan_id=scan_id,
+            region__in=regions,
         )
 
-        compliance_data = {}
-        framework_info = {}
+        data = self._aggregate_compliance_overview(queryset)
+        if data:
+            return Response(data)
 
-        for item in queryset.values("compliance_id", "framework", "version").distinct():
-            framework_info[item["compliance_id"]] = {
-                "framework": item["framework"],
-                "version": item["version"],
-            }
+        task_response = self._task_response_if_running(scan_id)
+        if task_response:
+            return task_response
 
-        for item in requirement_status_subquery:
-            compliance_id = item["compliance_id"]
+        return Response(data)
 
-            if item["fail_count"] > 0:
-                req_status = "FAIL"
-            elif item["pass_count"] == item["total_count"]:
-                req_status = "PASS"
-            else:
-                req_status = "MANUAL"
+    def _list_without_region_aggregation(self, scan_id):
+        """
+        Fall back aggregation when compliance summaries don't exist yet.
+        Aggregates ComplianceRequirementOverview data across ALL regions.
+        """
+        queryset = self.filter_queryset(self.get_queryset()).filter(scan_id=scan_id)
+        compliance_template = self._get_compliance_template(scan_id=scan_id)
+        data = self._aggregate_compliance_overview(
+            queryset, template_metadata=compliance_template
+        )
+        if data:
+            return Response(data)
 
-            if compliance_id not in compliance_data:
-                compliance_data[compliance_id] = {
-                    "total_requirements": 0,
-                    "requirements_passed": 0,
-                    "requirements_failed": 0,
-                    "requirements_manual": 0,
-                }
+        task_response = self._task_response_if_running(scan_id)
+        if task_response:
+            return task_response
 
-            compliance_data[compliance_id]["total_requirements"] += 1
-            if req_status == "PASS":
-                compliance_data[compliance_id]["requirements_passed"] += 1
-            elif req_status == "FAIL":
-                compliance_data[compliance_id]["requirements_failed"] += 1
-            else:
-                compliance_data[compliance_id]["requirements_manual"] += 1
+        return Response(data)
 
+    def list(self, request, *args, **kwargs):
+        scan_id = request.query_params.get("filter[scan_id]")
+
+        # Specific scan requested - use optimized summaries with region support
+        region_filter = request.query_params.get(
+            "filter[region]"
+        ) or request.query_params.get("filter[region__in]")
+
+        if region_filter:
+            # Fall back to detailed query with region filtering
+            return self._list_with_region_filter(scan_id, region_filter)
+
+        summaries = list(self._compliance_summaries_queryset(scan_id))
+        if not summaries:
+            # Trigger async backfill for next time
+            backfill_compliance_summaries_task.delay(
+                tenant_id=self.request.tenant_id, scan_id=scan_id
+            )
+            # Use fallback aggregation for this request
+            return self._list_without_region_aggregation(scan_id)
+
+        # Get compliance template for provider to enrich with framework/version
+        compliance_template = self._get_compliance_template(scan_id=scan_id)
+
+        # Convert to response format with framework/version enrichment
         response_data = []
-        for compliance_id, data in compliance_data.items():
-            framework = framework_info.get(compliance_id, {})
-
+        for summary in summaries:
+            compliance_metadata = compliance_template.get(summary.compliance_id, {})
             response_data.append(
                 {
-                    "id": compliance_id,
-                    "compliance_id": compliance_id,
-                    "framework": framework.get("framework", ""),
-                    "version": framework.get("version", ""),
-                    "requirements_passed": data["requirements_passed"],
-                    "requirements_failed": data["requirements_failed"],
-                    "requirements_manual": data["requirements_manual"],
-                    "total_requirements": data["total_requirements"],
+                    "id": summary.compliance_id,
+                    "compliance_id": summary.compliance_id,
+                    "framework": compliance_metadata.get("framework", ""),
+                    "version": compliance_metadata.get("version", ""),
+                    "requirements_passed": summary.requirements_passed,
+                    "requirements_failed": summary.requirements_failed,
+                    "requirements_manual": summary.requirements_manual,
+                    "total_requirements": summary.total_requirements,
                 }
             )
 
@@ -4095,6 +4275,34 @@ class ComplianceOverviewViewSet(BaseRLSViewSet, TaskManagementMixin):
         ),
         filters=True,
     ),
+    findings_severity_timeseries=extend_schema(
+        summary="Get findings severity data over time",
+        description=(
+            "Retrieve daily aggregated findings data grouped by severity levels over a date range. "
+            "Returns one data point per day with counts of failed findings by severity (critical, high, "
+            "medium, low, informational) and muted findings. Days without scans are filled forward with "
+            "the most recent known values. Use date_from (required) and date_to filters to specify the range."
+        ),
+        filters=True,
+    ),
+    attack_surface=extend_schema(
+        summary="Get attack surface overview",
+        description="Retrieve aggregated attack surface metrics from latest completed scans per provider.",
+        tags=["Overview"],
+        filters=True,
+        responses={200: AttackSurfaceOverviewSerializer(many=True)},
+    ),
+    categories=extend_schema(
+        summary="Get category overview",
+        description=(
+            "Retrieve aggregated category metrics from latest completed scans per provider. "
+            "Returns one row per category with total, failed, and new failed findings counts, "
+            "plus a severity breakdown showing failed findings per severity level. "
+        ),
+        tags=["Overview"],
+        filters=True,
+        responses={200: CategoryOverviewSerializer(many=True)},
+    ),
 )
 @method_decorator(CACHE_DECORATOR, name="list")
 class OverviewViewSet(BaseRLSViewSet):
@@ -4112,7 +4320,16 @@ class OverviewViewSet(BaseRLSViewSet):
         if not role.unlimited_visibility:
             self.allowed_providers = providers
 
-        return ScanSummary.all_objects.filter(tenant_id=self.request.tenant_id)
+        tenant_id = self.request.tenant_id
+
+        # Return appropriate queryset per action
+        if self.action == "findings_severity_timeseries":
+            qs = DailySeveritySummary.objects.filter(tenant_id=tenant_id)
+            if hasattr(self, "allowed_providers"):
+                qs = qs.filter(provider_id__in=self.allowed_providers)
+            return qs
+
+        return ScanSummary.all_objects.filter(tenant_id=tenant_id)
 
     def get_serializer_class(self):
         if self.action == "providers":
@@ -4123,12 +4340,20 @@ class OverviewViewSet(BaseRLSViewSet):
             return OverviewFindingSerializer
         elif self.action == "findings_severity":
             return OverviewSeveritySerializer
+        elif self.action == "findings_severity_timeseries":
+            return FindingsSeverityOverTimeSerializer
         elif self.action == "services":
             return OverviewServiceSerializer
         elif self.action == "regions":
             return OverviewRegionSerializer
         elif self.action == "threatscore":
             return ThreatScoreSnapshotSerializer
+        elif self.action == "attack_surface":
+            return AttackSurfaceOverviewSerializer
+        elif self.action == "categories":
+            return CategoryOverviewSerializer
+        elif self.action == "compliance_watchlist":
+            return ComplianceWatchlistOverviewSerializer
         return super().get_serializer_class()
 
     def get_filterset_class(self):
@@ -4138,8 +4363,24 @@ class OverviewViewSet(BaseRLSViewSet):
             return ScanSummaryFilter
         elif self.action == "findings_severity":
             return ScanSummarySeverityFilter
+        elif self.action == "findings_severity_timeseries":
+            return DailySeveritySummaryFilter
+        elif self.action == "categories":
+            return CategoryOverviewFilter
+        elif self.action == "attack_surface":
+            return AttackSurfaceOverviewFilter
+        elif self.action == "compliance_watchlist":
+            return ComplianceWatchlistFilter
         return None
 
+    def filter_queryset(self, queryset):
+        # Skip OrderingFilter for findings_severity_timeseries (no inserted_at field)
+        if self.action == "findings_severity_timeseries":
+            return CustomDjangoFilterBackend().filter_queryset(
+                self.request, queryset, self
+            )
+        return super().filter_queryset(queryset)
+
     @extend_schema(exclude=True)
     def list(self, request, *args, **kwargs):
         raise MethodNotAllowed(method="GET")
@@ -4177,6 +4418,93 @@ class OverviewViewSet(BaseRLSViewSet):
             tenant_id=tenant_id, scan_id__in=latest_scan_ids
         )
 
+    def _normalize_jsonapi_params(self, query_params, exclude_keys=None):
+        """Convert JSON:API filter params (filter[X]) to flat params (X)."""
+        exclude_keys = exclude_keys or set()
+        normalized = QueryDict(mutable=True)
+        for key, values in query_params.lists():
+            normalized_key = (
+                key[7:-1] if key.startswith("filter[") and key.endswith("]") else key
+            )
+            if normalized_key not in exclude_keys:
+                normalized.setlist(normalized_key, values)
+        return normalized
+
+    def _ensure_allowed_providers(self):
+        """Populate allowed providers for RBAC-aware queries once per request."""
+        if getattr(self, "_providers_initialized", False):
+            return
+        self.get_queryset()
+        self._providers_initialized = True
+
+    def _get_provider_filter(self, provider_field="provider"):
+        self._ensure_allowed_providers()
+        if hasattr(self, "allowed_providers"):
+            return {f"{provider_field}__in": self.allowed_providers}
+        return {}
+
+    def _apply_provider_filter(self, queryset, provider_field="provider"):
+        provider_filter = self._get_provider_filter(provider_field)
+        if provider_filter:
+            return queryset.filter(**provider_filter)
+        return queryset
+
+    def _apply_filterset(self, queryset, filterset_class, exclude_keys=None):
+        normalized_params = self._normalize_jsonapi_params(
+            self.request.query_params, exclude_keys=set(exclude_keys or [])
+        )
+        filterset = filterset_class(normalized_params, queryset=queryset)
+        if not filterset.is_valid():
+            raise ValidationError(filterset.errors)
+        return filterset.qs
+
+    def _latest_scan_ids_for_allowed_providers(self, tenant_id, provider_filters=None):
+        provider_filter = self._get_provider_filter()
+        queryset = Scan.all_objects.filter(
+            tenant_id=tenant_id, state=StateChoices.COMPLETED, **provider_filter
+        )
+        if provider_filters:
+            queryset = queryset.filter(**provider_filters)
+        return (
+            queryset.order_by("provider_id", "-inserted_at")
+            .distinct("provider_id")
+            .values_list("id", flat=True)
+        )
+
+    def _extract_provider_filters_from_params(self):
+        """Extract and validate provider filters from query params."""
+        params = self.request.query_params
+        filters = {}
+        valid_provider_types = {c[0] for c in Provider.ProviderChoices.choices}
+
+        provider_id = params.get("filter[provider_id]")
+        if provider_id:
+            filters["provider_id"] = provider_id
+
+        provider_id_in = params.get("filter[provider_id__in]")
+        if provider_id_in:
+            filters["provider_id__in"] = provider_id_in.split(",")
+
+        provider_type = params.get("filter[provider_type]")
+        if provider_type:
+            if provider_type not in valid_provider_types:
+                raise ValidationError(
+                    {"provider_type": f"Invalid choice: {provider_type}"}
+                )
+            filters["provider__provider"] = provider_type
+
+        provider_type_in = params.get("filter[provider_type__in]")
+        if provider_type_in:
+            types = provider_type_in.split(",")
+            invalid = [t for t in types if t not in valid_provider_types]
+            if invalid:
+                raise ValidationError(
+                    {"provider_type__in": f"Invalid choices: {', '.join(invalid)}"}
+                )
+            filters["provider__provider__in"] = types
+
+        return filters
+
     @action(detail=False, methods=["get"], url_name="providers")
     def providers(self, request):
         tenant_id = self.request.tenant_id
@@ -4354,6 +4682,108 @@ class OverviewViewSet(BaseRLSViewSet):
 
         return Response(serializer.data, status=status.HTTP_200_OK)
 
+    @action(
+        detail=False,
+        methods=["get"],
+        url_path="findings_severity/timeseries",
+        url_name="findings_severity_timeseries",
+    )
+    def findings_severity_timeseries(self, request):
+        """
+        Daily severity trends for charts. Uses DailySeveritySummary pre-aggregation.
+        Requires date_from filter.
+        """
+        # Get queryset with RBAC, provider, and date filters applied
+        # Date validation is handled by DailySeveritySummaryFilter
+        daily_qs = self.filter_queryset(self.get_queryset())
+
+        date_from = request._date_from
+        date_to = request._date_to
+
+        if not daily_qs.exists():
+            # No data matches filters - return zeros
+            result = self._generate_zero_result(date_from, date_to)
+            serializer = self.get_serializer(result, many=True)
+            return Response(serializer.data, status=status.HTTP_200_OK)
+
+        # Fetch all data for fill-forward logic
+        daily_summaries = list(
+            daily_qs.order_by("provider_id", "-date").values(
+                "provider_id",
+                "scan_id",
+                "date",
+                "critical",
+                "high",
+                "medium",
+                "low",
+                "informational",
+                "muted",
+            )
+        )
+
+        if not daily_summaries:
+            result = self._generate_zero_result(date_from, date_to)
+            serializer = self.get_serializer(result, many=True)
+            return Response(serializer.data, status=status.HTTP_200_OK)
+
+        # Build provider_data: {provider_id: [(date, data), ...]} sorted by date desc
+        provider_data = defaultdict(list)
+        for summary in daily_summaries:
+            provider_data[summary["provider_id"]].append(summary)
+
+        # For each day, find the latest data per provider and sum values
+        result = []
+        current_date = date_from
+        while current_date <= date_to:
+            day_totals = {
+                "critical": 0,
+                "high": 0,
+                "medium": 0,
+                "low": 0,
+                "informational": 0,
+                "muted": 0,
+            }
+            day_scan_ids = []
+
+            for provider_id, summaries in provider_data.items():
+                # Find the latest data for this provider <= current_date
+                for summary in summaries:  # Already sorted by date desc
+                    if summary["date"] <= current_date:
+                        day_totals["critical"] += summary["critical"] or 0
+                        day_totals["high"] += summary["high"] or 0
+                        day_totals["medium"] += summary["medium"] or 0
+                        day_totals["low"] += summary["low"] or 0
+                        day_totals["informational"] += summary["informational"] or 0
+                        day_totals["muted"] += summary["muted"] or 0
+                        day_scan_ids.append(summary["scan_id"])
+                        break  # Found the latest data for this provider
+
+            result.append(
+                {"date": current_date, "scan_ids": day_scan_ids, **day_totals}
+            )
+            current_date += timedelta(days=1)
+
+        serializer = self.get_serializer(result, many=True)
+        return Response(serializer.data, status=status.HTTP_200_OK)
+
+    def _generate_zero_result(self, date_from, date_to):
+        """Generate a list of zero-filled results for each date in range."""
+        result = []
+        current_date = date_from
+        zero_values = {
+            "critical": 0,
+            "high": 0,
+            "medium": 0,
+            "low": 0,
+            "informational": 0,
+            "muted": 0,
+            "scan_ids": [],
+        }
+        while current_date <= date_to:
+            result.append({"date": current_date, **zero_values})
+            current_date += timedelta(days=1)
+        return result
+
     @extend_schema(
         summary="Get ThreatScore snapshots",
         description=(
@@ -4406,11 +4836,9 @@ class OverviewViewSet(BaseRLSViewSet):
         snapshot_id = request.query_params.get("snapshot_id")
 
         # Base queryset with RLS
-        base_queryset = ThreatScoreSnapshot.objects.filter(tenant_id=tenant_id)
-
-        # Apply RBAC filtering
-        if hasattr(self, "allowed_providers"):
-            base_queryset = base_queryset.filter(provider__in=self.allowed_providers)
+        base_queryset = self._apply_provider_filter(
+            ThreatScoreSnapshot.objects.filter(tenant_id=tenant_id)
+        )
 
         # Case 1: Specific snapshot requested
         if snapshot_id:
@@ -4426,17 +4854,9 @@ class OverviewViewSet(BaseRLSViewSet):
         # Case 2: Latest snapshot per provider (default)
         # Apply filters manually: this @action is outside the standard list endpoint flow,
         # so DRF's filter backends don't execute and we must flatten JSON:API params ourselves.
-        normalized_params = QueryDict(mutable=True)
-        for param_key, values in request.query_params.lists():
-            normalized_key = param_key
-            if param_key.startswith("filter[") and param_key.endswith("]"):
-                normalized_key = param_key[7:-1]
-            if normalized_key == "snapshot_id":
-                continue
-            normalized_params.setlist(normalized_key, values)
-
-        filterset = ThreatScoreSnapshotFilter(normalized_params, queryset=base_queryset)
-        filtered_queryset = filterset.qs
+        filtered_queryset = self._apply_filterset(
+            base_queryset, ThreatScoreSnapshotFilter, exclude_keys={"snapshot_id"}
+        )
 
         # Get distinct provider IDs from filtered queryset
         # Pick the latest snapshot per provider using Postgres DISTINCT ON pattern.
@@ -4680,6 +5100,203 @@ class OverviewViewSet(BaseRLSViewSet):
 
         return aggregated_snapshot
 
+    @action(
+        detail=False,
+        methods=["get"],
+        url_name="attack-surface",
+        url_path="attack-surfaces",
+    )
+    def attack_surface(self, request):
+        tenant_id = request.tenant_id
+        latest_scan_ids = self._latest_scan_ids_for_allowed_providers(tenant_id)
+
+        base_queryset = AttackSurfaceOverview.objects.filter(
+            tenant_id=tenant_id, scan_id__in=latest_scan_ids
+        )
+        filtered_queryset = self._apply_filterset(
+            base_queryset, AttackSurfaceOverviewFilter
+        )
+
+        aggregation = filtered_queryset.values("attack_surface_type").annotate(
+            total_findings=Coalesce(Sum("total_findings"), 0),
+            failed_findings=Coalesce(Sum("failed_findings"), 0),
+            muted_failed_findings=Coalesce(Sum("muted_failed_findings"), 0),
+        )
+
+        results = {
+            attack_surface_type: {
+                "total_findings": 0,
+                "failed_findings": 0,
+                "muted_failed_findings": 0,
+            }
+            for attack_surface_type in AttackSurfaceOverview.AttackSurfaceTypeChoices.values
+        }
+        for item in aggregation:
+            results[item["attack_surface_type"]] = {
+                "total_findings": item["total_findings"],
+                "failed_findings": item["failed_findings"],
+                "muted_failed_findings": item["muted_failed_findings"],
+            }
+
+        response_data = [
+            {"attack_surface_type": key, **value} for key, value in results.items()
+        ]
+
+        return Response(
+            self.get_serializer(response_data, many=True).data,
+            status=status.HTTP_200_OK,
+        )
+
+    @action(detail=False, methods=["get"], url_name="categories")
+    def categories(self, request):
+        tenant_id = request.tenant_id
+        provider_filters = self._extract_provider_filters_from_params()
+        latest_scan_ids = self._latest_scan_ids_for_allowed_providers(
+            tenant_id, provider_filters
+        )
+
+        base_queryset = ScanCategorySummary.objects.filter(
+            tenant_id=tenant_id, scan_id__in=latest_scan_ids
+        )
+        provider_filter_keys = {
+            "provider_id",
+            "provider_id__in",
+            "provider_type",
+            "provider_type__in",
+        }
+        filtered_queryset = self._apply_filterset(
+            base_queryset, CategoryOverviewFilter, exclude_keys=provider_filter_keys
+        )
+
+        aggregation = (
+            filtered_queryset.values("category", "severity")
+            .annotate(
+                total=Coalesce(Sum("total_findings"), 0),
+                failed=Coalesce(Sum("failed_findings"), 0),
+                new_failed=Coalesce(Sum("new_failed_findings"), 0),
+            )
+            .order_by("category", "severity")
+        )
+
+        category_data = defaultdict(
+            lambda: {
+                "total_findings": 0,
+                "failed_findings": 0,
+                "new_failed_findings": 0,
+                "severity": {
+                    "informational": 0,
+                    "low": 0,
+                    "medium": 0,
+                    "high": 0,
+                    "critical": 0,
+                },
+            }
+        )
+
+        for row in aggregation:
+            cat = row["category"]
+            sev = row["severity"]
+            category_data[cat]["total_findings"] += row["total"]
+            category_data[cat]["failed_findings"] += row["failed"]
+            category_data[cat]["new_failed_findings"] += row["new_failed"]
+            if sev in category_data[cat]["severity"]:
+                category_data[cat]["severity"][sev] = row["failed"]
+
+        response_data = [
+            {"category": cat, **data} for cat, data in sorted(category_data.items())
+        ]
+
+        return Response(
+            self.get_serializer(response_data, many=True).data,
+            status=status.HTTP_200_OK,
+        )
+
+    @action(
+        detail=False,
+        methods=["get"],
+        url_name="compliance-watchlist",
+        url_path="compliance-watchlist",
+    )
+    def compliance_watchlist(self, request):
+        """
+        Get compliance watchlist overview with FAIL-dominant aggregation.
+
+        Without filters: uses pre-aggregated TenantComplianceSummary (~70 rows).
+        With provider filters: queries ProviderComplianceScore with FAIL-dominant logic.
+        """
+        tenant_id = request.tenant_id
+        rbac_filter = self._get_provider_filter()
+        query_params = request.query_params
+
+        has_provider_filter = any(
+            key.startswith("filter[provider") for key in query_params.keys()
+        )
+        has_rbac_restriction = bool(rbac_filter)
+
+        if not has_provider_filter and not has_rbac_restriction:
+            response_data = list(
+                TenantComplianceSummary.objects.filter(tenant_id=tenant_id)
+                .values(
+                    "compliance_id",
+                    "requirements_passed",
+                    "requirements_failed",
+                    "requirements_manual",
+                    "total_requirements",
+                )
+                .order_by("compliance_id")
+            )
+        else:
+            base_queryset = ProviderComplianceScore.objects.filter(
+                tenant_id=tenant_id, **rbac_filter
+            )
+
+            filtered_queryset = self._apply_filterset(
+                base_queryset, ComplianceWatchlistFilter
+            )
+
+            aggregation = (
+                filtered_queryset.values("compliance_id", "requirement_id")
+                .annotate(
+                    has_fail=Sum(
+                        Case(When(requirement_status="FAIL", then=1), default=0)
+                    ),
+                    has_manual=Sum(
+                        Case(When(requirement_status="MANUAL", then=1), default=0)
+                    ),
+                )
+                .values("compliance_id", "requirement_id", "has_fail", "has_manual")
+            )
+
+            compliance_data = defaultdict(
+                lambda: {
+                    "requirements_passed": 0,
+                    "requirements_failed": 0,
+                    "requirements_manual": 0,
+                    "total_requirements": 0,
+                }
+            )
+
+            for row in aggregation:
+                cid = row["compliance_id"]
+                compliance_data[cid]["total_requirements"] += 1
+
+                if row["has_fail"] and row["has_fail"] > 0:
+                    compliance_data[cid]["requirements_failed"] += 1
+                elif row["has_manual"] and row["has_manual"] > 0:
+                    compliance_data[cid]["requirements_manual"] += 1
+                else:
+                    compliance_data[cid]["requirements_passed"] += 1
+
+            response_data = [
+                {"compliance_id": cid, **data}
+                for cid, data in sorted(compliance_data.items())
+            ]
+
+        return Response(
+            self.get_serializer(response_data, many=True).data,
+            status=status.HTTP_200_OK,
+        )
+
 
 @extend_schema(tags=["Schedule"])
 @extend_schema_view(

api/src/backend/config/django/devel.py

@@ -36,6 +36,14 @@ DATABASES = {
         "HOST": env("POSTGRES_REPLICA_HOST", default=default_db_host),
         "PORT": env("POSTGRES_REPLICA_PORT", default=default_db_port),
     },
+    "admin_replica": {
+        "ENGINE": "psqlextra.backend",
+        "NAME": env("POSTGRES_REPLICA_DB", default=default_db_name),
+        "USER": env("POSTGRES_ADMIN_USER", default="prowler"),
+        "PASSWORD": env("POSTGRES_ADMIN_PASSWORD", default="S3cret"),
+        "HOST": env("POSTGRES_REPLICA_HOST", default=default_db_host),
+        "PORT": env("POSTGRES_REPLICA_PORT", default=default_db_port),
+    },
     "neo4j": {
         "HOST": env.str("NEO4J_HOST", "neo4j"),
         "PORT": env.str("NEO4J_PORT", "7687"),

api/src/backend/config/django/production.py

@@ -37,6 +37,14 @@ DATABASES = {
         "HOST": env("POSTGRES_REPLICA_HOST", default=default_db_host),
         "PORT": env("POSTGRES_REPLICA_PORT", default=default_db_port),
     },
+    "admin_replica": {
+        "ENGINE": "psqlextra.backend",
+        "NAME": env("POSTGRES_REPLICA_DB", default=default_db_name),
+        "USER": env("POSTGRES_ADMIN_USER"),
+        "PASSWORD": env("POSTGRES_ADMIN_PASSWORD"),
+        "HOST": env("POSTGRES_REPLICA_HOST", default=default_db_host),
+        "PORT": env("POSTGRES_REPLICA_PORT", default=default_db_port),
+    },
     "neo4j": {
         "HOST": env.str("NEO4J_HOST"),
         "PORT": env.str("NEO4J_PORT"),

api/src/backend/config/settings/sentry.py

@@ -5,6 +5,9 @@ IGNORED_EXCEPTIONS = [
     # Provider is not connected due to credentials errors
     "is not connected",
     "ProviderConnectionError",
+    # Provider was deleted during a scan
+    "ProviderDeletedException",
+    "violates foreign key constraint",
     # Authentication Errors from AWS
     "InvalidToken",
     "AccessDeniedException",

api/src/backend/conftest.py

@@ -22,6 +22,7 @@ from api.attack_paths import (
 from api.db_utils import rls_transaction
 from api.models import (
     AttackPathsScan,
+    AttackSurfaceOverview,
     ComplianceOverview,
     ComplianceRequirementOverview,
     Finding,
@@ -33,6 +34,7 @@ from api.models import (
     MuteRule,
     Processor,
     Provider,
+    ProviderComplianceScore,
     ProviderGroup,
     ProviderSecret,
     Resource,
@@ -42,11 +44,13 @@ from api.models import (
     SAMLConfiguration,
     SAMLDomainIndex,
     Scan,
+    ScanCategorySummary,
     ScanSummary,
     StateChoices,
     StatusChoices,
     Task,
     TenantAPIKey,
+    TenantComplianceSummary,
     User,
     UserRoleRelationship,
 )
@@ -54,7 +58,7 @@ from api.rls import Tenant
 from api.v1.serializers import TokenSerializer
 from prowler.lib.check.models import Severity
 from prowler.lib.outputs.finding import Status
-from tasks.jobs.backfill import backfill_resource_scan_summaries
+from tasks.jobs.backfill import backfill_resource_scan_summaries, backfill_scan_category_summaries
 
 TODAY = str(datetime.today().date())
 API_JSON_CONTENT_TYPE = "application/vnd.api+json"
@@ -518,6 +522,12 @@ def providers_fixture(tenants_fixture):
         alias="mongodbatlas_testing",
         tenant_id=tenant.id,
     )
+    provider9 = Provider.objects.create(
+        provider="alibabacloud",
+        uid="1234567890123456",
+        alias="alibabacloud_testing",
+        tenant_id=tenant.id,
+    )
 
     return (
         provider1,
@@ -528,6 +538,7 @@ def providers_fixture(tenants_fixture):
         provider6,
         provider7,
         provider8,
+        provider9,
     )
 
 
@@ -1276,6 +1287,113 @@ def latest_scan_finding(authenticated_client, providers_fixture, resources_fixtu
     return finding
 
 
+@pytest.fixture(scope="function")
+def findings_with_categories(scans_fixture, resources_fixture):
+    scan = scans_fixture[0]
+    resource = resources_fixture[0]
+
+    finding = Finding.objects.create(
+        tenant_id=scan.tenant_id,
+        uid="finding_with_categories_1",
+        scan=scan,
+        delta=None,
+        status=Status.FAIL,
+        status_extended="test status",
+        impact=Severity.critical,
+        impact_extended="test impact",
+        severity=Severity.critical,
+        raw_result={"status": Status.FAIL},
+        check_id="genai_check",
+        check_metadata={"CheckId": "genai_check"},
+        categories=["gen-ai", "security"],
+        first_seen_at="2024-01-02T00:00:00Z",
+    )
+    finding.add_resources([resource])
+    backfill_resource_scan_summaries(str(scan.tenant_id), str(scan.id))
+    return finding
+
+
+@pytest.fixture(scope="function")
+def findings_with_multiple_categories(scans_fixture, resources_fixture):
+    scan = scans_fixture[0]
+    resource1, resource2 = resources_fixture[:2]
+
+    finding1 = Finding.objects.create(
+        tenant_id=scan.tenant_id,
+        uid="finding_multi_cat_1",
+        scan=scan,
+        delta=None,
+        status=Status.FAIL,
+        status_extended="test status",
+        impact=Severity.critical,
+        impact_extended="test impact",
+        severity=Severity.critical,
+        raw_result={"status": Status.FAIL},
+        check_id="genai_check",
+        check_metadata={"CheckId": "genai_check"},
+        categories=["gen-ai", "security"],
+        first_seen_at="2024-01-02T00:00:00Z",
+    )
+    finding1.add_resources([resource1])
+
+    finding2 = Finding.objects.create(
+        tenant_id=scan.tenant_id,
+        uid="finding_multi_cat_2",
+        scan=scan,
+        delta=None,
+        status=Status.FAIL,
+        status_extended="test status 2",
+        impact=Severity.high,
+        impact_extended="test impact 2",
+        severity=Severity.high,
+        raw_result={"status": Status.FAIL},
+        check_id="iam_check",
+        check_metadata={"CheckId": "iam_check"},
+        categories=["iam", "security"],
+        first_seen_at="2024-01-02T00:00:00Z",
+    )
+    finding2.add_resources([resource2])
+
+    backfill_resource_scan_summaries(str(scan.tenant_id), str(scan.id))
+    return finding1, finding2
+
+
+@pytest.fixture(scope="function")
+def latest_scan_finding_with_categories(
+    authenticated_client, providers_fixture, resources_fixture
+):
+    provider = providers_fixture[0]
+    tenant_id = str(providers_fixture[0].tenant_id)
+    resource = resources_fixture[0]
+    scan = Scan.objects.create(
+        name="latest completed scan with categories",
+        provider=provider,
+        trigger=Scan.TriggerChoices.MANUAL,
+        state=StateChoices.COMPLETED,
+        tenant_id=tenant_id,
+    )
+    finding = Finding.objects.create(
+        tenant_id=tenant_id,
+        uid="latest_finding_with_categories",
+        scan=scan,
+        delta="new",
+        status=Status.FAIL,
+        status_extended="test status",
+        impact=Severity.critical,
+        impact_extended="test impact",
+        severity=Severity.critical,
+        raw_result={"status": Status.FAIL},
+        check_id="genai_iam_check",
+        check_metadata={"CheckId": "genai_iam_check"},
+        categories=["gen-ai", "iam"],
+        first_seen_at="2024-01-02T00:00:00Z",
+    )
+    finding.add_resources([resource])
+    backfill_resource_scan_summaries(tenant_id, str(scan.id))
+    backfill_scan_category_summaries(tenant_id, str(scan.id))
+    return finding
+
+
 @pytest.fixture(scope="function")
 def latest_scan_resource(authenticated_client, providers_fixture):
     provider = providers_fixture[0]
@@ -1573,10 +1691,152 @@ def attack_paths_graph_stub_classes():
     )
 
 
+@pytest.fixture
+def create_attack_surface_overview():
+    def _create(tenant, scan, attack_surface_type, total=10, failed=5, muted_failed=2):
+        return AttackSurfaceOverview.objects.create(
+            tenant=tenant,
+            scan=scan,
+            attack_surface_type=attack_surface_type,
+            total_findings=total,
+            failed_findings=failed,
+            muted_failed_findings=muted_failed,
+        )
+
+    return _create
+
+
+@pytest.fixture
+def create_scan_category_summary():
+    def _create(
+        tenant,
+        scan,
+        category,
+        severity,
+        total_findings=10,
+        failed_findings=5,
+        new_failed_findings=2,
+    ):
+        return ScanCategorySummary.objects.create(
+            tenant=tenant,
+            scan=scan,
+            category=category,
+            severity=severity,
+            total_findings=total_findings,
+            failed_findings=failed_findings,
+            new_failed_findings=new_failed_findings,
+        )
+
+    return _create
+
+
+@pytest.fixture
 def get_authorization_header(access_token: str) -> dict:
     return {"Authorization": f"Bearer {access_token}"}
 
 
+@pytest.fixture
+def provider_compliance_scores_fixture(
+    tenants_fixture, providers_fixture, scans_fixture
+):
+    """Create ProviderComplianceScore entries for compliance watchlist tests."""
+    tenant = tenants_fixture[0]
+    provider1, provider2, *_ = providers_fixture
+    scan1, _, scan3 = scans_fixture
+
+    scan1.completed_at = datetime.now(timezone.utc) - timedelta(hours=1)
+    scan1.save()
+    scan3.state = StateChoices.COMPLETED
+    scan3.completed_at = datetime.now(timezone.utc)
+    scan3.save()
+
+    scores = [
+        ProviderComplianceScore.objects.create(
+            tenant_id=tenant.id,
+            provider=provider1,
+            scan=scan1,
+            compliance_id="aws_cis_2.0",
+            requirement_id="req_1",
+            requirement_status=StatusChoices.PASS,
+            scan_completed_at=scan1.completed_at,
+        ),
+        ProviderComplianceScore.objects.create(
+            tenant_id=tenant.id,
+            provider=provider1,
+            scan=scan1,
+            compliance_id="aws_cis_2.0",
+            requirement_id="req_2",
+            requirement_status=StatusChoices.FAIL,
+            scan_completed_at=scan1.completed_at,
+        ),
+        ProviderComplianceScore.objects.create(
+            tenant_id=tenant.id,
+            provider=provider1,
+            scan=scan1,
+            compliance_id="aws_cis_2.0",
+            requirement_id="req_3",
+            requirement_status=StatusChoices.MANUAL,
+            scan_completed_at=scan1.completed_at,
+        ),
+        ProviderComplianceScore.objects.create(
+            tenant_id=tenant.id,
+            provider=provider2,
+            scan=scan3,
+            compliance_id="aws_cis_2.0",
+            requirement_id="req_1",
+            requirement_status=StatusChoices.FAIL,
+            scan_completed_at=scan3.completed_at,
+        ),
+        ProviderComplianceScore.objects.create(
+            tenant_id=tenant.id,
+            provider=provider2,
+            scan=scan3,
+            compliance_id="aws_cis_2.0",
+            requirement_id="req_2",
+            requirement_status=StatusChoices.PASS,
+            scan_completed_at=scan3.completed_at,
+        ),
+        ProviderComplianceScore.objects.create(
+            tenant_id=tenant.id,
+            provider=provider1,
+            scan=scan1,
+            compliance_id="gdpr_aws",
+            requirement_id="gdpr_req_1",
+            requirement_status=StatusChoices.PASS,
+            scan_completed_at=scan1.completed_at,
+        ),
+    ]
+
+    return scores
+
+
+@pytest.fixture
+def tenant_compliance_summary_fixture(tenants_fixture):
+    """Create TenantComplianceSummary entries for compliance watchlist tests."""
+    tenant = tenants_fixture[0]
+
+    summaries = [
+        TenantComplianceSummary.objects.create(
+            tenant_id=tenant.id,
+            compliance_id="aws_cis_2.0",
+            requirements_passed=1,
+            requirements_failed=2,
+            requirements_manual=1,
+            total_requirements=4,
+        ),
+        TenantComplianceSummary.objects.create(
+            tenant_id=tenant.id,
+            compliance_id="gdpr_aws",
+            requirements_passed=5,
+            requirements_failed=0,
+            requirements_manual=2,
+            total_requirements=7,
+        ),
+    ]
+
+    return summaries
+
+
 def pytest_collection_modifyitems(items):
     """Ensure test_rbac.py is executed first."""
     items.sort(key=lambda item: 0 if "test_rbac.py" in item.nodeid else 1)

api/src/backend/tasks/beat.py

@@ -68,4 +68,5 @@ def schedule_provider_scan(provider_instance: Provider):
             "tenant_id": str(provider_instance.tenant_id),
             "provider_id": provider_id,
         },
+        countdown=5,  # Avoid race conditions between the worker and the database
     )

api/src/backend/tasks/jobs/backfill.py

@@ -1,17 +1,39 @@
 from collections import defaultdict
+from datetime import timedelta
 
-from api.db_router import READ_REPLICA_ALIAS
-from api.db_utils import rls_transaction
+from celery.utils.log import get_task_logger
+from django.db.models import Sum
+from django.utils import timezone
+from tasks.jobs.queries import (
+    COMPLIANCE_UPSERT_PROVIDER_SCORE_SQL,
+    COMPLIANCE_UPSERT_TENANT_SUMMARY_ALL_SQL,
+)
+from tasks.jobs.scan import aggregate_category_counts
+
+from api.db_router import READ_REPLICA_ALIAS, MainRouter
+from api.db_utils import (
+    POSTGRES_TENANT_VAR,
+    SET_CONFIG_QUERY,
+    psycopg_connection,
+    rls_transaction,
+)
 from api.models import (
     ComplianceOverviewSummary,
     ComplianceRequirementOverview,
+    DailySeveritySummary,
+    Finding,
+    ProviderComplianceScore,
     Resource,
     ResourceFindingMapping,
     ResourceScanSummary,
     Scan,
+    ScanCategorySummary,
+    ScanSummary,
     StateChoices,
 )
 
+logger = get_task_logger(__name__)
+
 
 def backfill_resource_scan_summaries(tenant_id: str, scan_id: str):
     with rls_transaction(tenant_id, using=READ_REPLICA_ALIAS):
@@ -175,3 +197,271 @@ def backfill_compliance_summaries(tenant_id: str, scan_id: str):
             )
 
     return {"status": "backfilled", "inserted": len(summary_objects)}
+
+
+def backfill_daily_severity_summaries(tenant_id: str, days: int = None):
+    """
+    Backfill DailySeveritySummary from completed scans.
+    Groups by provider+date, keeps latest scan per day.
+    """
+    created_count = 0
+    updated_count = 0
+
+    with rls_transaction(tenant_id, using=READ_REPLICA_ALIAS):
+        scan_filter = {
+            "tenant_id": tenant_id,
+            "state": StateChoices.COMPLETED,
+            "completed_at__isnull": False,
+        }
+
+        if days is not None:
+            cutoff_date = timezone.now() - timedelta(days=days)
+            scan_filter["completed_at__gte"] = cutoff_date
+
+        completed_scans = (
+            Scan.objects.filter(**scan_filter)
+            .order_by("provider_id", "-completed_at")
+            .values("id", "provider_id", "completed_at")
+        )
+
+        if not completed_scans:
+            return {"status": "no scans to backfill"}
+
+        # Keep only latest scan per provider/day
+        latest_scans_by_day = {}
+        for scan in completed_scans:
+            key = (scan["provider_id"], scan["completed_at"].date())
+            if key not in latest_scans_by_day:
+                latest_scans_by_day[key] = scan
+
+    # Process each provider/day
+    for (provider_id, scan_date), scan in latest_scans_by_day.items():
+        scan_id = scan["id"]
+
+        with rls_transaction(tenant_id, using=READ_REPLICA_ALIAS):
+            severity_totals = (
+                ScanSummary.objects.filter(
+                    tenant_id=tenant_id,
+                    scan_id=scan_id,
+                )
+                .values("severity")
+                .annotate(total_fail=Sum("fail"), total_muted=Sum("muted"))
+            )
+
+            severity_data = {
+                "critical": 0,
+                "high": 0,
+                "medium": 0,
+                "low": 0,
+                "informational": 0,
+                "muted": 0,
+            }
+
+            for row in severity_totals:
+                severity = row["severity"]
+                if severity in severity_data:
+                    severity_data[severity] = row["total_fail"] or 0
+                severity_data["muted"] += row["total_muted"] or 0
+
+        with rls_transaction(tenant_id):
+            _, created = DailySeveritySummary.objects.update_or_create(
+                tenant_id=tenant_id,
+                provider_id=provider_id,
+                date=scan_date,
+                defaults={
+                    "scan_id": scan_id,
+                    "critical": severity_data["critical"],
+                    "high": severity_data["high"],
+                    "medium": severity_data["medium"],
+                    "low": severity_data["low"],
+                    "informational": severity_data["informational"],
+                    "muted": severity_data["muted"],
+                },
+            )
+
+            if created:
+                created_count += 1
+            else:
+                updated_count += 1
+
+    return {
+        "status": "backfilled",
+        "created": created_count,
+        "updated": updated_count,
+        "total_days": len(latest_scans_by_day),
+    }
+
+
+def backfill_scan_category_summaries(tenant_id: str, scan_id: str):
+    """
+    Backfill ScanCategorySummary for a completed scan.
+
+    Aggregates category counts from all findings in the scan and creates
+    one ScanCategorySummary row per (category, severity) combination.
+
+    Args:
+        tenant_id: Target tenant UUID
+        scan_id: Scan UUID to backfill
+
+    Returns:
+        dict: Status indicating whether backfill was performed
+    """
+    with rls_transaction(tenant_id, using=READ_REPLICA_ALIAS):
+        if ScanCategorySummary.objects.filter(
+            tenant_id=tenant_id, scan_id=scan_id
+        ).exists():
+            return {"status": "already backfilled"}
+
+        if not Scan.objects.filter(
+            tenant_id=tenant_id,
+            id=scan_id,
+            state__in=(StateChoices.COMPLETED, StateChoices.FAILED),
+        ).exists():
+            return {"status": "scan is not completed"}
+
+        category_counts: dict[tuple[str, str], dict[str, int]] = {}
+        for finding in Finding.all_objects.filter(
+            tenant_id=tenant_id, scan_id=scan_id
+        ).values("categories", "severity", "status", "delta", "muted"):
+            aggregate_category_counts(
+                categories=finding.get("categories") or [],
+                severity=finding.get("severity"),
+                status=finding.get("status"),
+                delta=finding.get("delta"),
+                muted=finding.get("muted", False),
+                cache=category_counts,
+            )
+
+        if not category_counts:
+            return {"status": "no categories to backfill"}
+
+    category_summaries = [
+        ScanCategorySummary(
+            tenant_id=tenant_id,
+            scan_id=scan_id,
+            category=category,
+            severity=severity,
+            total_findings=counts["total"],
+            failed_findings=counts["failed"],
+            new_failed_findings=counts["new_failed"],
+        )
+        for (category, severity), counts in category_counts.items()
+    ]
+
+    with rls_transaction(tenant_id):
+        ScanCategorySummary.objects.bulk_create(
+            category_summaries, batch_size=500, ignore_conflicts=True
+        )
+
+    return {"status": "backfilled", "categories_count": len(category_counts)}
+
+
+def backfill_provider_compliance_scores(tenant_id: str) -> dict:
+    """
+    Backfill ProviderComplianceScore from latest completed scan per provider.
+
+    For each provider with completed scans, finds the most recent scan and
+    upserts compliance requirement statuses with FAIL-dominant aggregation.
+
+    Args:
+        tenant_id: Target tenant UUID
+
+    Returns:
+        dict: Statistics about the backfill operation
+    """
+    with rls_transaction(tenant_id, using=READ_REPLICA_ALIAS):
+        completed_scans = Scan.all_objects.filter(
+            tenant_id=tenant_id,
+            state=StateChoices.COMPLETED,
+            completed_at__isnull=False,
+        )
+        if not completed_scans.exists():
+            return {"status": "no completed scans"}
+
+        existing_providers = set(
+            ProviderComplianceScore.objects.filter(tenant_id=tenant_id)
+            .values_list("provider_id", flat=True)
+            .distinct()
+        )
+
+        if existing_providers:
+            completed_scans = completed_scans.exclude(
+                provider_id__in=existing_providers
+            )
+
+        scan_info = list(
+            completed_scans.order_by("provider_id", "-completed_at")
+            .distinct("provider_id")
+            .values("id", "provider_id", "completed_at")
+        )
+
+        if not scan_info:
+            return {"status": "no scans to process"}
+
+    total_upserted = 0
+    providers_processed = 0
+    providers_skipped = 0
+
+    for scan in scan_info:
+        provider_id = scan["provider_id"]
+
+        scan_id = scan["id"]
+
+        try:
+            with psycopg_connection(MainRouter.default_db) as connection:
+                connection.autocommit = False
+                try:
+                    with connection.cursor() as cursor:
+                        cursor.execute(
+                            SET_CONFIG_QUERY, [POSTGRES_TENANT_VAR, tenant_id]
+                        )
+                        cursor.execute(
+                            COMPLIANCE_UPSERT_PROVIDER_SCORE_SQL,
+                            [tenant_id, str(scan_id)],
+                        )
+                        upserted = cursor.rowcount
+                    connection.commit()
+                    total_upserted += upserted
+                    providers_processed += 1
+                except Exception:
+                    connection.rollback()
+                    raise
+        except Exception as e:
+            providers_skipped += 1
+            logger.exception(
+                "Error backfilling provider %s for tenant %s: %s",
+                provider_id,
+                tenant_id,
+                e,
+            )
+
+    # Recalculate tenant summary after all providers are backfilled
+    if providers_processed > 0:
+        with psycopg_connection(MainRouter.default_db) as connection:
+            connection.autocommit = False
+            try:
+                with connection.cursor() as cursor:
+                    cursor.execute(SET_CONFIG_QUERY, [POSTGRES_TENANT_VAR, tenant_id])
+                    # Advisory lock to prevent race conditions
+                    cursor.execute(
+                        "SELECT pg_advisory_xact_lock(hashtext(%s))", [tenant_id]
+                    )
+                    cursor.execute(
+                        COMPLIANCE_UPSERT_TENANT_SUMMARY_ALL_SQL,
+                        [tenant_id, tenant_id],
+                    )
+                    tenant_summary_count = cursor.rowcount
+                connection.commit()
+            except Exception:
+                connection.rollback()
+                raise
+    else:
+        tenant_summary_count = 0
+
+    return {
+        "status": "backfilled",
+        "providers_processed": providers_processed,
+        "providers_skipped": providers_skipped,
+        "total_upserted": total_upserted,
+        "tenant_summary_count": tenant_summary_count,
+    }

api/src/backend/tasks/jobs/export.py

@@ -27,6 +27,7 @@ from prowler.lib.outputs.compliance.c5.c5_gcp import GCPC5
 from prowler.lib.outputs.compliance.ccc.ccc_aws import CCC_AWS
 from prowler.lib.outputs.compliance.ccc.ccc_azure import CCC_Azure
 from prowler.lib.outputs.compliance.ccc.ccc_gcp import CCC_GCP
+from prowler.lib.outputs.compliance.cis.cis_alibabacloud import AlibabaCloudCIS
 from prowler.lib.outputs.compliance.cis.cis_aws import AWSCIS
 from prowler.lib.outputs.compliance.cis.cis_azure import AzureCIS
 from prowler.lib.outputs.compliance.cis.cis_gcp import GCPCIS
@@ -50,6 +51,9 @@ from prowler.lib.outputs.compliance.mitre_attack.mitre_attack_azure import (
     AzureMitreAttack,
 )
 from prowler.lib.outputs.compliance.mitre_attack.mitre_attack_gcp import GCPMitreAttack
+from prowler.lib.outputs.compliance.prowler_threatscore.prowler_threatscore_alibaba import (
+    ProwlerThreatScoreAlibaba,
+)
 from prowler.lib.outputs.compliance.prowler_threatscore.prowler_threatscore_aws import (
     ProwlerThreatScoreAWS,
 )
@@ -128,6 +132,13 @@ COMPLIANCE_CLASS_MAP = {
     "oraclecloud": [
         (lambda name: name.startswith("cis_"), OracleCloudCIS),
     ],
+    "alibabacloud": [
+        (lambda name: name.startswith("cis_"), AlibabaCloudCIS),
+        (
+            lambda name: name == "prowler_threatscore_alibabacloud",
+            ProwlerThreatScoreAlibaba,
+        ),
+    ],
 }
 
 

api/src/backend/tasks/jobs/integrations.py

@@ -19,6 +19,9 @@ from prowler.providers.aws.aws_provider import AwsProvider
 from prowler.providers.aws.lib.s3.s3 import S3
 from prowler.providers.aws.lib.security_hub.security_hub import SecurityHub
 from prowler.providers.common.models import Connection
+from prowler.providers.aws.lib.security_hub.exceptions.exceptions import (
+    SecurityHubNoEnabledRegionsError,
+)
 
 logger = get_task_logger(__name__)
 
@@ -222,8 +225,9 @@ def get_security_hub_client_from_integration(
         )
         return True, security_hub
     else:
-        # Reset regions information if connection fails
+        # Reset regions information if connection fails and integration is not connected
         with rls_transaction(tenant_id, using=MainRouter.default_db):
+            integration.connected = False
             integration.configuration["regions"] = {}
             integration.save()
 
@@ -330,15 +334,18 @@ def upload_security_hub_integration(
                                 )
 
                                 if not connected:
-                                    logger.error(
-                                        f"Security Hub connection failed for integration {integration.id}: "
-                                        f"{security_hub.error}"
-                                    )
-                                    with rls_transaction(
-                                        tenant_id, using=MainRouter.default_db
+                                    if isinstance(
+                                        security_hub.error,
+                                        SecurityHubNoEnabledRegionsError,
                                     ):
-                                        integration.connected = False
-                                        integration.save()
+                                        logger.warning(
+                                            f"Security Hub integration {integration.id} has no enabled regions"
+                                        )
+                                    else:
+                                        logger.error(
+                                            f"Security Hub connection failed for integration {integration.id}: "
+                                            f"{security_hub.error}"
+                                        )
                                     break  # Skip this integration
 
                                 security_hub_client = security_hub
@@ -409,22 +416,16 @@ def upload_security_hub_integration(
                             logger.warning(
                                 f"Failed to archive previous findings: {str(archive_error)}"
                             )
-
             except Exception as e:
                 logger.error(
                     f"Security Hub integration {integration.id} failed: {str(e)}"
                 )
-                continue
 
         result = integration_executions == len(integrations)
         if result:
             logger.info(
                 f"All Security Hub integrations completed successfully for provider {provider_id}"
             )
-        else:
-            logger.error(
-                f"Some Security Hub integrations failed for provider {provider_id}"
-            )
 
         return result
 

api/src/backend/tasks/jobs/lighthouse_providers.py

@@ -2,6 +2,8 @@ from typing import Dict
 
 import boto3
 import openai
+from botocore import UNSIGNED
+from botocore.config import Config
 from botocore.exceptions import BotoCoreError, ClientError
 from celery.utils.log import get_task_logger
 
@@ -9,6 +11,74 @@ from api.models import LighthouseProviderConfiguration, LighthouseProviderModels
 
 logger = get_task_logger(__name__)
 
+# OpenAI model prefixes to exclude from Lighthouse model selection.
+# These models don't support text chat completions and tool calling.
+EXCLUDED_OPENAI_MODEL_PREFIXES = (
+    "dall-e",  # Image generation
+    "whisper",  # Audio transcription
+    "tts-",  # Text-to-speech (tts-1, tts-1-hd, etc.)
+    "sora",  # Text-to-video (sora-2, sora-2-pro, etc.)
+    "text-embedding",  # Embeddings
+    "embedding",  # Embeddings (alternative naming)
+    "text-moderation",  # Content moderation
+    "omni-moderation",  # Content moderation
+    "text-davinci",  # Legacy completion models
+    "text-curie",  # Legacy completion models
+    "text-babbage",  # Legacy completion models
+    "text-ada",  # Legacy completion models
+    "davinci",  # Legacy completion models
+    "curie",  # Legacy completion models
+    "babbage",  # Legacy completion models
+    "ada",  # Legacy completion models
+    "computer-use",  # Computer control agent
+    "gpt-image",  # Image generation
+    "gpt-audio",  # Audio models
+    "gpt-realtime",  # Realtime voice API
+)
+
+# OpenAI model substrings to exclude (patterns that can appear anywhere in model ID).
+# These patterns identify non-chat model variants.
+EXCLUDED_OPENAI_MODEL_SUBSTRINGS = (
+    "-audio-",  # Audio preview models (gpt-4o-audio-preview, etc.)
+    "-realtime-",  # Realtime preview models (gpt-4o-realtime-preview, etc.)
+    "-transcribe",  # Transcription models (gpt-4o-transcribe, etc.)
+    "-tts",  # TTS models (gpt-4o-mini-tts)
+    "-instruct",  # Legacy instruct models (gpt-3.5-turbo-instruct, etc.)
+)
+
+
+def _extract_error_message(e: Exception) -> str:
+    """
+    Extract a user-friendly error message from various exception types.
+
+    This function handles exceptions from different providers (OpenAI, AWS Bedrock)
+    and extracts the most relevant error message for display to users.
+
+    Args:
+        e: The exception to extract a message from.
+
+    Returns:
+        str: A user-friendly error message.
+    """
+    # For OpenAI SDK errors (>= v1.0)
+    # OpenAI exceptions have a 'body' attribute with error details
+    if hasattr(e, "body") and isinstance(e.body, dict):
+        if "message" in e.body:
+            return e.body["message"]
+        # Sometimes nested under 'error' key
+        if "error" in e.body and isinstance(e.body["error"], dict):
+            return e.body["error"].get("message", str(e))
+
+    # For boto3 ClientError
+    # Boto3 exceptions have a 'response' attribute with error details
+    if hasattr(e, "response") and isinstance(e.response, dict):
+        error_info = e.response.get("Error", {})
+        if error_info.get("Message"):
+            return error_info["Message"]
+
+    # Fallback to string representation for unknown error types
+    return str(e)
+
 
 def _extract_openai_api_key(
     provider_cfg: LighthouseProviderConfiguration,
@@ -56,21 +126,39 @@ def _extract_bedrock_credentials(
     """
     Safely extract AWS Bedrock credentials from a provider configuration.
 
+    Supports two authentication methods:
+    1. AWS access key + secret key + region
+    2. Bedrock API key (bearer token) + region
+
     Args:
         provider_cfg (LighthouseProviderConfiguration): The provider configuration instance
             containing the credentials.
 
     Returns:
-        Dict[str, str] | None: Dictionary with 'access_key_id', 'secret_access_key', and
-            'region' if present and valid, otherwise None.
+        Dict[str, str] | None: Dictionary with either:
+            - 'access_key_id', 'secret_access_key', and 'region' for access key auth
+            - 'api_key' and 'region' for API key (bearer token) auth
+            Returns None if credentials are invalid or missing.
     """
     creds = provider_cfg.credentials_decoded
     if not isinstance(creds, dict):
         return None
 
+    region = creds.get("region")
+    if not isinstance(region, str) or not region:
+        return None
+
+    # Check for API key authentication first
+    api_key = creds.get("api_key")
+    if isinstance(api_key, str) and api_key:
+        return {
+            "api_key": api_key,
+            "region": region,
+        }
+
+    # Fall back to access key authentication
     access_key_id = creds.get("access_key_id")
     secret_access_key = creds.get("secret_access_key")
-    region = creds.get("region")
 
     # Validate all required fields are present and are strings
     if (
@@ -78,8 +166,6 @@ def _extract_bedrock_credentials(
         or not access_key_id
         or not isinstance(secret_access_key, str)
         or not secret_access_key
-        or not isinstance(region, str)
-        or not region
     ):
         return None
 
@@ -90,6 +176,51 @@ def _extract_bedrock_credentials(
     }
 
 
+def _create_bedrock_client(
+    bedrock_creds: Dict[str, str], service_name: str = "bedrock"
+):
+    """
+    Create a boto3 Bedrock client with the appropriate authentication method.
+
+    Supports two authentication methods:
+    1. API key (bearer token) - uses unsigned requests with Authorization header
+    2. AWS access key + secret key - uses standard SigV4 signing
+
+    Args:
+        bedrock_creds: Dictionary with either:
+            - 'api_key' and 'region' for API key (bearer token) auth
+            - 'access_key_id', 'secret_access_key', and 'region' for access key auth
+        service_name: The Bedrock service name. Use 'bedrock' for control plane
+            operations (list_foundation_models, etc.) or 'bedrock-runtime' for
+            inference operations.
+
+    Returns:
+        boto3 client configured for the specified Bedrock service.
+    """
+    region = bedrock_creds["region"]
+
+    if "api_key" in bedrock_creds:
+        bearer_token = bedrock_creds["api_key"]
+        client = boto3.client(
+            service_name=service_name,
+            region_name=region,
+            config=Config(signature_version=UNSIGNED),
+        )
+
+        def inject_bearer_token(request, **kwargs):
+            request.headers["Authorization"] = f"Bearer {bearer_token}"
+
+        client.meta.events.register("before-send.*.*", inject_bearer_token)
+        return client
+
+    return boto3.client(
+        service_name=service_name,
+        region_name=region,
+        aws_access_key_id=bedrock_creds["access_key_id"],
+        aws_secret_access_key=bedrock_creds["secret_access_key"],
+    )
+
+
 def check_lighthouse_provider_connection(provider_config_id: str) -> Dict:
     """
     Validate a Lighthouse provider configuration by calling the provider API and
@@ -141,12 +272,7 @@ def check_lighthouse_provider_connection(provider_config_id: str) -> Dict:
                 }
 
             # Test connection by listing foundation models
-            bedrock_client = boto3.client(
-                "bedrock",
-                aws_access_key_id=bedrock_creds["access_key_id"],
-                aws_secret_access_key=bedrock_creds["secret_access_key"],
-                region_name=bedrock_creds["region"],
-            )
+            bedrock_client = _create_bedrock_client(bedrock_creds)
             _ = bedrock_client.list_foundation_models()
 
         elif (
@@ -179,32 +305,54 @@ def check_lighthouse_provider_connection(provider_config_id: str) -> Dict:
         return {"connected": True, "error": None}
 
     except Exception as e:
+        error_message = _extract_error_message(e)
         logger.warning(
-            "%s connection check failed: %s", provider_cfg.provider_type, str(e)
+            "%s connection check failed: %s", provider_cfg.provider_type, error_message
         )
         provider_cfg.is_active = False
         provider_cfg.save()
-        return {"connected": False, "error": str(e)}
+        return {"connected": False, "error": error_message}
 
 
 def _fetch_openai_models(api_key: str) -> Dict[str, str]:
     """
     Fetch available models from OpenAI API.
 
+    Filters out models that don't support text input/output and tool calling,
+    such as image generation (DALL-E), audio transcription (Whisper),
+    text-to-speech (TTS), embeddings, and moderation models.
+
     Args:
         api_key: OpenAI API key for authentication.
 
     Returns:
         Dict mapping model_id to model_name. For OpenAI, both are the same
-        as the API doesn't provide separate display names.
+        as the API doesn't provide separate display names. Only includes
+        models that support text input, text output or tool calling.
 
     Raises:
         Exception: If the API call fails.
     """
     client = openai.OpenAI(api_key=api_key)
     models = client.models.list()
-    # OpenAI uses model.id for both ID and display name
-    return {m.id: m.id for m in getattr(models, "data", [])}
+
+    # Filter models to only include those supporting chat completions + tool calling
+    filtered_models = {}
+    for model in getattr(models, "data", []):
+        model_id = model.id
+
+        # Skip if model ID starts with excluded prefixes
+        if model_id.startswith(EXCLUDED_OPENAI_MODEL_PREFIXES):
+            continue
+
+        # Skip if model ID contains excluded substrings
+        if any(substring in model_id for substring in EXCLUDED_OPENAI_MODEL_SUBSTRINGS):
+            continue
+
+        # Include model (supports chat completions + tool calling)
+        filtered_models[model_id] = model_id
+
+    return filtered_models
 
 
 def _fetch_openai_compatible_models(base_url: str, api_key: str) -> Dict[str, str]:
@@ -232,105 +380,219 @@ def _fetch_openai_compatible_models(base_url: str, api_key: str) -> Dict[str, st
     return available_models
 
 
-def _fetch_bedrock_models(bedrock_creds: Dict[str, str]) -> Dict[str, str]:
+def _get_region_prefix(region: str) -> str:
     """
-    Fetch available models from AWS Bedrock with entitlement verification.
+    Determine geographic prefix for AWS region.
 
-    This function:
-    1. Lists foundation models with TEXT modality support
-    2. Lists inference profiles with TEXT modality support
-    3. Verifies user has entitlement access to each model
+    Examples: ap-south-1 -> apac, us-east-1 -> us, eu-west-1 -> eu
+    """
+    if region.startswith(("us-", "ca-", "sa-")):
+        return "us"
+    elif region.startswith("eu-"):
+        return "eu"
+    elif region.startswith("ap-"):
+        return "apac"
+    return "global"
 
-    Args:
-        bedrock_creds: Dictionary with 'access_key_id', 'secret_access_key', and 'region'.
+
+def _clean_inference_profile_name(profile_name: str) -> str:
+    """
+    Remove geographic prefix from inference profile name.
+
+    AWS includes geographic prefixes in profile names which are redundant
+    since the profile ID already contains this information.
+
+    Examples:
+        "APAC Anthropic Claude 3.5 Sonnet" -> "Anthropic Claude 3.5 Sonnet"
+        "GLOBAL Claude Sonnet 4.5" -> "Claude Sonnet 4.5"
+        "US Anthropic Claude 3 Haiku" -> "Anthropic Claude 3 Haiku"
+    """
+    prefixes = ["APAC ", "GLOBAL ", "US ", "EU ", "APAC-", "GLOBAL-", "US-", "EU-"]
+
+    for prefix in prefixes:
+        if profile_name.upper().startswith(prefix.upper()):
+            return profile_name[len(prefix) :].strip()
+
+    return profile_name
+
+
+def _supports_text_modality(input_modalities: list, output_modalities: list) -> bool:
+    """Check if model supports TEXT for both input and output."""
+    return "TEXT" in input_modalities and "TEXT" in output_modalities
+
+
+def _get_foundation_model_modalities(
+    bedrock_client, model_id: str
+) -> tuple[list, list] | None:
+    """
+    Fetch input and output modalities for a foundation model.
 
     Returns:
-        Dict mapping model_id to model_name for all accessible models.
-
-    Raises:
-        BotoCoreError, ClientError: If AWS API calls fail.
+        (input_modalities, output_modalities) or None if fetch fails
     """
-    bedrock_client = boto3.client(
-        "bedrock",
-        aws_access_key_id=bedrock_creds["access_key_id"],
-        aws_secret_access_key=bedrock_creds["secret_access_key"],
-        region_name=bedrock_creds["region"],
-    )
+    try:
+        model_info = bedrock_client.get_foundation_model(modelIdentifier=model_id)
+        model_details = model_info.get("modelDetails", {})
+        input_mods = model_details.get("inputModalities", [])
+        output_mods = model_details.get("outputModalities", [])
+        return (input_mods, output_mods)
+    except (BotoCoreError, ClientError) as e:
+        logger.debug("Could not fetch model details for %s: %s", model_id, str(e))
+        return None
 
-    models_to_check: Dict[str, str] = {}
 
-    # Step 1: Get foundation models with TEXT modality
+def _extract_foundation_model_ids(profile_models: list) -> list[str]:
+    """
+    Extract foundation model IDs from inference profile model ARNs.
+
+    Args:
+        profile_models: List of model references from inference profile
+
+    Returns:
+        List of foundation model IDs extracted from ARNs
+    """
+    model_ids = []
+    for model_ref in profile_models:
+        model_arn = model_ref.get("modelArn", "")
+        if "foundation-model/" in model_arn:
+            model_id = model_arn.split("foundation-model/")[1]
+            model_ids.append(model_id)
+    return model_ids
+
+
+def _build_inference_profile_map(
+    bedrock_client, region: str
+) -> Dict[str, tuple[str, str]]:
+    """
+    Build map of foundation_model_id -> best inference profile.
+
+    Returns:
+        Dict mapping foundation_model_id to (profile_id, profile_name)
+        Only includes profiles with TEXT modality support
+        Prefers region-matched profiles over others
+    """
+    region_prefix = _get_region_prefix(region)
+    model_to_profile: Dict[str, tuple[str, str]] = {}
+
+    try:
+        response = bedrock_client.list_inference_profiles()
+        profiles = response.get("inferenceProfileSummaries", [])
+
+        for profile in profiles:
+            profile_id = profile.get("inferenceProfileId")
+            profile_name = profile.get("inferenceProfileName")
+
+            if not profile_id or not profile_name:
+                continue
+
+            profile_models = profile.get("models", [])
+            if not profile_models:
+                continue
+
+            foundation_model_ids = _extract_foundation_model_ids(profile_models)
+            if not foundation_model_ids:
+                continue
+
+            modalities = _get_foundation_model_modalities(
+                bedrock_client, foundation_model_ids[0]
+            )
+            if not modalities:
+                continue
+
+            input_mods, output_mods = modalities
+            if not _supports_text_modality(input_mods, output_mods):
+                continue
+
+            is_preferred = profile_id.startswith(f"{region_prefix}.")
+            clean_name = _clean_inference_profile_name(profile_name)
+
+            for foundation_model_id in foundation_model_ids:
+                if foundation_model_id not in model_to_profile:
+                    model_to_profile[foundation_model_id] = (profile_id, clean_name)
+                elif is_preferred and not model_to_profile[foundation_model_id][
+                    0
+                ].startswith(f"{region_prefix}."):
+                    model_to_profile[foundation_model_id] = (profile_id, clean_name)
+
+    except (BotoCoreError, ClientError) as e:
+        logger.info("Could not fetch inference profiles in %s: %s", region, str(e))
+
+    return model_to_profile
+
+
+def _check_on_demand_availability(bedrock_client, model_id: str) -> bool:
+    """Check if an ON_DEMAND foundation model is entitled and available."""
+    try:
+        availability = bedrock_client.get_foundation_model_availability(
+            modelId=model_id
+        )
+        entitlement = availability.get("entitlementAvailability")
+        return entitlement == "AVAILABLE"
+    except (BotoCoreError, ClientError) as e:
+        logger.debug("Could not check availability for %s: %s", model_id, str(e))
+        return False
+
+
+def _fetch_bedrock_models(bedrock_creds: Dict[str, str]) -> Dict[str, str]:
+    """
+    Fetch available models from AWS Bedrock, preferring inference profiles over ON_DEMAND.
+
+    Strategy:
+    1. Build map of foundation_model -> best_inference_profile (with TEXT validation)
+    2. For each TEXT-capable foundation model:
+       - Use inference profile ID if available (preferred - better throughput)
+       - Fallback to foundation model ID if only ON_DEMAND available
+    3. Verify entitlement for ON_DEMAND models
+
+    Args:
+        bedrock_creds: Dict with 'region' and auth credentials
+
+    Returns:
+        Dict mapping model_id to model_name. IDs can be:
+        - Inference profile IDs (e.g., "apac.anthropic.claude-3-5-sonnet-20240620-v1:0")
+        - Foundation model IDs (e.g., "anthropic.claude-3-5-sonnet-20240620-v1:0")
+    """
+    bedrock_client = _create_bedrock_client(bedrock_creds)
+    region = bedrock_creds["region"]
+
+    model_to_profile = _build_inference_profile_map(bedrock_client, region)
+
     foundation_response = bedrock_client.list_foundation_models()
     model_summaries = foundation_response.get("modelSummaries", [])
 
-    for model in model_summaries:
-        # Check if model supports TEXT input and output modality
-        input_modalities = model.get("inputModalities", [])
-        output_modalities = model.get("outputModalities", [])
+    models_to_return: Dict[str, str] = {}
+    on_demand_models: set[str] = set()
 
-        if "TEXT" not in input_modalities or "TEXT" not in output_modalities:
+    for model in model_summaries:
+        input_mods = model.get("inputModalities", [])
+        output_mods = model.get("outputModalities", [])
+
+        if not _supports_text_modality(input_mods, output_mods):
             continue
 
         model_id = model.get("modelId")
-        if not model_id:
+        model_name = model.get("modelName")
+
+        if not model_id or not model_name:
             continue
 
-        inference_types = model.get("inferenceTypesSupported", [])
+        if model_id in model_to_profile:
+            profile_id, profile_name = model_to_profile[model_id]
+            models_to_return[profile_id] = profile_name
+        else:
+            inference_types = model.get("inferenceTypesSupported", [])
+            if "ON_DEMAND" in inference_types:
+                models_to_return[model_id] = model_name
+                on_demand_models.add(model_id)
 
-        # Only include models with ON_DEMAND inference support
-        if "ON_DEMAND" in inference_types:
-            models_to_check[model_id] = model["modelName"]
-
-    # Step 2: Get inference profiles
-    try:
-        inference_profiles_response = bedrock_client.list_inference_profiles()
-        inference_profiles = inference_profiles_response.get(
-            "inferenceProfileSummaries", []
-        )
-
-        for profile in inference_profiles:
-            # Check if profile supports TEXT modality
-            input_modalities = profile.get("inputModalities", [])
-            output_modalities = profile.get("outputModalities", [])
-
-            if "TEXT" not in input_modalities or "TEXT" not in output_modalities:
-                continue
-
-            profile_id = profile.get("inferenceProfileId")
-            if profile_id:
-                models_to_check[profile_id] = profile["inferenceProfileName"]
-
-    except (BotoCoreError, ClientError) as e:
-        logger.info(
-            "Could not fetch inference profiles in %s: %s",
-            bedrock_creds["region"],
-            str(e),
-        )
-
-    # Step 3: Verify entitlement availability for each model
     available_models: Dict[str, str] = {}
 
-    for model_id, model_name in models_to_check.items():
-        try:
-            availability = bedrock_client.get_foundation_model_availability(
-                modelId=model_id
-            )
-
-            entitlement = availability.get("entitlementAvailability")
-
-            # Only include models user has access to
-            if entitlement == "AVAILABLE":
+    for model_id, model_name in models_to_return.items():
+        if model_id in on_demand_models:
+            if _check_on_demand_availability(bedrock_client, model_id):
                 available_models[model_id] = model_name
-            else:
-                logger.debug(
-                    "Skipping model %s - entitlement status: %s", model_id, entitlement
-                )
-
-        except (BotoCoreError, ClientError) as e:
-            logger.debug(
-                "Could not check availability for model %s: %s", model_id, str(e)
-            )
-            continue
+        else:
+            available_models[model_id] = model_name
 
     return available_models
 
@@ -359,7 +621,6 @@ def refresh_lighthouse_provider_models(provider_config_id: str) -> Dict:
     provider_cfg = LighthouseProviderConfiguration.objects.get(pk=provider_config_id)
     fetched_models: Dict[str, str] = {}
 
-    # Fetch models from the appropriate provider
     try:
         if (
             provider_cfg.provider_type
@@ -414,12 +675,13 @@ def refresh_lighthouse_provider_models(provider_config_id: str) -> Dict:
             }
 
     except Exception as e:
+        error_message = _extract_error_message(e)
         logger.warning(
             "Unexpected error refreshing %s models: %s",
             provider_cfg.provider_type,
-            str(e),
+            error_message,
         )
-        return {"created": 0, "updated": 0, "deleted": 0, "error": str(e)}
+        return {"created": 0, "updated": 0, "deleted": 0, "error": error_message}
 
     # Upsert models into the catalog
     created = 0

api/src/backend/tasks/jobs/queries.py

@@ -0,0 +1,134 @@
+"""
+Shared SQL queries for tasks.
+
+This module centralizes raw SQL queries used across multiple task modules
+to ensure consistency and maintainability.
+"""
+
+# =============================================================================
+# COMPLIANCE SCORE QUERIES
+# =============================================================================
+
+# Upsert provider compliance scores from a scan's compliance requirements.
+# Uses FAIL-dominant aggregation: FAIL > MANUAL > PASS
+# Parameters: [tenant_id, scan_id]
+COMPLIANCE_UPSERT_PROVIDER_SCORE_SQL = """
+    INSERT INTO provider_compliance_scores
+        (id, tenant_id, provider_id, scan_id, compliance_id, requirement_id,
+         requirement_status, scan_completed_at)
+    SELECT
+        gen_random_uuid(),
+        agg.tenant_id,
+        agg.provider_id,
+        agg.scan_id,
+        agg.compliance_id,
+        agg.requirement_id,
+        agg.requirement_status,
+        agg.completed_at
+    FROM (
+        SELECT DISTINCT ON (cro.compliance_id, cro.requirement_id)
+            cro.tenant_id,
+            s.provider_id,
+            cro.scan_id,
+            cro.compliance_id,
+            cro.requirement_id,
+            (CASE
+                WHEN bool_or(cro.requirement_status = 'FAIL')
+                    OVER (PARTITION BY cro.compliance_id, cro.requirement_id) THEN 'FAIL'
+                WHEN bool_or(cro.requirement_status = 'MANUAL')
+                    OVER (PARTITION BY cro.compliance_id, cro.requirement_id) THEN 'MANUAL'
+                ELSE 'PASS'
+            END)::status as requirement_status,
+            s.completed_at
+        FROM compliance_requirements_overviews cro
+        JOIN scans s ON s.id = cro.scan_id
+        WHERE cro.tenant_id = %s AND cro.scan_id = %s
+        ORDER BY cro.compliance_id, cro.requirement_id
+    ) agg
+    ON CONFLICT (tenant_id, provider_id, compliance_id, requirement_id)
+    DO UPDATE SET
+        requirement_status = EXCLUDED.requirement_status,
+        scan_id = EXCLUDED.scan_id,
+        scan_completed_at = EXCLUDED.scan_completed_at
+    WHERE EXCLUDED.scan_completed_at > provider_compliance_scores.scan_completed_at
+"""
+
+# Upsert tenant compliance summary for specific compliance IDs.
+# Aggregates across all providers with FAIL-dominant logic at requirement level.
+# Parameters: [tenant_id, tenant_id, compliance_ids_array]
+COMPLIANCE_UPSERT_TENANT_SUMMARY_SQL = """
+    INSERT INTO tenant_compliance_summaries
+        (id, tenant_id, compliance_id,
+         requirements_passed, requirements_failed, requirements_manual,
+         total_requirements, updated_at)
+    SELECT
+        gen_random_uuid(),
+        %s as tenant_id,
+        compliance_id,
+        COUNT(*) FILTER (WHERE req_status = 'PASS') as requirements_passed,
+        COUNT(*) FILTER (WHERE req_status = 'FAIL') as requirements_failed,
+        COUNT(*) FILTER (WHERE req_status = 'MANUAL') as requirements_manual,
+        COUNT(*) as total_requirements,
+        NOW() as updated_at
+    FROM (
+        SELECT
+            compliance_id,
+            requirement_id,
+            CASE
+                WHEN bool_or(requirement_status = 'FAIL') THEN 'FAIL'
+                WHEN bool_or(requirement_status = 'MANUAL') THEN 'MANUAL'
+                ELSE 'PASS'
+            END as req_status
+        FROM provider_compliance_scores
+        WHERE tenant_id = %s AND compliance_id = ANY(%s)
+        GROUP BY compliance_id, requirement_id
+    ) req_agg
+    GROUP BY compliance_id
+    ON CONFLICT (tenant_id, compliance_id)
+    DO UPDATE SET
+        requirements_passed = EXCLUDED.requirements_passed,
+        requirements_failed = EXCLUDED.requirements_failed,
+        requirements_manual = EXCLUDED.requirements_manual,
+        total_requirements = EXCLUDED.total_requirements,
+        updated_at = NOW()
+"""
+
+# Upsert tenant compliance summary for ALL compliance IDs in tenant.
+# Used by backfill when recalculating entire tenant summary.
+# Parameters: [tenant_id, tenant_id]
+COMPLIANCE_UPSERT_TENANT_SUMMARY_ALL_SQL = """
+    INSERT INTO tenant_compliance_summaries
+        (id, tenant_id, compliance_id,
+         requirements_passed, requirements_failed, requirements_manual,
+         total_requirements, updated_at)
+    SELECT
+        gen_random_uuid(),
+        %s as tenant_id,
+        compliance_id,
+        COUNT(*) FILTER (WHERE req_status = 'PASS') as requirements_passed,
+        COUNT(*) FILTER (WHERE req_status = 'FAIL') as requirements_failed,
+        COUNT(*) FILTER (WHERE req_status = 'MANUAL') as requirements_manual,
+        COUNT(*) as total_requirements,
+        NOW() as updated_at
+    FROM (
+        SELECT
+            compliance_id,
+            requirement_id,
+            CASE
+                WHEN bool_or(requirement_status = 'FAIL') THEN 'FAIL'
+                WHEN bool_or(requirement_status = 'MANUAL') THEN 'MANUAL'
+                ELSE 'PASS'
+            END as req_status
+        FROM provider_compliance_scores
+        WHERE tenant_id = %s
+        GROUP BY compliance_id, requirement_id
+    ) req_agg
+    GROUP BY compliance_id
+    ON CONFLICT (tenant_id, compliance_id)
+    DO UPDATE SET
+        requirements_passed = EXCLUDED.requirements_passed,
+        requirements_failed = EXCLUDED.requirements_failed,
+        requirements_manual = EXCLUDED.requirements_manual,
+        total_requirements = EXCLUDED.total_requirements,
+        updated_at = NOW()
+"""

api/src/backend/tasks/jobs/report.py

@@ -243,15 +243,28 @@ def _safe_getattr(obj, attr: str, default: str = "N/A") -> str:
 
 
 def _create_info_table_style() -> TableStyle:
-    """Create a reusable table style for information/metadata tables."""
+    """Create a reusable table style for information/metadata tables.
+
+    ReportLab TableStyle coordinate system:
+        - Format: (COMMAND, (start_col, start_row), (end_col, end_row), value)
+        - Coordinates use (column, row) format, starting at (0, 0) for top-left cell
+        - Negative indices work like Python slicing: -1 means "last row/column"
+        - (0, 0) to (0, -1) = entire first column (all rows)
+        - (0, 0) to (-1, 0) = entire first row (all columns)
+        - (0, 0) to (-1, -1) = entire table
+        - Styles are applied in order; later rules override earlier ones
+    """
     return TableStyle(
         [
+            # Column 0 (labels): blue background with white text
             ("BACKGROUND", (0, 0), (0, -1), COLOR_BLUE),
             ("TEXTCOLOR", (0, 0), (0, -1), COLOR_WHITE),
             ("FONTNAME", (0, 0), (0, -1), "FiraCode"),
+            # Column 1 (values): light blue background with gray text
             ("BACKGROUND", (1, 0), (1, -1), COLOR_BG_BLUE),
             ("TEXTCOLOR", (1, 0), (1, -1), COLOR_GRAY),
             ("FONTNAME", (1, 0), (1, -1), "PlusJakartaSans"),
+            # Apply to entire table
             ("ALIGN", (0, 0), (-1, -1), "LEFT"),
             ("VALIGN", (0, 0), (-1, -1), "TOP"),
             ("FONTSIZE", (0, 0), (-1, -1), 11),
@@ -265,19 +278,30 @@ def _create_info_table_style() -> TableStyle:
 
 
 def _create_header_table_style(header_color: colors.Color = None) -> TableStyle:
-    """Create a reusable table style for tables with headers."""
+    """Create a reusable table style for tables with headers.
+
+    ReportLab TableStyle coordinate system:
+        - Format: (COMMAND, (start_col, start_row), (end_col, end_row), value)
+        - (0, 0) to (-1, 0) = entire first row (header row)
+        - (1, 1) to (-1, -1) = all data cells (excludes header row and first column)
+        - See _create_info_table_style() for full coordinate system documentation
+    """
     if header_color is None:
         header_color = COLOR_BLUE
 
     return TableStyle(
         [
+            # Header row (row 0): colored background with white text
             ("BACKGROUND", (0, 0), (-1, 0), header_color),
             ("TEXTCOLOR", (0, 0), (-1, 0), COLOR_WHITE),
             ("FONTNAME", (0, 0), (-1, 0), "FiraCode"),
             ("FONTSIZE", (0, 0), (-1, 0), 10),
+            # Apply to entire table
             ("ALIGN", (0, 0), (-1, -1), "CENTER"),
             ("VALIGN", (0, 0), (-1, -1), "MIDDLE"),
+            # Data cells (excluding header): smaller font
             ("FONTSIZE", (1, 1), (-1, -1), 9),
+            # Apply to entire table
             ("GRID", (0, 0), (-1, -1), 1, COLOR_GRID_GRAY),
             ("LEFTPADDING", (0, 0), (-1, -1), PADDING_MEDIUM),
             ("RIGHTPADDING", (0, 0), (-1, -1), PADDING_MEDIUM),
@@ -288,18 +312,30 @@ def _create_header_table_style(header_color: colors.Color = None) -> TableStyle:
 
 
 def _create_findings_table_style() -> TableStyle:
-    """Create a reusable table style for findings tables."""
+    """Create a reusable table style for findings tables.
+
+    ReportLab TableStyle coordinate system:
+        - Format: (COMMAND, (start_col, start_row), (end_col, end_row), value)
+        - (0, 0) to (-1, 0) = entire first row (header row)
+        - (0, 0) to (0, 0) = only the top-left cell
+        - See _create_info_table_style() for full coordinate system documentation
+    """
     return TableStyle(
         [
+            # Header row (row 0): colored background with white text
             ("BACKGROUND", (0, 0), (-1, 0), COLOR_BLUE),
             ("TEXTCOLOR", (0, 0), (-1, 0), COLOR_WHITE),
             ("FONTNAME", (0, 0), (-1, 0), "FiraCode"),
+            # Only top-left cell centered (for index/number column)
             ("ALIGN", (0, 0), (0, 0), "CENTER"),
+            # Apply to entire table
             ("VALIGN", (0, 0), (-1, -1), "MIDDLE"),
             ("FONTSIZE", (0, 0), (-1, -1), 9),
             ("GRID", (0, 0), (-1, -1), 0.1, COLOR_BORDER_GRAY),
+            # Remove padding only from top-left cell
             ("LEFTPADDING", (0, 0), (0, 0), 0),
             ("RIGHTPADDING", (0, 0), (0, 0), 0),
+            # Apply to entire table
             ("TOPPADDING", (0, 0), (-1, -1), PADDING_SMALL),
             ("BOTTOMPADDING", (0, 0), (-1, -1), PADDING_SMALL),
         ]
@@ -1103,11 +1139,15 @@ def generate_threatscore_report(
         elements.append(Spacer(1, 0.5 * inch))
 
         # Add compliance information table
+        provider_alias = provider_obj.alias or "N/A"
         info_data = [
             ["Framework:", compliance_framework],
             ["ID:", compliance_id],
             ["Name:", Paragraph(compliance_name, normal_center)],
             ["Version:", compliance_version],
+            ["Provider:", provider_type.upper()],
+            ["Account ID:", provider_obj.uid],
+            ["Alias:", provider_alias],
             ["Scan ID:", scan_id],
             ["Description:", Paragraph(compliance_description, normal_center)],
         ]
@@ -2059,12 +2099,15 @@ def generate_ens_report(
         elements.append(Spacer(1, 0.5 * inch))
 
         # Add compliance information table
+        provider_alias = provider_obj.alias or "N/A"
         info_data = [
             ["Framework:", compliance_framework],
             ["ID:", compliance_id],
             ["Nombre:", Paragraph(compliance_name, normal_center)],
             ["Versión:", compliance_version],
             ["Proveedor:", provider_type.upper()],
+            ["Account ID:", provider_obj.uid],
+            ["Alias:", provider_alias],
             ["Scan ID:", scan_id],
             ["Descripción:", Paragraph(compliance_description, normal_center)],
         ]
@@ -2072,12 +2115,12 @@ def generate_ens_report(
         info_table.setStyle(
             TableStyle(
                 [
-                    ("BACKGROUND", (0, 0), (0, 6), colors.Color(0.2, 0.4, 0.6)),
-                    ("TEXTCOLOR", (0, 0), (0, 6), colors.white),
-                    ("FONTNAME", (0, 0), (0, 6), "FiraCode"),
-                    ("BACKGROUND", (1, 0), (1, 6), colors.Color(0.95, 0.97, 1.0)),
-                    ("TEXTCOLOR", (1, 0), (1, 6), colors.Color(0.2, 0.2, 0.2)),
-                    ("FONTNAME", (1, 0), (1, 6), "PlusJakartaSans"),
+                    ("BACKGROUND", (0, 0), (0, -1), colors.Color(0.2, 0.4, 0.6)),
+                    ("TEXTCOLOR", (0, 0), (0, -1), colors.white),
+                    ("FONTNAME", (0, 0), (0, -1), "FiraCode"),
+                    ("BACKGROUND", (1, 0), (1, -1), colors.Color(0.95, 0.97, 1.0)),
+                    ("TEXTCOLOR", (1, 0), (1, -1), colors.Color(0.2, 0.2, 0.2)),
+                    ("FONTNAME", (1, 0), (1, -1), "PlusJakartaSans"),
                     ("ALIGN", (0, 0), (-1, -1), "LEFT"),
                     ("VALIGN", (0, 0), (-1, -1), "TOP"),
                     ("FONTSIZE", (0, 0), (-1, -1), 11),
@@ -2238,12 +2281,20 @@ def generate_ens_report(
             [
                 "CUMPLE",
                 str(passed_requirements),
-                f"{(passed_requirements / total_requirements * 100):.1f}%",
+                (
+                    f"{(passed_requirements / total_requirements * 100):.1f}%"
+                    if total_requirements > 0
+                    else "0.0%"
+                ),
             ],
             [
                 "NO CUMPLE",
                 str(failed_requirements),
-                f"{(failed_requirements / total_requirements * 100):.1f}%",
+                (
+                    f"{(failed_requirements / total_requirements * 100):.1f}%"
+                    if total_requirements > 0
+                    else "0.0%"
+                ),
             ],
             ["TOTAL", str(total_requirements), "100%"],
         ]
@@ -2989,11 +3040,14 @@ def generate_nis2_report(
         elements.append(Spacer(1, 0.3 * inch))
 
         # Compliance metadata table
+        provider_alias = provider_obj.alias or "N/A"
         metadata_data = [
             ["Framework:", compliance_framework],
             ["Name:", Paragraph(compliance_name, normal_center)],
             ["Version:", compliance_version or "N/A"],
             ["Provider:", provider_type.upper()],
+            ["Account ID:", provider_obj.uid],
+            ["Alias:", provider_alias],
             ["Scan ID:", scan_id],
             ["Description:", Paragraph(compliance_description, normal_center)],
         ]
@@ -3477,6 +3531,7 @@ def generate_compliance_reports(
         "gcp",
         "m365",
         "kubernetes",
+        "alibabacloud",
     ]:
         logger.info(
             f"Provider {provider_id} ({provider_type}) is not supported for ThreatScore report"

api/src/backend/tasks/jobs/scan.py

@@ -8,11 +8,16 @@ from collections import defaultdict
 from datetime import datetime, timezone
 from typing import Any
 
+import sentry_sdk
 from celery.utils.log import get_task_logger
 from config.env import env
 from config.settings.celery import CELERY_DEADLOCK_ATTEMPTS
 from django.db import IntegrityError, OperationalError
-from django.db.models import Case, Count, IntegerField, Prefetch, Sum, When
+from django.db.models import Case, Count, IntegerField, Prefetch, Q, Sum, When
+from tasks.jobs.queries import (
+    COMPLIANCE_UPSERT_PROVIDER_SCORE_SQL,
+    COMPLIANCE_UPSERT_TENANT_SUMMARY_SQL,
+)
 from tasks.utils import CustomEncoder
 
 from api.compliance import PROWLER_COMPLIANCE_OVERVIEW_TEMPLATE
@@ -26,8 +31,10 @@ from api.db_utils import (
 )
 from api.exceptions import ProviderConnectionError
 from api.models import (
+    AttackSurfaceOverview,
     ComplianceOverviewSummary,
     ComplianceRequirementOverview,
+    DailySeveritySummary,
     Finding,
     MuteRule,
     Processor,
@@ -37,12 +44,14 @@ from api.models import (
     ResourceScanSummary,
     ResourceTag,
     Scan,
+    ScanCategorySummary,
     ScanSummary,
     StateChoices,
 )
 from api.models import StatusChoices as FindingStatus
 from api.utils import initialize_prowler_provider, return_prowler_provider
 from api.v1.serializers import ScanTaskSerializer
+from prowler.lib.check.models import CheckMetadata
 from prowler.lib.outputs.finding import Finding as ProwlerFinding
 from prowler.lib.scan.scan import Scan as ProwlerScan
 
@@ -74,6 +83,77 @@ FINDINGS_MICRO_BATCH_SIZE = env.int("DJANGO_FINDINGS_MICRO_BATCH_SIZE", default=
 # Controls how many rows each ORM bulk_create/bulk_update call sends to Postgres
 SCAN_DB_BATCH_SIZE = env.int("DJANGO_SCAN_DB_BATCH_SIZE", default=500)
 
+ATTACK_SURFACE_PROVIDER_COMPATIBILITY = {
+    "internet-exposed": None,  # Compatible with all providers
+    "secrets": None,  # Compatible with all providers
+    "privilege-escalation": ["aws", "kubernetes"],
+    "ec2-imdsv1": ["aws"],
+}
+
+_ATTACK_SURFACE_MAPPING_CACHE: dict[str, dict] = {}
+
+
+def aggregate_category_counts(
+    categories: list[str],
+    severity: str,
+    status: str,
+    delta: str | None,
+    muted: bool,
+    cache: dict[tuple[str, str], dict[str, int]],
+) -> None:
+    """
+    Increment category counters in-place for a finding.
+
+    Args:
+        categories: List of categories from finding metadata.
+        severity: Severity level (e.g., "high", "medium").
+        status: Finding status as string ("FAIL", "PASS").
+        delta: Delta value as string ("new", "changed") or None.
+        muted: Whether the finding is muted.
+        cache: Dict {(category, severity): {"total", "failed", "new_failed"}} to update.
+    """
+    is_failed = status == "FAIL" and not muted
+    is_new_failed = is_failed and delta == "new"
+
+    for cat in categories:
+        key = (cat, severity)
+        if key not in cache:
+            cache[key] = {"total": 0, "failed": 0, "new_failed": 0}
+        if not muted:
+            cache[key]["total"] += 1
+        if is_failed:
+            cache[key]["failed"] += 1
+        if is_new_failed:
+            cache[key]["new_failed"] += 1
+
+
+def _get_attack_surface_mapping_from_provider(provider_type: str) -> dict:
+    global _ATTACK_SURFACE_MAPPING_CACHE
+
+    if provider_type in _ATTACK_SURFACE_MAPPING_CACHE:
+        return _ATTACK_SURFACE_MAPPING_CACHE[provider_type]
+
+    attack_surface_check_mappings = {
+        "internet-exposed": None,
+        "secrets": None,
+        "privilege-escalation": {
+            "iam_policy_allows_privilege_escalation",
+            "iam_inline_policy_allows_privilege_escalation",
+        },
+        "ec2-imdsv1": {
+            "ec2_instance_imdsv2_enabled"
+        },  # AWS only - IMDSv1 enabled findings
+    }
+    for category_name, check_ids in attack_surface_check_mappings.items():
+        if check_ids is None:
+            sdk_check_ids = CheckMetadata.list(
+                provider=provider_type, category=category_name
+            )
+            attack_surface_check_mappings[category_name] = sdk_check_ids
+
+    _ATTACK_SURFACE_MAPPING_CACHE[provider_type] = attack_surface_check_mappings
+    return attack_surface_check_mappings
+
 
 def _create_finding_delta(
     last_status: FindingStatus | None | str, new_status: FindingStatus | None
@@ -330,7 +410,7 @@ def _create_compliance_summaries(
     if summary_objects:
         with rls_transaction(tenant_id):
             ComplianceOverviewSummary.objects.bulk_create(
-                summary_objects, batch_size=500
+                summary_objects, batch_size=500, ignore_conflicts=True
             )
 
 
@@ -357,6 +437,7 @@ def _process_finding_micro_batch(
     unique_resources: set,
     scan_resource_cache: set,
     mute_rules_cache: dict,
+    scan_categories_cache: dict[tuple[str, str], dict[str, int]],
 ) -> None:
     """
     Process a micro-batch of findings and persist them using bulk operations.
@@ -377,6 +458,7 @@ def _process_finding_micro_batch(
         unique_resources: Set tracking (uid, region) pairs seen in the scan.
         scan_resource_cache: Set of tuples used to create `ResourceScanSummary` rows.
         mute_rules_cache: Map of finding UID -> mute reason gathered before the scan.
+        scan_categories_cache: Dict tracking category counts {(category, severity): {"total", "failed", "new_failed"}}.
     """
     # Accumulate objects for bulk operations
     findings_to_create = []
@@ -532,11 +614,12 @@ def _process_finding_micro_batch(
             resource_failed_findings_cache[resource_uid] += 1
 
         # Create finding object (don't save yet)
+        check_metadata = finding.get_metadata()
         finding_instance = Finding(
             tenant_id=tenant_id,
             uid=finding_uid,
             delta=delta,
-            check_metadata=finding.get_metadata(),
+            check_metadata=check_metadata,
             status=status,
             status_extended=finding.status_extended,
             severity=finding.severity,
@@ -549,6 +632,7 @@ def _process_finding_micro_batch(
             muted_at=datetime.now(tz=timezone.utc) if is_muted else None,
             muted_reason=muted_reason,
             compliance=finding.compliance,
+            categories=check_metadata.get("categories", []) or [],
         )
         findings_to_create.append(finding_instance)
         resource_denormalized_data.append((finding_instance, resource_instance))
@@ -563,6 +647,16 @@ def _process_finding_micro_batch(
             )
         )
 
+        # Track categories with counts for ScanCategorySummary by (category, severity)
+        aggregate_category_counts(
+            categories=check_metadata.get("categories", []) or [],
+            severity=finding.severity.value,
+            status=status.value,
+            delta=delta.value if delta else None,
+            muted=is_muted,
+            cache=scan_categories_cache,
+        )
+
     # Bulk operations within single transaction
     with rls_transaction(tenant_id):
         # Bulk create findings
@@ -662,6 +756,7 @@ def perform_prowler_scan(
     exception = None
     unique_resources = set()
     scan_resource_cache: set[tuple[str, str, str, str]] = set()
+    scan_categories_cache: dict[tuple[str, str], dict[str, int]] = {}
     start_time = time.time()
     exc = None
 
@@ -751,6 +846,7 @@ def perform_prowler_scan(
                     unique_resources=unique_resources,
                     scan_resource_cache=scan_resource_cache,
                     mute_rules_cache=mute_rules_cache,
+                    scan_categories_cache=scan_categories_cache,
                 )
 
             # Update scan progress
@@ -810,13 +906,33 @@ def perform_prowler_scan(
                 resource_scan_summaries, batch_size=500, ignore_conflicts=True
             )
     except Exception as filter_exception:
-        import sentry_sdk
-
         sentry_sdk.capture_exception(filter_exception)
         logger.error(
             f"Error storing filter values for scan {scan_id}: {filter_exception}"
         )
 
+    try:
+        if scan_categories_cache:
+            category_summaries = [
+                ScanCategorySummary(
+                    tenant_id=tenant_id,
+                    scan_id=scan_id,
+                    category=category,
+                    severity=severity,
+                    total_findings=counts["total"],
+                    failed_findings=counts["failed"],
+                    new_failed_findings=counts["new_failed"],
+                )
+                for (category, severity), counts in scan_categories_cache.items()
+            ]
+            with rls_transaction(tenant_id):
+                ScanCategorySummary.objects.bulk_create(
+                    category_summaries, batch_size=500, ignore_conflicts=True
+                )
+    except Exception as cat_exception:
+        sentry_sdk.capture_exception(cat_exception)
+        logger.error(f"Error storing categories for scan {scan_id}: {cat_exception}")
+
     serializer = ScanTaskSerializer(instance=scan_instance)
     return serializer.data
 
@@ -979,11 +1095,14 @@ def _aggregate_findings_by_region(
     findings_count_by_compliance = {}
 
     with rls_transaction(tenant_id, using=READ_REPLICA_ALIAS):
-        # Fetch findings with resources in a single efficient query
-        # Use select_related for finding fields and prefetch_related for many-to-many resources
+        # Fetch only PASS/FAIL findings (optimized query reduces data transfer)
+        # Other statuses are not needed for check_status or ThreatScore calculation
         findings = (
             Finding.all_objects.filter(
-                tenant_id=tenant_id, scan_id=scan_id, muted=False
+                tenant_id=tenant_id,
+                scan_id=scan_id,
+                muted=False,
+                status__in=["PASS", "FAIL"],
             )
             .only("id", "check_id", "status", "compliance")
             .prefetch_related(
@@ -1001,6 +1120,8 @@ def _aggregate_findings_by_region(
         )
 
         for finding in findings:
+            status = finding.status
+
             for resource in finding.small_resources:
                 region = resource.region
 
@@ -1008,7 +1129,7 @@ def _aggregate_findings_by_region(
                 current_status = check_status_by_region.setdefault(region, {})
                 # Priority: FAIL > any other status
                 if current_status.get(finding.check_id) != "FAIL":
-                    current_status[finding.check_id] = finding.status
+                    current_status[finding.check_id] = status
 
                 # Aggregate ThreatScore compliance counts
                 if modeled_threatscore_compliance_id in (finding.compliance or {}):
@@ -1023,7 +1144,7 @@ def _aggregate_findings_by_region(
                             requirement_id, {"total": 0, "pass": 0}
                         )
                         requirement_stats["total"] += 1
-                        if finding.status == "PASS":
+                        if status == "PASS":
                             requirement_stats["pass"] += 1
 
     return check_status_by_region, findings_count_by_compliance
@@ -1191,3 +1312,321 @@ def create_compliance_requirements(tenant_id: str, scan_id: str):
     except Exception as e:
         logger.error(f"Error creating compliance requirements for scan {scan_id}: {e}")
         raise e
+
+
+def aggregate_attack_surface(tenant_id: str, scan_id: str):
+    """
+    Aggregate findings into attack surface overview records.
+
+    Creates one AttackSurfaceOverview record per attack surface type
+    for the given scan, based on check_id mappings.
+
+    Args:
+        tenant_id: Tenant that owns the scan.
+        scan_id: Scan UUID whose findings should be aggregated.
+    """
+    with rls_transaction(tenant_id, using=READ_REPLICA_ALIAS):
+        scan_instance = Scan.all_objects.select_related("provider").get(pk=scan_id)
+        provider_type = scan_instance.provider.provider
+
+    provider_attack_surface_mapping = _get_attack_surface_mapping_from_provider(
+        provider_type=provider_type
+    )
+
+    # Filter out attack surfaces that are not compatible or have no resolved check IDs
+    supported_mappings: dict[str, list[str]] = {}
+    for attack_surface_type, check_ids in provider_attack_surface_mapping.items():
+        compatible_providers = ATTACK_SURFACE_PROVIDER_COMPATIBILITY.get(
+            attack_surface_type
+        )
+        if (
+            compatible_providers is not None
+            and provider_type not in compatible_providers
+        ):
+            logger.info(
+                f"Skipping {attack_surface_type} - not supported for {provider_type}"
+            )
+            continue
+
+        if not check_ids:
+            logger.info(
+                f"Skipping {attack_surface_type} - no check IDs resolved for {provider_type}"
+            )
+            continue
+
+        supported_mappings[attack_surface_type] = list(check_ids)
+
+    if not supported_mappings:
+        logger.info(
+            f"No attack surface mappings available for scan {scan_id} and provider {provider_type}"
+        )
+        logger.info(f"No attack surface overview records created for scan {scan_id}")
+        return
+
+    # Map every check_id to its attack surface, so we can aggregate with a single query
+    check_id_to_surface: dict[str, str] = {}
+    for attack_surface_type, check_ids in supported_mappings.items():
+        for check_id in check_ids:
+            check_id_to_surface[check_id] = attack_surface_type
+
+    aggregated_counts = {
+        attack_surface_type: {"total": 0, "failed": 0, "muted": 0}
+        for attack_surface_type in supported_mappings.keys()
+    }
+
+    with rls_transaction(tenant_id, using=READ_REPLICA_ALIAS):
+        finding_stats = (
+            Finding.all_objects.filter(
+                tenant_id=tenant_id,
+                scan_id=scan_id,
+                check_id__in=list(check_id_to_surface.keys()),
+            )
+            .values("check_id")
+            .annotate(
+                total=Count("id"),
+                failed=Count("id", filter=Q(status="FAIL", muted=False)),
+                muted=Count("id", filter=Q(status="FAIL", muted=True)),
+            )
+        )
+
+        for stats in finding_stats:
+            attack_surface_type = check_id_to_surface.get(stats["check_id"])
+            if not attack_surface_type:
+                continue
+
+            aggregated_counts[attack_surface_type]["total"] += stats["total"] or 0
+            aggregated_counts[attack_surface_type]["failed"] += stats["failed"] or 0
+            aggregated_counts[attack_surface_type]["muted"] += stats["muted"] or 0
+
+    overview_objects = []
+    for attack_surface_type, counts in aggregated_counts.items():
+        total = counts["total"]
+        if not total:
+            continue
+
+        overview_objects.append(
+            AttackSurfaceOverview(
+                tenant_id=tenant_id,
+                scan_id=scan_id,
+                attack_surface_type=attack_surface_type,
+                total_findings=total,
+                failed_findings=counts["failed"],
+                muted_failed_findings=counts["muted"],
+            )
+        )
+
+    # Bulk create overview records
+    if overview_objects:
+        with rls_transaction(tenant_id):
+            AttackSurfaceOverview.objects.bulk_create(overview_objects, batch_size=500)
+            logger.info(
+                f"Created {len(overview_objects)} attack surface overview records for scan {scan_id}"
+            )
+    else:
+        logger.info(f"No attack surface overview records created for scan {scan_id}")
+
+
+def aggregate_daily_severity(tenant_id: str, scan_id: str):
+    """Aggregate scan severity counts into DailySeveritySummary (one record per provider/day)."""
+    with rls_transaction(tenant_id, using=READ_REPLICA_ALIAS):
+        scan = Scan.objects.filter(
+            tenant_id=tenant_id,
+            id=scan_id,
+            state=StateChoices.COMPLETED,
+        ).first()
+
+        if not scan:
+            logger.warning(f"Scan {scan_id} not found or not completed")
+            return {"status": "scan is not completed"}
+
+        provider_id = scan.provider_id
+        scan_date = scan.completed_at.date()
+
+        severity_totals = (
+            ScanSummary.objects.filter(
+                tenant_id=tenant_id,
+                scan_id=scan_id,
+            )
+            .values("severity")
+            .annotate(total_fail=Sum("fail"), total_muted=Sum("muted"))
+        )
+
+        severity_data = {
+            "critical": 0,
+            "high": 0,
+            "medium": 0,
+            "low": 0,
+            "informational": 0,
+            "muted": 0,
+        }
+
+        for row in severity_totals:
+            severity = row["severity"]
+            if severity in severity_data:
+                severity_data[severity] = row["total_fail"] or 0
+            severity_data["muted"] += row["total_muted"] or 0
+
+    with rls_transaction(tenant_id):
+        summary, created = DailySeveritySummary.objects.update_or_create(
+            tenant_id=tenant_id,
+            provider_id=provider_id,
+            date=scan_date,
+            defaults={
+                "scan_id": scan_id,
+                "critical": severity_data["critical"],
+                "high": severity_data["high"],
+                "medium": severity_data["medium"],
+                "low": severity_data["low"],
+                "informational": severity_data["informational"],
+                "muted": severity_data["muted"],
+            },
+        )
+
+    action = "created" if created else "updated"
+    logger.info(
+        f"Daily severity summary {action} for provider {provider_id} on {scan_date}"
+    )
+
+    return {
+        "status": action,
+        "provider_id": str(provider_id),
+        "date": str(scan_date),
+        "severity_data": severity_data,
+    }
+
+
+def update_provider_compliance_scores(tenant_id: str, scan_id: str):
+    """
+    Update ProviderComplianceScore with requirement statuses from a completed scan.
+
+    Uses atomic SQL upsert with ON CONFLICT for concurrency safety. Only updates
+    if the new scan is more recent than existing data. Also cleans up stale
+    requirements that no longer exist in the new scan.
+
+    Reads from primary DB (not replica) to avoid replication lag issues since
+    this runs immediately after create_compliance_requirements_task.
+
+    Args:
+        tenant_id: Tenant that owns the scan.
+        scan_id: Scan UUID whose compliance data should be materialized.
+
+    Returns:
+        dict: Statistics about the upsert operation.
+    """
+    with rls_transaction(tenant_id):
+        scan = (
+            Scan.all_objects.filter(
+                tenant_id=tenant_id,
+                id=scan_id,
+                state=StateChoices.COMPLETED,
+            )
+            .select_related("provider")
+            .first()
+        )
+
+        if not scan:
+            logger.warning(
+                f"Scan {scan_id} not found or not completed for compliance score update"
+            )
+            return {"status": "skipped", "reason": "scan not completed"}
+
+        if not scan.completed_at:
+            logger.warning(f"Scan {scan_id} has no completed_at timestamp")
+            return {"status": "skipped", "reason": "no completed_at"}
+
+        provider_id = str(scan.provider_id)
+        scan_completed_at = scan.completed_at
+
+    delete_stale_sql = """
+        DELETE FROM provider_compliance_scores pcs
+        WHERE pcs.tenant_id = %s
+          AND pcs.provider_id = %s
+          AND pcs.scan_completed_at < %s
+          AND NOT EXISTS (
+              SELECT 1 FROM compliance_requirements_overviews cro
+              WHERE cro.tenant_id = pcs.tenant_id
+                AND cro.scan_id = %s
+                AND cro.compliance_id = pcs.compliance_id
+                AND cro.requirement_id = pcs.requirement_id
+          )
+        RETURNING compliance_id
+    """
+
+    compliance_ids_sql = """
+        SELECT DISTINCT compliance_id
+        FROM compliance_requirements_overviews
+        WHERE tenant_id = %s AND scan_id = %s
+    """
+
+    try:
+        with psycopg_connection(MainRouter.default_db) as connection:
+            connection.autocommit = False
+            try:
+                with connection.cursor() as cursor:
+                    cursor.execute(SET_CONFIG_QUERY, [POSTGRES_TENANT_VAR, tenant_id])
+
+                    # Update requirement-level scores per provider
+                    cursor.execute(
+                        COMPLIANCE_UPSERT_PROVIDER_SCORE_SQL, [tenant_id, scan_id]
+                    )
+                    upserted_count = cursor.rowcount
+
+                    cursor.execute(compliance_ids_sql, [tenant_id, scan_id])
+                    scan_rows = cursor.fetchall()
+                    if not isinstance(scan_rows, (list, tuple)):
+                        scan_rows = []
+                    scan_compliance_ids = {row[0] for row in scan_rows}
+
+                    cursor.execute(
+                        delete_stale_sql,
+                        [tenant_id, provider_id, scan_completed_at, scan_id],
+                    )
+                    deleted_rows = cursor.fetchall()
+                    if not isinstance(deleted_rows, (list, tuple)):
+                        deleted_rows = []
+                    deleted_ids = {row[0] for row in deleted_rows}
+                    stale_deleted = len(deleted_ids)
+
+                    impacted_compliance_ids = sorted(scan_compliance_ids | deleted_ids)
+
+                    if impacted_compliance_ids:
+                        # Advisory lock on tenant to prevent race conditions when
+                        # multiple scans complete simultaneously for the same tenant
+                        cursor.execute(
+                            "SELECT pg_advisory_xact_lock(hashtext(%s))", [tenant_id]
+                        )
+
+                        # Recalculate tenant-level summary (FAIL-dominant across all providers)
+                        cursor.execute(
+                            COMPLIANCE_UPSERT_TENANT_SUMMARY_SQL,
+                            [tenant_id, tenant_id, impacted_compliance_ids],
+                        )
+                        tenant_summary_count = cursor.rowcount
+                    else:
+                        tenant_summary_count = 0
+
+                connection.commit()
+            except Exception:
+                connection.rollback()
+                raise
+
+        logger.info(
+            f"Provider compliance scores updated for scan {scan_id}: "
+            f"{upserted_count} upserted, {stale_deleted} stale deleted, "
+            f"{tenant_summary_count} tenant summaries upserted"
+        )
+
+        return {
+            "status": "completed",
+            "scan_id": str(scan_id),
+            "provider_id": provider_id,
+            "upserted": upserted_count,
+            "stale_deleted": stale_deleted,
+            "tenant_summary_count": tenant_summary_count,
+        }
+
+    except Exception as e:
+        logger.error(
+            f"Error updating provider compliance scores for scan {scan_id}: {e}"
+        )
+        raise

api/src/backend/tasks/tasks.py

@@ -11,7 +11,7 @@ from django_celery_beat.models import PeriodicTask
 from api.compliance import get_compliance_frameworks
 from api.db_router import READ_REPLICA_ALIAS
 from api.db_utils import rls_transaction
-from api.decorators import set_tenant
+from api.decorators import handle_provider_deletion, set_tenant
 from api.models import Finding, Integration, Provider, Scan, ScanSummary, StateChoices
 from api.utils import initialize_prowler_provider
 from api.v1.serializers import ScanTaskSerializer
@@ -23,7 +23,10 @@ from prowler.lib.outputs.finding import Finding as FindingOutput
 from tasks.jobs.attack_paths import attack_paths_scan
 from tasks.jobs.backfill import (
     backfill_compliance_summaries,
+    backfill_daily_severity_summaries,
+    backfill_provider_compliance_scores,
     backfill_resource_scan_summaries,
+    backfill_scan_category_summaries,
 )
 from tasks.jobs.connection import (
     check_integration_connection,
@@ -50,15 +53,70 @@ from tasks.jobs.lighthouse_providers import (
 from tasks.jobs.muting import mute_historical_findings
 from tasks.jobs.report import generate_compliance_reports_job
 from tasks.jobs.scan import (
+    aggregate_attack_surface,
+    aggregate_daily_severity,
     aggregate_findings,
     create_compliance_requirements,
     perform_prowler_scan,
+    update_provider_compliance_scores,
 )
 from tasks.utils import batched, get_next_execution_datetime
 
 logger = get_task_logger(__name__)
 
 
+def _cleanup_orphan_scheduled_scans(
+    tenant_id: str,
+    provider_id: str,
+    scheduler_task_id: int,
+) -> int:
+    """
+    TEMPORARY WORKAROUND: Clean up orphan AVAILABLE scans.
+
+    Detects and removes AVAILABLE scans that were never used due to an
+    issue during the first scheduled scan setup.
+
+    An AVAILABLE scan is considered orphan if there's also a SCHEDULED scan for
+    the same provider with the same scheduler_task_id. This situation indicates
+    that the first scan execution didn't find the AVAILABLE scan (because it
+    wasn't committed yet, probably) and created a new one, leaving the AVAILABLE orphaned.
+
+    Args:
+        tenant_id: The tenant ID.
+        provider_id: The provider ID.
+        scheduler_task_id: The PeriodicTask ID that triggers these scans.
+
+    Returns:
+        Number of orphan scans deleted (0 if none found).
+    """
+    orphan_available_scans = Scan.objects.filter(
+        tenant_id=tenant_id,
+        provider_id=provider_id,
+        trigger=Scan.TriggerChoices.SCHEDULED,
+        state=StateChoices.AVAILABLE,
+        scheduler_task_id=scheduler_task_id,
+    )
+
+    scheduled_scan_exists = Scan.objects.filter(
+        tenant_id=tenant_id,
+        provider_id=provider_id,
+        trigger=Scan.TriggerChoices.SCHEDULED,
+        state=StateChoices.SCHEDULED,
+        scheduler_task_id=scheduler_task_id,
+    ).exists()
+
+    if scheduled_scan_exists and orphan_available_scans.exists():
+        orphan_count = orphan_available_scans.count()
+        logger.warning(
+            f"[WORKAROUND] Found {orphan_count} orphan AVAILABLE scan(s) for "
+            f"provider {provider_id} alongside a SCHEDULED scan. Cleaning up orphans..."
+        )
+        orphan_available_scans.delete()
+        return orphan_count
+
+    return 0
+
+
 def _perform_scan_complete_tasks(tenant_id: str, scan_id: str, provider_id: str):
     """
     Helper function to perform tasks after a scan is completed.
@@ -68,13 +126,20 @@ def _perform_scan_complete_tasks(tenant_id: str, scan_id: str, provider_id: str)
         scan_id (str): The ID of the scan that was performed.
         provider_id (str): The primary key of the Provider instance that was scanned.
     """
-    create_compliance_requirements_task.apply_async(
+    chain(
+        create_compliance_requirements_task.si(tenant_id=tenant_id, scan_id=scan_id),
+        update_provider_compliance_scores_task.si(tenant_id=tenant_id, scan_id=scan_id),
+    ).apply_async()
+    aggregate_attack_surface_task.apply_async(
         kwargs={"tenant_id": tenant_id, "scan_id": scan_id}
     )
     chain(
         perform_scan_summary_task.si(tenant_id=tenant_id, scan_id=scan_id),
-        generate_outputs_task.si(
-            scan_id=scan_id, provider_id=provider_id, tenant_id=tenant_id
+        group(
+            aggregate_daily_severity_task.si(tenant_id=tenant_id, scan_id=scan_id),
+            generate_outputs_task.si(
+                scan_id=scan_id, provider_id=provider_id, tenant_id=tenant_id
+            ),
         ),
         group(
             # Use optimized task that generates both reports with shared queries
@@ -145,6 +210,7 @@ def delete_provider_task(provider_id: str, tenant_id: str):
 
 
 @shared_task(base=RLSTask, name="scan-perform", queue="scans")
+@handle_provider_deletion
 def perform_scan_task(
     tenant_id: str, scan_id: str, provider_id: str, checks_to_execute: list[str] = None
 ):
@@ -177,6 +243,7 @@ def perform_scan_task(
 
 
 @shared_task(base=RLSTask, bind=True, name="scan-perform-scheduled", queue="scans")
+@handle_provider_deletion
 def perform_scheduled_scan_task(self, tenant_id: str, provider_id: str):
     """
     Task to perform a scheduled Prowler scan on a given provider.
@@ -240,6 +307,14 @@ def perform_scheduled_scan_task(self, tenant_id: str, provider_id: str):
             return serializer.data
 
         next_scan_datetime = get_next_execution_datetime(task_id, provider_id)
+
+        # TEMPORARY WORKAROUND: Clean up orphan scans from transaction isolation issue
+        _cleanup_orphan_scheduled_scans(
+            tenant_id=tenant_id,
+            provider_id=provider_id,
+            scheduler_task_id=periodic_task_instance.id,
+        )
+
         scan_instance, _ = Scan.objects.get_or_create(
             tenant_id=tenant_id,
             provider_id=provider_id,
@@ -282,6 +357,7 @@ def perform_scheduled_scan_task(self, tenant_id: str, provider_id: str):
 
 
 @shared_task(name="scan-summary", queue="overview")
+@handle_provider_deletion
 def perform_scan_summary_task(tenant_id: str, scan_id: str):
     return aggregate_findings(tenant_id=tenant_id, scan_id=scan_id)
 
@@ -316,6 +392,7 @@ def delete_tenant_task(tenant_id: str):
     queue="scan-reports",
 )
 @set_tenant(keep_tenant=True)
+@handle_provider_deletion
 def generate_outputs_task(scan_id: str, provider_id: str, tenant_id: str):
     """
     Process findings in batches and generate output files in multiple formats.
@@ -511,6 +588,7 @@ def generate_outputs_task(scan_id: str, provider_id: str, tenant_id: str):
 
 
 @shared_task(name="backfill-scan-resource-summaries", queue="backfill")
+@handle_provider_deletion
 def backfill_scan_resource_summaries_task(tenant_id: str, scan_id: str):
     """
     Tries to backfill the resource scan summaries table for a given scan.
@@ -523,6 +601,7 @@ def backfill_scan_resource_summaries_task(tenant_id: str, scan_id: str):
 
 
 @shared_task(name="backfill-compliance-summaries", queue="backfill")
+@handle_provider_deletion
 def backfill_compliance_summaries_task(tenant_id: str, scan_id: str):
     """
     Tries to backfill compliance overview summaries for a completed scan.
@@ -537,7 +616,43 @@ def backfill_compliance_summaries_task(tenant_id: str, scan_id: str):
     return backfill_compliance_summaries(tenant_id=tenant_id, scan_id=scan_id)
 
 
+@shared_task(name="backfill-daily-severity-summaries", queue="backfill")
+def backfill_daily_severity_summaries_task(tenant_id: str, days: int = None):
+    """Backfill DailySeveritySummary from historical scans. Use days param to limit scope."""
+    return backfill_daily_severity_summaries(tenant_id=tenant_id, days=days)
+
+
+@shared_task(name="backfill-scan-category-summaries", queue="backfill")
+@handle_provider_deletion
+def backfill_scan_category_summaries_task(tenant_id: str, scan_id: str):
+    """
+    Backfill ScanCategorySummary for a completed scan.
+
+    Aggregates unique categories from findings and creates a summary row.
+
+    Args:
+        tenant_id (str): The tenant identifier.
+        scan_id (str): The scan identifier.
+    """
+    return backfill_scan_category_summaries(tenant_id=tenant_id, scan_id=scan_id)
+
+
+@shared_task(name="backfill-provider-compliance-scores", queue="backfill")
+def backfill_provider_compliance_scores_task(tenant_id: str):
+    """
+    Backfill ProviderComplianceScore from latest completed scan per provider.
+
+    Used to populate the compliance watchlist materialized table for tenants
+    that had scans before the feature was deployed.
+
+    Args:
+        tenant_id: Target tenant UUID.
+    """
+    return backfill_provider_compliance_scores(tenant_id=tenant_id)
+
+
 @shared_task(base=RLSTask, name="scan-compliance-overviews", queue="compliance")
+@handle_provider_deletion
 def create_compliance_requirements_task(tenant_id: str, scan_id: str):
     """
     Creates detailed compliance requirement records for a scan.
@@ -553,6 +668,44 @@ def create_compliance_requirements_task(tenant_id: str, scan_id: str):
     return create_compliance_requirements(tenant_id=tenant_id, scan_id=scan_id)
 
 
+@shared_task(name="scan-attack-surface-overviews", queue="overview")
+@handle_provider_deletion
+def aggregate_attack_surface_task(tenant_id: str, scan_id: str):
+    """
+    Creates attack surface overview records for a scan.
+
+    This task processes findings and aggregates them into attack surface categories
+    (internet-exposed, secrets, privilege-escalation, ec2-imdsv1) for quick overview queries.
+
+    Args:
+        tenant_id (str): The tenant ID for which to create records.
+        scan_id (str): The ID of the scan for which to create records.
+    """
+    return aggregate_attack_surface(tenant_id=tenant_id, scan_id=scan_id)
+
+
+@shared_task(name="scan-provider-compliance-scores", queue="compliance")
+def update_provider_compliance_scores_task(tenant_id: str, scan_id: str):
+    """
+    Update provider compliance scores from a completed scan.
+
+    This task materializes compliance requirement statuses into ProviderComplianceScore
+    for efficient watchlist queries. Uses atomic upsert with concurrency protection.
+
+    Args:
+        tenant_id (str): The tenant ID for which to update scores.
+        scan_id (str): The ID of the scan whose data should be materialized.
+    """
+    return update_provider_compliance_scores(tenant_id=tenant_id, scan_id=scan_id)
+
+
+@shared_task(name="scan-daily-severity", queue="overview")
+@handle_provider_deletion
+def aggregate_daily_severity_task(tenant_id: str, scan_id: str):
+    """Aggregate scan severity into DailySeveritySummary for findings_severity/timeseries endpoint."""
+    return aggregate_daily_severity(tenant_id=tenant_id, scan_id=scan_id)
+
+
 @shared_task(base=RLSTask, name="lighthouse-connection-check")
 @set_tenant
 def check_lighthouse_connection_task(lighthouse_config_id: str, tenant_id: str = None):
@@ -591,6 +744,7 @@ def refresh_lighthouse_provider_models_task(
 
 
 @shared_task(name="integration-check")
+@handle_provider_deletion
 def check_integrations_task(tenant_id: str, provider_id: str, scan_id: str = None):
     """
     Check and execute all configured integrations for a provider.
@@ -655,6 +809,7 @@ def check_integrations_task(tenant_id: str, provider_id: str, scan_id: str = Non
     name="integration-s3",
     queue="integrations",
 )
+@handle_provider_deletion
 def s3_integration_task(
     tenant_id: str,
     provider_id: str,
@@ -714,6 +869,7 @@ def jira_integration_task(
     name="scan-compliance-reports",
     queue="scan-reports",
 )
+@handle_provider_deletion
 def generate_compliance_reports_task(tenant_id: str, scan_id: str, provider_id: str):
     """
     Optimized task to generate ThreatScore, ENS, and NIS2 reports with shared queries.

api/src/backend/tasks/tests/test_backfill.py

@@ -1,17 +1,25 @@
+from datetime import datetime, timezone
+from unittest.mock import MagicMock, patch
 from uuid import uuid4
 
 import pytest
 from tasks.jobs.backfill import (
     backfill_compliance_summaries,
+    backfill_provider_compliance_scores,
     backfill_resource_scan_summaries,
+    backfill_scan_category_summaries,
 )
 
 from api.models import (
     ComplianceOverviewSummary,
+    Finding,
     ResourceScanSummary,
     Scan,
+    ScanCategorySummary,
     StateChoices,
 )
+from prowler.lib.check.models import Severity
+from prowler.lib.outputs.finding import Status
 
 
 @pytest.fixture(scope="function")
@@ -46,6 +54,45 @@ def get_not_completed_scans(providers_fixture):
     return scan_1, scan_2
 
 
+@pytest.fixture(scope="function")
+def findings_with_categories_fixture(scans_fixture, resources_fixture):
+    scan = scans_fixture[0]
+    resource = resources_fixture[0]
+
+    finding = Finding.objects.create(
+        tenant_id=scan.tenant_id,
+        uid="finding_with_categories",
+        scan=scan,
+        delta="new",
+        status=Status.FAIL,
+        status_extended="test status",
+        impact=Severity.critical,
+        impact_extended="test impact",
+        severity=Severity.critical,
+        raw_result={"status": Status.FAIL},
+        check_id="test_check",
+        check_metadata={"CheckId": "test_check"},
+        categories=["gen-ai", "security"],
+        first_seen_at="2024-01-02T00:00:00Z",
+    )
+    finding.add_resources([resource])
+    return finding
+
+
+@pytest.fixture(scope="function")
+def scan_category_summary_fixture(scans_fixture):
+    scan = scans_fixture[0]
+    return ScanCategorySummary.objects.create(
+        tenant_id=scan.tenant_id,
+        scan=scan,
+        category="existing-category",
+        severity=Severity.critical,
+        total_findings=1,
+        failed_findings=0,
+        new_failed_findings=0,
+    )
+
+
 @pytest.mark.django_db
 class TestBackfillResourceScanSummaries:
     def test_already_backfilled(self, resource_scan_summary_data):
@@ -172,3 +219,106 @@ class TestBackfillComplianceSummaries:
             assert summary.requirements_failed == expected_counts["requirements_failed"]
             assert summary.requirements_manual == expected_counts["requirements_manual"]
             assert summary.total_requirements == expected_counts["total_requirements"]
+
+
+@pytest.mark.django_db
+class TestBackfillScanCategorySummaries:
+    def test_already_backfilled(self, scan_category_summary_fixture):
+        tenant_id = scan_category_summary_fixture.tenant_id
+        scan_id = scan_category_summary_fixture.scan_id
+
+        result = backfill_scan_category_summaries(str(tenant_id), str(scan_id))
+
+        assert result == {"status": "already backfilled"}
+
+    def test_not_completed_scan(self, get_not_completed_scans):
+        for scan in get_not_completed_scans:
+            result = backfill_scan_category_summaries(str(scan.tenant_id), str(scan.id))
+            assert result == {"status": "scan is not completed"}
+
+    def test_no_categories_to_backfill(self, scans_fixture):
+        scan = scans_fixture[1]  # Failed scan with no findings
+        result = backfill_scan_category_summaries(str(scan.tenant_id), str(scan.id))
+        assert result == {"status": "no categories to backfill"}
+
+    def test_successful_backfill(self, findings_with_categories_fixture):
+        finding = findings_with_categories_fixture
+        tenant_id = str(finding.tenant_id)
+        scan_id = str(finding.scan_id)
+
+        result = backfill_scan_category_summaries(tenant_id, scan_id)
+
+        # 2 categories × 1 severity = 2 rows
+        assert result == {"status": "backfilled", "categories_count": 2}
+
+        summaries = ScanCategorySummary.objects.filter(
+            tenant_id=tenant_id, scan_id=scan_id
+        )
+        assert summaries.count() == 2
+        categories = set(summaries.values_list("category", flat=True))
+        assert categories == {"gen-ai", "security"}
+
+        for summary in summaries:
+            assert summary.severity == Severity.critical
+            assert summary.total_findings == 1
+            assert summary.failed_findings == 1
+            assert summary.new_failed_findings == 1
+
+
+@pytest.mark.django_db
+class TestBackfillProviderComplianceScores:
+    def test_no_completed_scans(self, tenants_fixture):
+        tenant = tenants_fixture[2]
+        result = backfill_provider_compliance_scores(str(tenant.id))
+        assert result == {"status": "no completed scans"}
+
+    def test_no_scans_to_process(self, tenants_fixture, scans_fixture):
+        tenant = tenants_fixture[0]
+        scan = scans_fixture[0]
+        scan.completed_at = None
+        scan.save()
+
+        result = backfill_provider_compliance_scores(str(tenant.id))
+        assert result == {"status": "no completed scans"}
+
+    @patch("tasks.jobs.backfill.psycopg_connection")
+    def test_successful_backfill_executes_sql_queries(
+        self,
+        mock_psycopg_connection,
+        tenants_fixture,
+        scans_fixture,
+        settings,
+    ):
+        """Test successful backfill executes SQL queries and returns correct stats."""
+        settings.DATABASES.setdefault("admin", settings.DATABASES["default"])
+        tenant = tenants_fixture[0]
+        scan = scans_fixture[0]
+
+        # Set completed_at to make the scan eligible for backfill
+        scan.completed_at = datetime.now(timezone.utc)
+        scan.save()
+
+        connection = MagicMock()
+        cursor = MagicMock()
+        cursor_context = MagicMock()
+        cursor_context.__enter__.return_value = cursor
+        cursor_context.__exit__.return_value = False
+        connection.cursor.return_value = cursor_context
+        connection.__enter__.return_value = connection
+        connection.__exit__.return_value = False
+        connection.autocommit = True
+
+        context_manager = MagicMock()
+        context_manager.__enter__.return_value = connection
+        context_manager.__exit__.return_value = False
+        mock_psycopg_connection.return_value = context_manager
+
+        cursor.rowcount = 5
+
+        result = backfill_provider_compliance_scores(str(tenant.id))
+
+        assert result["status"] == "backfilled"
+        assert result["providers_processed"] == 1
+        assert result["providers_skipped"] == 0
+        assert result["total_upserted"] == 5
+        assert result["tenant_summary_count"] == 5

api/src/backend/tasks/tests/test_beat.py

@@ -28,6 +28,7 @@ class TestScheduleProviderScan:
                     "tenant_id": str(provider_instance.tenant_id),
                     "provider_id": str(provider_instance.id),
                 },
+                countdown=5,
             )
 
             task_name = f"scan-perform-scheduled-{provider_instance.id}"

api/src/backend/tasks/tests/test_integrations.py

@@ -1199,9 +1199,6 @@ class TestSecurityHubIntegrationUploads:
                     )
 
         assert result is False
-        # Integration should be marked as disconnected
-        integration.save.assert_called_once()
-        assert integration.connected is False
 
     @patch("tasks.jobs.integrations.ASFF")
     @patch("tasks.jobs.integrations.FindingOutput")

api/src/backend/tasks/tests/test_scan.py

@@ -9,17 +9,22 @@ from unittest.mock import MagicMock, patch
 
 import pytest
 from tasks.jobs.scan import (
+    _ATTACK_SURFACE_MAPPING_CACHE,
     _aggregate_findings_by_region,
     _copy_compliance_requirement_rows,
     _create_compliance_summaries,
     _create_finding_delta,
+    _get_attack_surface_mapping_from_provider,
     _normalized_compliance_key,
     _persist_compliance_requirement_rows,
     _process_finding_micro_batch,
     _store_resources,
+    aggregate_attack_surface,
+    aggregate_category_counts,
     aggregate_findings,
     create_compliance_requirements,
     perform_prowler_scan,
+    update_provider_compliance_scores,
 )
 from tasks.utils import CustomEncoder
 
@@ -1374,6 +1379,7 @@ class TestProcessFindingMicroBatch:
         unique_resources: set[tuple[str, str]] = set()
         scan_resource_cache: set[tuple[str, str, str, str]] = set()
         mute_rules_cache = {}
+        scan_categories_cache: dict[tuple[str, str], dict[str, int]] = {}
 
         with (
             patch("tasks.jobs.scan.rls_transaction", new=noop_rls_transaction),
@@ -1391,6 +1397,7 @@ class TestProcessFindingMicroBatch:
                 unique_resources,
                 scan_resource_cache,
                 mute_rules_cache,
+                scan_categories_cache,
             )
 
         created_finding = Finding.objects.get(uid=finding.uid)
@@ -1483,6 +1490,7 @@ class TestProcessFindingMicroBatch:
         unique_resources: set[tuple[str, str]] = set()
         scan_resource_cache: set[tuple[str, str, str, str]] = set()
         mute_rules_cache = {finding.uid: "Muted via rule"}
+        scan_categories_cache: dict[tuple[str, str], dict[str, int]] = {}
 
         with (
             patch("tasks.jobs.scan.rls_transaction", new=noop_rls_transaction),
@@ -1500,6 +1508,7 @@ class TestProcessFindingMicroBatch:
                 unique_resources,
                 scan_resource_cache,
                 mute_rules_cache,
+                scan_categories_cache,
             )
 
         existing_resource.refresh_from_db()
@@ -1607,6 +1616,7 @@ class TestProcessFindingMicroBatch:
         unique_resources: set[tuple[str, str]] = set()
         scan_resource_cache: set[tuple[str, str, str, str]] = set()
         mute_rules_cache = {}
+        scan_categories_cache: dict[tuple[str, str], dict[str, int]] = {}
 
         with (
             patch("tasks.jobs.scan.rls_transaction", new=noop_rls_transaction),
@@ -1625,6 +1635,7 @@ class TestProcessFindingMicroBatch:
                 unique_resources,
                 scan_resource_cache,
                 mute_rules_cache,
+                scan_categories_cache,
             )
 
         # Verify the long UID finding was NOT created
@@ -1645,6 +1656,118 @@ class TestProcessFindingMicroBatch:
             for call in warning_calls
         )
 
+    def test_process_finding_micro_batch_tracks_categories(
+        self, tenants_fixture, scans_fixture
+    ):
+        tenant = tenants_fixture[0]
+        scan = scans_fixture[0]
+        provider = scan.provider
+
+        finding1 = FakeFinding(
+            uid="finding-cat-1",
+            status=StatusChoices.PASS,
+            status_extended="all good",
+            severity=Severity.low,
+            check_id="genai_check",
+            resource_uid="arn:aws:bedrock:::model/test",
+            resource_name="test-model",
+            region="us-east-1",
+            service_name="bedrock",
+            resource_type="model",
+            resource_tags={},
+            resource_metadata={},
+            resource_details={},
+            partition="aws",
+            raw={},
+            compliance={},
+            metadata={"categories": ["gen-ai", "security"]},
+            muted=False,
+        )
+
+        finding2 = FakeFinding(
+            uid="finding-cat-2",
+            status=StatusChoices.FAIL,
+            status_extended="bad",
+            severity=Severity.high,
+            check_id="iam_check",
+            resource_uid="arn:aws:iam:::user/test",
+            resource_name="test-user",
+            region="us-east-1",
+            service_name="iam",
+            resource_type="user",
+            resource_tags={},
+            resource_metadata={},
+            resource_details={},
+            partition="aws",
+            raw={},
+            compliance={},
+            metadata={"categories": ["security", "iam"]},
+            muted=False,
+        )
+
+        resource_cache = {}
+        tag_cache = {}
+        last_status_cache = {}
+        resource_failed_findings_cache = {}
+        unique_resources: set[tuple[str, str]] = set()
+        scan_resource_cache: set[tuple[str, str, str, str]] = set()
+        mute_rules_cache = {}
+        scan_categories_cache: dict[tuple[str, str], dict[str, int]] = {}
+
+        with (
+            patch("tasks.jobs.scan.rls_transaction", new=noop_rls_transaction),
+            patch("api.db_utils.rls_transaction", new=noop_rls_transaction),
+        ):
+            _process_finding_micro_batch(
+                str(tenant.id),
+                [finding1, finding2],
+                scan,
+                provider,
+                resource_cache,
+                tag_cache,
+                last_status_cache,
+                resource_failed_findings_cache,
+                unique_resources,
+                scan_resource_cache,
+                mute_rules_cache,
+                scan_categories_cache,
+            )
+
+        # finding1: PASS, severity=low, categories=["gen-ai", "security"]
+        # finding2: FAIL, severity=high, categories=["security", "iam"]
+        # Keys are (category, severity) tuples
+        assert set(scan_categories_cache.keys()) == {
+            ("gen-ai", "low"),
+            ("security", "low"),
+            ("security", "high"),
+            ("iam", "high"),
+        }
+        assert scan_categories_cache[("gen-ai", "low")] == {
+            "total": 1,
+            "failed": 0,
+            "new_failed": 0,
+        }
+        assert scan_categories_cache[("security", "low")] == {
+            "total": 1,
+            "failed": 0,
+            "new_failed": 0,
+        }
+        assert scan_categories_cache[("security", "high")] == {
+            "total": 1,
+            "failed": 1,
+            "new_failed": 1,
+        }
+        assert scan_categories_cache[("iam", "high")] == {
+            "total": 1,
+            "failed": 1,
+            "new_failed": 1,
+        }
+
+        created_finding1 = Finding.objects.get(uid="finding-cat-1")
+        created_finding2 = Finding.objects.get(uid="finding-cat-2")
+        assert set(created_finding1.categories) == {"gen-ai", "security"}
+        assert set(created_finding2.categories) == {"security", "iam"}
+
 
 @pytest.mark.django_db
 class TestCreateComplianceRequirements:
@@ -3338,7 +3461,10 @@ class TestAggregateFindingsByRegion:
 
         # Verify filter was called with muted=False
         mock_findings_filter.assert_called_once_with(
-            tenant_id=tenant_id, scan_id=scan_id, muted=False
+            tenant_id=tenant_id,
+            scan_id=scan_id,
+            muted=False,
+            status__in=["PASS", "FAIL"],
         )
 
     @patch("tasks.jobs.scan.Finding.all_objects.filter")
@@ -3471,3 +3597,549 @@ class TestAggregateFindingsByRegion:
 
         assert check_status_by_region == {}
         assert findings_count_by_compliance == {}
+
+
+@pytest.mark.django_db
+class TestAggregateAttackSurface:
+    """Test aggregate_attack_surface function and related caching."""
+
+    def setup_method(self):
+        """Clear cache before each test."""
+        _ATTACK_SURFACE_MAPPING_CACHE.clear()
+
+    def teardown_method(self):
+        """Clear cache after each test."""
+        _ATTACK_SURFACE_MAPPING_CACHE.clear()
+
+    @patch("tasks.jobs.scan.CheckMetadata.list")
+    def test_get_attack_surface_mapping_caches_result(self, mock_check_metadata_list):
+        """Test that _get_attack_surface_mapping_from_provider caches results."""
+        mock_check_metadata_list.return_value = {"check_internet_exposed_1"}
+
+        # First call should hit CheckMetadata.list
+        result1 = _get_attack_surface_mapping_from_provider("aws")
+        assert mock_check_metadata_list.call_count == 2  # internet-exposed, secrets
+
+        # Second call should use cache
+        result2 = _get_attack_surface_mapping_from_provider("aws")
+        assert mock_check_metadata_list.call_count == 2  # No additional calls
+
+        assert result1 is result2
+        assert "aws" in _ATTACK_SURFACE_MAPPING_CACHE
+
+    @patch("tasks.jobs.scan.CheckMetadata.list")
+    def test_get_attack_surface_mapping_different_providers(
+        self, mock_check_metadata_list
+    ):
+        """Test caching works independently for different providers."""
+        mock_check_metadata_list.return_value = {"check_1"}
+
+        _get_attack_surface_mapping_from_provider("aws")
+        aws_call_count = mock_check_metadata_list.call_count
+
+        _get_attack_surface_mapping_from_provider("gcp")
+        gcp_call_count = mock_check_metadata_list.call_count
+
+        # Both providers should have made calls
+        assert gcp_call_count > aws_call_count
+        assert "aws" in _ATTACK_SURFACE_MAPPING_CACHE
+        assert "gcp" in _ATTACK_SURFACE_MAPPING_CACHE
+
+    @patch("tasks.jobs.scan.CheckMetadata.list")
+    def test_get_attack_surface_mapping_returns_hardcoded_checks(
+        self, mock_check_metadata_list
+    ):
+        """Test that hardcoded check IDs are returned for privilege-escalation and ec2-imdsv1."""
+        mock_check_metadata_list.return_value = set()
+
+        result = _get_attack_surface_mapping_from_provider("aws")
+
+        # Hardcoded checks should be present
+        assert (
+            "iam_policy_allows_privilege_escalation" in result["privilege-escalation"]
+        )
+        assert (
+            "iam_inline_policy_allows_privilege_escalation"
+            in result["privilege-escalation"]
+        )
+        assert "ec2_instance_imdsv2_enabled" in result["ec2-imdsv1"]
+
+    @patch("tasks.jobs.scan.AttackSurfaceOverview.objects.bulk_create")
+    @patch("tasks.jobs.scan.Finding.all_objects.filter")
+    @patch("tasks.jobs.scan._get_attack_surface_mapping_from_provider")
+    @patch("tasks.jobs.scan.rls_transaction")
+    def test_aggregate_attack_surface_creates_overview_records(
+        self,
+        mock_rls_transaction,
+        mock_get_mapping,
+        mock_findings_filter,
+        mock_bulk_create,
+        tenants_fixture,
+        scans_fixture,
+    ):
+        """Test that aggregate_attack_surface creates AttackSurfaceOverview records."""
+        tenant = tenants_fixture[0]
+        scan = scans_fixture[0]
+        scan.provider.provider = "aws"
+        scan.provider.save()
+
+        mock_get_mapping.return_value = {
+            "internet-exposed": {"check_internet_1", "check_internet_2"},
+            "secrets": {"check_secrets_1"},
+            "privilege-escalation": {"check_privesc_1"},
+            "ec2-imdsv1": {"check_imdsv1_1"},
+        }
+
+        # Mock findings aggregation
+        mock_queryset = MagicMock()
+        mock_queryset.values.return_value = mock_queryset
+        mock_queryset.annotate.return_value = [
+            {"check_id": "check_internet_1", "total": 10, "failed": 3, "muted": 1},
+            {"check_id": "check_secrets_1", "total": 5, "failed": 2, "muted": 0},
+        ]
+
+        ctx = MagicMock()
+        ctx.__enter__.return_value = None
+        ctx.__exit__.return_value = False
+        mock_rls_transaction.return_value = ctx
+        mock_findings_filter.return_value = mock_queryset
+
+        aggregate_attack_surface(str(tenant.id), str(scan.id))
+
+        mock_bulk_create.assert_called_once()
+        args, kwargs = mock_bulk_create.call_args
+        objects = args[0]
+
+        # Should create records for internet-exposed and secrets (the ones with findings)
+        assert len(objects) == 2
+        assert kwargs["batch_size"] == 500
+
+    @patch("tasks.jobs.scan.AttackSurfaceOverview.objects.bulk_create")
+    @patch("tasks.jobs.scan.Finding.all_objects.filter")
+    @patch("tasks.jobs.scan._get_attack_surface_mapping_from_provider")
+    @patch("tasks.jobs.scan.rls_transaction")
+    def test_aggregate_attack_surface_skips_unsupported_provider(
+        self,
+        mock_rls_transaction,
+        mock_get_mapping,
+        mock_findings_filter,
+        mock_bulk_create,
+        tenants_fixture,
+        scans_fixture,
+    ):
+        """Test that ec2-imdsv1 is skipped for non-AWS providers."""
+        tenant = tenants_fixture[0]
+        scan = scans_fixture[0]
+        scan.provider.provider = "gcp"
+        scan.provider.uid = "gcp-test-project-id"
+        scan.provider.save()
+
+        mock_get_mapping.return_value = {
+            "internet-exposed": {"check_internet_1"},
+            "secrets": {"check_secrets_1"},
+            "privilege-escalation": set(),  # Not supported for GCP
+            "ec2-imdsv1": {"check_imdsv1_1"},  # Should be skipped for GCP
+        }
+
+        mock_queryset = MagicMock()
+        mock_queryset.values.return_value = mock_queryset
+        mock_queryset.annotate.return_value = [
+            {"check_id": "check_internet_1", "total": 5, "failed": 1, "muted": 0},
+        ]
+
+        ctx = MagicMock()
+        ctx.__enter__.return_value = None
+        ctx.__exit__.return_value = False
+        mock_rls_transaction.return_value = ctx
+        mock_findings_filter.return_value = mock_queryset
+
+        aggregate_attack_surface(str(tenant.id), str(scan.id))
+
+        # ec2-imdsv1 check_ids should not be in the filter
+        filter_call = mock_findings_filter.call_args
+        check_ids_in_filter = filter_call[1]["check_id__in"]
+        assert "check_imdsv1_1" not in check_ids_in_filter
+
+    @patch("tasks.jobs.scan.AttackSurfaceOverview.objects.bulk_create")
+    @patch("tasks.jobs.scan.Finding.all_objects.filter")
+    @patch("tasks.jobs.scan._get_attack_surface_mapping_from_provider")
+    @patch("tasks.jobs.scan.rls_transaction")
+    def test_aggregate_attack_surface_no_findings(
+        self,
+        mock_rls_transaction,
+        mock_get_mapping,
+        mock_findings_filter,
+        mock_bulk_create,
+        tenants_fixture,
+        scans_fixture,
+    ):
+        """Test that no records are created when there are no findings."""
+        tenant = tenants_fixture[0]
+        scan = scans_fixture[0]
+
+        mock_get_mapping.return_value = {
+            "internet-exposed": {"check_1"},
+            "secrets": {"check_2"},
+            "privilege-escalation": set(),
+            "ec2-imdsv1": set(),
+        }
+
+        mock_queryset = MagicMock()
+        mock_queryset.values.return_value = mock_queryset
+        mock_queryset.annotate.return_value = []  # No findings
+
+        ctx = MagicMock()
+        ctx.__enter__.return_value = None
+        ctx.__exit__.return_value = False
+        mock_rls_transaction.return_value = ctx
+        mock_findings_filter.return_value = mock_queryset
+
+        aggregate_attack_surface(str(tenant.id), str(scan.id))
+
+        mock_bulk_create.assert_not_called()
+
+    @patch("tasks.jobs.scan.AttackSurfaceOverview.objects.bulk_create")
+    @patch("tasks.jobs.scan.Finding.all_objects.filter")
+    @patch("tasks.jobs.scan._get_attack_surface_mapping_from_provider")
+    @patch("tasks.jobs.scan.rls_transaction")
+    def test_aggregate_attack_surface_aggregates_counts_correctly(
+        self,
+        mock_rls_transaction,
+        mock_get_mapping,
+        mock_findings_filter,
+        mock_bulk_create,
+        tenants_fixture,
+        scans_fixture,
+    ):
+        """Test that counts from multiple check_ids are aggregated per attack surface type."""
+        tenant = tenants_fixture[0]
+        scan = scans_fixture[0]
+        scan.provider.provider = "aws"
+        scan.provider.save()
+
+        mock_get_mapping.return_value = {
+            "internet-exposed": {"check_internet_1", "check_internet_2"},
+            "secrets": set(),
+            "privilege-escalation": set(),
+            "ec2-imdsv1": set(),
+        }
+
+        mock_queryset = MagicMock()
+        mock_queryset.values.return_value = mock_queryset
+        mock_queryset.annotate.return_value = [
+            {"check_id": "check_internet_1", "total": 10, "failed": 3, "muted": 1},
+            {"check_id": "check_internet_2", "total": 5, "failed": 2, "muted": 0},
+        ]
+
+        ctx = MagicMock()
+        ctx.__enter__.return_value = None
+        ctx.__exit__.return_value = False
+        mock_rls_transaction.return_value = ctx
+        mock_findings_filter.return_value = mock_queryset
+
+        aggregate_attack_surface(str(tenant.id), str(scan.id))
+
+        args, kwargs = mock_bulk_create.call_args
+        objects = args[0]
+
+        assert len(objects) == 1
+        overview = objects[0]
+        assert overview.attack_surface_type == "internet-exposed"
+        assert overview.total_findings == 15  # 10 + 5
+        assert overview.failed_findings == 5  # 3 + 2
+        assert overview.muted_failed_findings == 1  # 1 + 0
+
+    @patch("tasks.jobs.scan.Scan.all_objects.select_related")
+    @patch("tasks.jobs.scan.rls_transaction")
+    def test_aggregate_attack_surface_uses_select_related(
+        self, mock_rls_transaction, mock_select_related, tenants_fixture, scans_fixture
+    ):
+        """Test that select_related is used to avoid N+1 query."""
+        tenant = tenants_fixture[0]
+        scan = scans_fixture[0]
+
+        mock_scan = MagicMock()
+        mock_scan.provider.provider = "aws"
+
+        mock_select_related.return_value.get.return_value = mock_scan
+
+        ctx = MagicMock()
+        ctx.__enter__.return_value = None
+        ctx.__exit__.return_value = False
+        mock_rls_transaction.return_value = ctx
+
+        with patch(
+            "tasks.jobs.scan._get_attack_surface_mapping_from_provider"
+        ) as mock_map:
+            mock_map.return_value = {}
+
+            aggregate_attack_surface(str(tenant.id), str(scan.id))
+
+        mock_select_related.assert_called_once_with("provider")
+
+
+class TestAggregateCategoryCounts:
+    """Test aggregate_category_counts helper function."""
+
+    def test_aggregate_category_counts_basic(self):
+        """Test basic category counting for a non-muted PASS finding."""
+        cache: dict[tuple[str, str], dict[str, int]] = {}
+        aggregate_category_counts(
+            categories=["security", "iam"],
+            severity="high",
+            status="PASS",
+            delta=None,
+            muted=False,
+            cache=cache,
+        )
+
+        assert ("security", "high") in cache
+        assert ("iam", "high") in cache
+        assert cache[("security", "high")] == {"total": 1, "failed": 0, "new_failed": 0}
+        assert cache[("iam", "high")] == {"total": 1, "failed": 0, "new_failed": 0}
+
+    def test_aggregate_category_counts_fail_not_muted(self):
+        """Test category counting for a non-muted FAIL finding."""
+        cache: dict[tuple[str, str], dict[str, int]] = {}
+        aggregate_category_counts(
+            categories=["security"],
+            severity="critical",
+            status="FAIL",
+            delta=None,
+            muted=False,
+            cache=cache,
+        )
+
+        assert cache[("security", "critical")] == {
+            "total": 1,
+            "failed": 1,
+            "new_failed": 0,
+        }
+
+    def test_aggregate_category_counts_new_fail(self):
+        """Test category counting for a new FAIL finding (delta='new')."""
+        cache: dict[tuple[str, str], dict[str, int]] = {}
+        aggregate_category_counts(
+            categories=["gen-ai"],
+            severity="high",
+            status="FAIL",
+            delta="new",
+            muted=False,
+            cache=cache,
+        )
+
+        assert cache[("gen-ai", "high")] == {"total": 1, "failed": 1, "new_failed": 1}
+
+    def test_aggregate_category_counts_muted_finding(self):
+        """Test that muted findings are excluded from all counts."""
+        cache: dict[tuple[str, str], dict[str, int]] = {}
+        aggregate_category_counts(
+            categories=["security"],
+            severity="high",
+            status="FAIL",
+            delta="new",
+            muted=True,
+            cache=cache,
+        )
+
+        assert cache[("security", "high")] == {"total": 0, "failed": 0, "new_failed": 0}
+
+    def test_aggregate_category_counts_accumulates(self):
+        """Test that multiple calls accumulate counts."""
+        cache: dict[tuple[str, str], dict[str, int]] = {}
+
+        # First finding: PASS
+        aggregate_category_counts(
+            categories=["security"],
+            severity="high",
+            status="PASS",
+            delta=None,
+            muted=False,
+            cache=cache,
+        )
+
+        # Second finding: FAIL (new)
+        aggregate_category_counts(
+            categories=["security"],
+            severity="high",
+            status="FAIL",
+            delta="new",
+            muted=False,
+            cache=cache,
+        )
+
+        # Third finding: FAIL (changed)
+        aggregate_category_counts(
+            categories=["security"],
+            severity="high",
+            status="FAIL",
+            delta="changed",
+            muted=False,
+            cache=cache,
+        )
+
+        assert cache[("security", "high")] == {"total": 3, "failed": 2, "new_failed": 1}
+
+    def test_aggregate_category_counts_empty_categories(self):
+        """Test with empty categories list."""
+        cache: dict[tuple[str, str], dict[str, int]] = {}
+        aggregate_category_counts(
+            categories=[],
+            severity="high",
+            status="FAIL",
+            delta="new",
+            muted=False,
+            cache=cache,
+        )
+
+        assert cache == {}
+
+    def test_aggregate_category_counts_changed_delta(self):
+        """Test that changed delta increments failed but not new_failed."""
+        cache: dict[tuple[str, str], dict[str, int]] = {}
+        aggregate_category_counts(
+            categories=["iam"],
+            severity="medium",
+            status="FAIL",
+            delta="changed",
+            muted=False,
+            cache=cache,
+        )
+
+        assert cache[("iam", "medium")] == {"total": 1, "failed": 1, "new_failed": 0}
+
+    def test_aggregate_category_counts_multiple_categories_single_finding(self):
+        """Test single finding with multiple categories."""
+        cache: dict[tuple[str, str], dict[str, int]] = {}
+        aggregate_category_counts(
+            categories=["security", "compliance", "data-protection"],
+            severity="low",
+            status="FAIL",
+            delta="new",
+            muted=False,
+            cache=cache,
+        )
+
+        assert len(cache) == 3
+        for cat in ["security", "compliance", "data-protection"]:
+            assert cache[(cat, "low")] == {"total": 1, "failed": 1, "new_failed": 1}
+
+
+@pytest.mark.django_db
+class TestUpdateProviderComplianceScores:
+    @patch("tasks.jobs.scan.psycopg_connection")
+    def test_update_provider_compliance_scores_basic(
+        self,
+        mock_psycopg_connection,
+        tenants_fixture,
+        scans_fixture,
+        settings,
+    ):
+        settings.DATABASES.setdefault("admin", settings.DATABASES["default"])
+        tenant = tenants_fixture[0]
+        scan = scans_fixture[0]
+        tenant_id = str(tenant.id)
+        scan_id = str(scan.id)
+
+        scan.state = StateChoices.COMPLETED
+        scan.completed_at = datetime.now(timezone.utc)
+        scan.save()
+
+        connection = MagicMock()
+        cursor = MagicMock()
+        cursor_context = MagicMock()
+        cursor_context.__enter__.return_value = cursor
+        cursor_context.__exit__.return_value = False
+        connection.cursor.return_value = cursor_context
+        connection.__enter__.return_value = connection
+        connection.__exit__.return_value = False
+        connection.autocommit = True
+
+        context_manager = MagicMock()
+        context_manager.__enter__.return_value = connection
+        context_manager.__exit__.return_value = False
+        mock_psycopg_connection.return_value = context_manager
+
+        cursor.rowcount = 2
+
+        result = update_provider_compliance_scores(tenant_id, scan_id)
+
+        assert result["status"] == "completed"
+        assert result["upserted"] == 2
+        assert cursor.execute.call_count >= 3
+        connection.commit.assert_called_once()
+
+    def test_update_provider_compliance_scores_skips_incomplete_scan(
+        self, tenants_fixture, scans_fixture
+    ):
+        tenant = tenants_fixture[0]
+        scan = scans_fixture[1]
+        tenant_id = str(tenant.id)
+        scan_id = str(scan.id)
+
+        result = update_provider_compliance_scores(tenant_id, scan_id)
+
+        assert result["status"] == "skipped"
+        assert result["reason"] == "scan not completed"
+
+    def test_update_provider_compliance_scores_skips_no_completed_at(
+        self, tenants_fixture, scans_fixture
+    ):
+        tenant = tenants_fixture[0]
+        scan = scans_fixture[0]
+        tenant_id = str(tenant.id)
+        scan_id = str(scan.id)
+
+        scan.state = StateChoices.COMPLETED
+        scan.completed_at = None
+        scan.save()
+
+        result = update_provider_compliance_scores(tenant_id, scan_id)
+
+        assert result["status"] == "skipped"
+        assert result["reason"] == "no completed_at"
+
+    @patch("tasks.jobs.scan.psycopg_connection")
+    def test_update_provider_compliance_scores_executes_sql_queries(
+        self,
+        mock_psycopg_connection,
+        tenants_fixture,
+        providers_fixture,
+        scans_fixture,
+        settings,
+    ):
+        settings.DATABASES.setdefault("admin", settings.DATABASES["default"])
+        tenant = tenants_fixture[0]
+        scan = scans_fixture[0]
+        tenant_id = str(tenant.id)
+        scan_id = str(scan.id)
+
+        scan.state = StateChoices.COMPLETED
+        scan.completed_at = datetime.now(timezone.utc)
+        scan.save()
+
+        connection = MagicMock()
+        cursor = MagicMock()
+        cursor_context = MagicMock()
+        cursor_context.__enter__.return_value = cursor
+        cursor_context.__exit__.return_value = False
+        connection.cursor.return_value = cursor_context
+        connection.__enter__.return_value = connection
+        connection.__exit__.return_value = False
+
+        context_manager = MagicMock()
+        context_manager.__enter__.return_value = connection
+        context_manager.__exit__.return_value = False
+        mock_psycopg_connection.return_value = context_manager
+
+        cursor.rowcount = 1
+        cursor.fetchall.side_effect = [[("aws_cis_2.0",)], []]
+
+        result = update_provider_compliance_scores(tenant_id, scan_id)
+
+        assert result["status"] == "completed"
+
+        calls = [str(c) for c in cursor.execute.call_args_list]
+        assert any("provider_compliance_scores" in c for c in calls)
+        assert any("tenant_compliance_summaries" in c for c in calls)
+        assert any("pg_advisory_xact_lock" in c for c in calls)

api/src/backend/tasks/tests/test_tasks.py

@@ -7,13 +7,21 @@ import openai
 import pytest
 
 from botocore.exceptions import ClientError
+from django_celery_beat.models import IntervalSchedule, PeriodicTask
 
 from api.models import (
     Integration,
     LighthouseProviderConfiguration,
     LighthouseProviderModels,
+    Scan,
+    StateChoices,
+)
+from tasks.jobs.lighthouse_providers import (
+    _create_bedrock_client,
+    _extract_bedrock_credentials,
 )
 from tasks.tasks import (
+    _cleanup_orphan_scheduled_scans,
     _perform_scan_complete_tasks,
     check_integrations_task,
     check_lighthouse_provider_connection_task,
@@ -25,6 +33,198 @@ from tasks.tasks import (
 )
 
 
+@pytest.mark.django_db
+class TestExtractBedrockCredentials:
+    """Unit tests for _extract_bedrock_credentials helper function."""
+
+    def test_extract_access_key_credentials(self, tenants_fixture):
+        """Test extraction of access key + secret key credentials."""
+        provider_cfg = LighthouseProviderConfiguration(
+            tenant_id=tenants_fixture[0].id,
+            provider_type=LighthouseProviderConfiguration.LLMProviderChoices.BEDROCK,
+            is_active=True,
+        )
+        provider_cfg.credentials_decoded = {
+            "access_key_id": "AKIAIOSFODNN7EXAMPLE",
+            "secret_access_key": "wJalrXUtnFEMI/K7MDENG/bPxRfiCYEXAMPLEKEY",
+            "region": "us-east-1",
+        }
+        provider_cfg.save()
+
+        result = _extract_bedrock_credentials(provider_cfg)
+
+        assert result is not None
+        assert result["access_key_id"] == "AKIAIOSFODNN7EXAMPLE"
+        assert result["secret_access_key"] == "wJalrXUtnFEMI/K7MDENG/bPxRfiCYEXAMPLEKEY"
+        assert result["region"] == "us-east-1"
+        assert "api_key" not in result
+
+    def test_extract_api_key_credentials(self, tenants_fixture):
+        """Test extraction of API key (bearer token) credentials."""
+        valid_api_key = "ABSKQmVkcm9ja0FQSUtleS" + ("A" * 110)
+        provider_cfg = LighthouseProviderConfiguration(
+            tenant_id=tenants_fixture[0].id,
+            provider_type=LighthouseProviderConfiguration.LLMProviderChoices.BEDROCK,
+            is_active=True,
+        )
+        provider_cfg.credentials_decoded = {
+            "api_key": valid_api_key,
+            "region": "us-west-2",
+        }
+        provider_cfg.save()
+
+        result = _extract_bedrock_credentials(provider_cfg)
+
+        assert result is not None
+        assert result["api_key"] == valid_api_key
+        assert result["region"] == "us-west-2"
+        assert "access_key_id" not in result
+        assert "secret_access_key" not in result
+
+    def test_api_key_takes_precedence_over_access_keys(self, tenants_fixture):
+        """Test that API key is preferred when both auth methods are present."""
+        valid_api_key = "ABSKQmVkcm9ja0FQSUtleS" + ("B" * 110)
+        provider_cfg = LighthouseProviderConfiguration(
+            tenant_id=tenants_fixture[0].id,
+            provider_type=LighthouseProviderConfiguration.LLMProviderChoices.BEDROCK,
+            is_active=True,
+        )
+        provider_cfg.credentials_decoded = {
+            "api_key": valid_api_key,
+            "access_key_id": "AKIAIOSFODNN7EXAMPLE",
+            "secret_access_key": "wJalrXUtnFEMI/K7MDENG/bPxRfiCYEXAMPLEKEY",
+            "region": "eu-west-1",
+        }
+        provider_cfg.save()
+
+        result = _extract_bedrock_credentials(provider_cfg)
+
+        assert result is not None
+        assert result["api_key"] == valid_api_key
+        assert result["region"] == "eu-west-1"
+        assert "access_key_id" not in result
+
+    def test_missing_region_returns_none(self, tenants_fixture):
+        """Test that missing region returns None."""
+        provider_cfg = LighthouseProviderConfiguration(
+            tenant_id=tenants_fixture[0].id,
+            provider_type=LighthouseProviderConfiguration.LLMProviderChoices.BEDROCK,
+            is_active=True,
+        )
+        provider_cfg.credentials_decoded = {
+            "api_key": "ABSKQmVkcm9ja0FQSUtleS" + ("A" * 110),
+        }
+        provider_cfg.save()
+
+        result = _extract_bedrock_credentials(provider_cfg)
+
+        assert result is None
+
+    def test_empty_credentials_returns_none(self, tenants_fixture):
+        """Test that empty credentials dict returns None (region only is not enough)."""
+        provider_cfg = LighthouseProviderConfiguration(
+            tenant_id=tenants_fixture[0].id,
+            provider_type=LighthouseProviderConfiguration.LLMProviderChoices.BEDROCK,
+            is_active=True,
+        )
+        # Only region, no auth credentials - should return None
+        provider_cfg.credentials_decoded = {
+            "region": "us-east-1",
+        }
+        provider_cfg.save()
+
+        result = _extract_bedrock_credentials(provider_cfg)
+
+        assert result is None
+
+    def test_non_dict_credentials_returns_none(self, tenants_fixture):
+        """Test that non-dict credentials returns None."""
+        provider_cfg = LighthouseProviderConfiguration(
+            tenant_id=tenants_fixture[0].id,
+            provider_type=LighthouseProviderConfiguration.LLMProviderChoices.BEDROCK,
+            is_active=True,
+        )
+        # Store valid credentials first to pass model validation
+        provider_cfg.credentials_decoded = {
+            "api_key": "ABSKQmVkcm9ja0FQSUtleS" + ("A" * 110),
+            "region": "us-east-1",
+        }
+        provider_cfg.save()
+
+        # Mock the credentials_decoded property to return a non-dict value
+        # This simulates corrupted/invalid stored data
+        with patch.object(
+            type(provider_cfg),
+            "credentials_decoded",
+            new_callable=lambda: property(lambda self: "invalid"),
+        ):
+            result = _extract_bedrock_credentials(provider_cfg)
+
+        assert result is None
+
+
+class TestCreateBedrockClient:
+    """Unit tests for _create_bedrock_client helper function."""
+
+    @patch("tasks.jobs.lighthouse_providers.boto3.client")
+    def test_create_client_with_access_keys(self, mock_boto_client):
+        """Test creating client with access key authentication."""
+        mock_client = MagicMock()
+        mock_boto_client.return_value = mock_client
+
+        creds = {
+            "access_key_id": "AKIAIOSFODNN7EXAMPLE",
+            "secret_access_key": "wJalrXUtnFEMI/K7MDENG/bPxRfiCYEXAMPLEKEY",
+            "region": "us-east-1",
+        }
+
+        result = _create_bedrock_client(creds)
+
+        assert result == mock_client
+        mock_boto_client.assert_called_once_with(
+            service_name="bedrock",
+            region_name="us-east-1",
+            aws_access_key_id="AKIAIOSFODNN7EXAMPLE",
+            aws_secret_access_key="wJalrXUtnFEMI/K7MDENG/bPxRfiCYEXAMPLEKEY",
+        )
+
+    @patch("tasks.jobs.lighthouse_providers.Config")
+    @patch("tasks.jobs.lighthouse_providers.boto3.client")
+    def test_create_client_with_api_key(self, mock_boto_client, mock_config):
+        """Test creating client with API key authentication."""
+        mock_client = MagicMock()
+        mock_events = MagicMock()
+        mock_client.meta.events = mock_events
+        mock_boto_client.return_value = mock_client
+        mock_config_instance = MagicMock()
+        mock_config.return_value = mock_config_instance
+        valid_api_key = "ABSKQmVkcm9ja0FQSUtleS" + ("A" * 110)
+
+        creds = {
+            "api_key": valid_api_key,
+            "region": "us-west-2",
+        }
+
+        result = _create_bedrock_client(creds)
+
+        assert result == mock_client
+        mock_boto_client.assert_called_once_with(
+            service_name="bedrock",
+            region_name="us-west-2",
+            config=mock_config_instance,
+        )
+        mock_events.register.assert_called_once()
+        call_args = mock_events.register.call_args
+        assert call_args[0][0] == "before-send.*.*"
+
+        # Verify handler injects bearer token
+        handler_fn = call_args[0][1]
+        mock_request = MagicMock()
+        mock_request.headers = {}
+        handler_fn(mock_request)
+        assert mock_request.headers["Authorization"] == f"Bearer {valid_api_key}"
+
+
 # TODO Move this to outputs/reports jobs
 @pytest.mark.django_db
 class TestGenerateOutputs:
@@ -533,8 +733,15 @@ class TestGenerateOutputs:
 
 
 class TestScanCompleteTasks:
+<<<<<<< HEAD
     @patch("tasks.tasks.perform_attack_paths_scan_task.apply_async")
     @patch("tasks.tasks.create_compliance_requirements_task.apply_async")
+=======
+    @patch("tasks.tasks.aggregate_attack_surface_task.apply_async")
+    @patch("tasks.tasks.chain")
+    @patch("tasks.tasks.create_compliance_requirements_task.si")
+    @patch("tasks.tasks.update_provider_compliance_scores_task.si")
+>>>>>>> 1bf49747adaefcb19db66274478f6933342112c1
     @patch("tasks.tasks.perform_scan_summary_task.si")
     @patch("tasks.tasks.generate_outputs_task.si")
     @patch("tasks.tasks.generate_compliance_reports_task.si")
@@ -545,14 +752,30 @@ class TestScanCompleteTasks:
         mock_compliance_reports_task,
         mock_outputs_task,
         mock_scan_summary_task,
+        mock_update_compliance_scores_task,
         mock_compliance_requirements_task,
+<<<<<<< HEAD
         mock_attack_paths_task,
+=======
+        mock_chain,
+        mock_attack_surface_task,
+>>>>>>> 1bf49747adaefcb19db66274478f6933342112c1
     ):
         """Test that scan complete tasks are properly orchestrated with optimized reports."""
         _perform_scan_complete_tasks("tenant-id", "scan-id", "provider-id")
 
-        # Verify compliance requirements task is called
+        # Verify compliance requirements task is called via chain
         mock_compliance_requirements_task.assert_called_once_with(
+            tenant_id="tenant-id", scan_id="scan-id"
+        )
+
+        # Verify update provider compliance scores task is called via chain
+        mock_update_compliance_scores_task.assert_called_once_with(
+            tenant_id="tenant-id", scan_id="scan-id"
+        )
+
+        # Verify attack surface task is called
+        mock_attack_surface_task.assert_called_once_with(
             kwargs={"tenant_id": "tenant-id", "scan_id": "scan-id"},
         )
 
@@ -1213,6 +1436,16 @@ class TestCheckLighthouseProviderConnectionTask:
                 None,
                 {"connected": True, "error": None},
             ),
+            # Bedrock API key authentication
+            (
+                LighthouseProviderConfiguration.LLMProviderChoices.BEDROCK,
+                {
+                    "api_key": "ABSKQmVkcm9ja0FQSUtleS" + ("A" * 110),
+                    "region": "us-east-1",
+                },
+                None,
+                {"connected": True, "error": None},
+            ),
         ],
     )
     def test_check_connection_success_all_providers(
@@ -1281,6 +1514,24 @@ class TestCheckLighthouseProviderConnectionTask:
                     "list_foundation_models",
                 ),
             ),
+            # Bedrock API key authentication failure
+            (
+                LighthouseProviderConfiguration.LLMProviderChoices.BEDROCK,
+                {
+                    "api_key": "ABSKQmVkcm9ja0FQSUtleS" + ("X" * 110),
+                    "region": "us-east-1",
+                },
+                None,
+                ClientError(
+                    {
+                        "Error": {
+                            "Code": "UnrecognizedClientException",
+                            "Message": "Invalid API key",
+                        }
+                    },
+                    "list_foundation_models",
+                ),
+            ),
         ],
     )
     def test_check_connection_api_failure(
@@ -1405,6 +1656,17 @@ class TestRefreshLighthouseProviderModelsTask:
                 {"openai.gpt-oss-120b-1:0": "gpt-oss-120b"},
                 1,
             ),
+            # Bedrock API key authentication
+            (
+                LighthouseProviderConfiguration.LLMProviderChoices.BEDROCK,
+                {
+                    "api_key": "ABSKQmVkcm9ja0FQSUtleS" + ("A" * 110),
+                    "region": "us-east-1",
+                },
+                None,
+                {"anthropic.claude-v3": "Claude 3"},
+                1,
+            ),
         ],
     )
     def test_refresh_models_create_new(
@@ -1541,3 +1803,343 @@ class TestRefreshLighthouseProviderModelsTask:
             assert result["deleted"] == 0
             assert "error" in result
             assert result["error"] is not None
+
+
+@pytest.mark.django_db
+class TestCleanupOrphanScheduledScans:
+    """Unit tests for _cleanup_orphan_scheduled_scans helper function."""
+
+    def _create_periodic_task(self, provider_id, tenant_id):
+        """Helper to create a PeriodicTask for testing."""
+        interval, _ = IntervalSchedule.objects.get_or_create(every=24, period="hours")
+        return PeriodicTask.objects.create(
+            name=f"scan-perform-scheduled-{provider_id}",
+            task="scan-perform-scheduled",
+            interval=interval,
+            kwargs=f'{{"tenant_id": "{tenant_id}", "provider_id": "{provider_id}"}}',
+            enabled=True,
+        )
+
+    def test_cleanup_deletes_orphan_when_both_available_and_scheduled_exist(
+        self, tenants_fixture, providers_fixture
+    ):
+        """Test that AVAILABLE scan is deleted when SCHEDULED also exists."""
+        tenant = tenants_fixture[0]
+        provider = providers_fixture[0]
+        periodic_task = self._create_periodic_task(provider.id, tenant.id)
+
+        # Create orphan AVAILABLE scan
+        orphan_scan = Scan.objects.create(
+            tenant_id=tenant.id,
+            provider=provider,
+            name="Daily scheduled scan",
+            trigger=Scan.TriggerChoices.SCHEDULED,
+            state=StateChoices.AVAILABLE,
+            scheduler_task_id=periodic_task.id,
+        )
+
+        # Create SCHEDULED scan (next execution)
+        scheduled_scan = Scan.objects.create(
+            tenant_id=tenant.id,
+            provider=provider,
+            name="Daily scheduled scan",
+            trigger=Scan.TriggerChoices.SCHEDULED,
+            state=StateChoices.SCHEDULED,
+            scheduler_task_id=periodic_task.id,
+        )
+
+        # Execute cleanup
+        deleted_count = _cleanup_orphan_scheduled_scans(
+            tenant_id=str(tenant.id),
+            provider_id=str(provider.id),
+            scheduler_task_id=periodic_task.id,
+        )
+
+        # Verify orphan was deleted
+        assert deleted_count == 1
+        assert not Scan.objects.filter(id=orphan_scan.id).exists()
+        assert Scan.objects.filter(id=scheduled_scan.id).exists()
+
+    def test_cleanup_does_not_delete_when_only_available_exists(
+        self, tenants_fixture, providers_fixture
+    ):
+        """Test that AVAILABLE scan is NOT deleted when no SCHEDULED exists."""
+        tenant = tenants_fixture[0]
+        provider = providers_fixture[0]
+        periodic_task = self._create_periodic_task(provider.id, tenant.id)
+
+        # Create only AVAILABLE scan (normal first scan scenario)
+        available_scan = Scan.objects.create(
+            tenant_id=tenant.id,
+            provider=provider,
+            name="Daily scheduled scan",
+            trigger=Scan.TriggerChoices.SCHEDULED,
+            state=StateChoices.AVAILABLE,
+            scheduler_task_id=periodic_task.id,
+        )
+
+        # Execute cleanup
+        deleted_count = _cleanup_orphan_scheduled_scans(
+            tenant_id=str(tenant.id),
+            provider_id=str(provider.id),
+            scheduler_task_id=periodic_task.id,
+        )
+
+        # Verify nothing was deleted
+        assert deleted_count == 0
+        assert Scan.objects.filter(id=available_scan.id).exists()
+
+    def test_cleanup_does_not_delete_when_only_scheduled_exists(
+        self, tenants_fixture, providers_fixture
+    ):
+        """Test that nothing is deleted when only SCHEDULED exists."""
+        tenant = tenants_fixture[0]
+        provider = providers_fixture[0]
+        periodic_task = self._create_periodic_task(provider.id, tenant.id)
+
+        # Create only SCHEDULED scan (normal subsequent scan scenario)
+        scheduled_scan = Scan.objects.create(
+            tenant_id=tenant.id,
+            provider=provider,
+            name="Daily scheduled scan",
+            trigger=Scan.TriggerChoices.SCHEDULED,
+            state=StateChoices.SCHEDULED,
+            scheduler_task_id=periodic_task.id,
+        )
+
+        # Execute cleanup
+        deleted_count = _cleanup_orphan_scheduled_scans(
+            tenant_id=str(tenant.id),
+            provider_id=str(provider.id),
+            scheduler_task_id=periodic_task.id,
+        )
+
+        # Verify nothing was deleted
+        assert deleted_count == 0
+        assert Scan.objects.filter(id=scheduled_scan.id).exists()
+
+    def test_cleanup_returns_zero_when_no_scans_exist(
+        self, tenants_fixture, providers_fixture
+    ):
+        """Test that cleanup returns 0 when no scans exist."""
+        tenant = tenants_fixture[0]
+        provider = providers_fixture[0]
+        periodic_task = self._create_periodic_task(provider.id, tenant.id)
+
+        # Execute cleanup with no scans
+        deleted_count = _cleanup_orphan_scheduled_scans(
+            tenant_id=str(tenant.id),
+            provider_id=str(provider.id),
+            scheduler_task_id=periodic_task.id,
+        )
+
+        assert deleted_count == 0
+
+    def test_cleanup_deletes_multiple_orphan_available_scans(
+        self, tenants_fixture, providers_fixture
+    ):
+        """Test that multiple AVAILABLE orphan scans are all deleted."""
+        tenant = tenants_fixture[0]
+        provider = providers_fixture[0]
+        periodic_task = self._create_periodic_task(provider.id, tenant.id)
+
+        # Create multiple orphan AVAILABLE scans
+        orphan_scan_1 = Scan.objects.create(
+            tenant_id=tenant.id,
+            provider=provider,
+            name="Daily scheduled scan",
+            trigger=Scan.TriggerChoices.SCHEDULED,
+            state=StateChoices.AVAILABLE,
+            scheduler_task_id=periodic_task.id,
+        )
+        orphan_scan_2 = Scan.objects.create(
+            tenant_id=tenant.id,
+            provider=provider,
+            name="Daily scheduled scan",
+            trigger=Scan.TriggerChoices.SCHEDULED,
+            state=StateChoices.AVAILABLE,
+            scheduler_task_id=periodic_task.id,
+        )
+
+        # Create SCHEDULED scan
+        scheduled_scan = Scan.objects.create(
+            tenant_id=tenant.id,
+            provider=provider,
+            name="Daily scheduled scan",
+            trigger=Scan.TriggerChoices.SCHEDULED,
+            state=StateChoices.SCHEDULED,
+            scheduler_task_id=periodic_task.id,
+        )
+
+        # Execute cleanup
+        deleted_count = _cleanup_orphan_scheduled_scans(
+            tenant_id=str(tenant.id),
+            provider_id=str(provider.id),
+            scheduler_task_id=periodic_task.id,
+        )
+
+        # Verify all orphans were deleted
+        assert deleted_count == 2
+        assert not Scan.objects.filter(id=orphan_scan_1.id).exists()
+        assert not Scan.objects.filter(id=orphan_scan_2.id).exists()
+        assert Scan.objects.filter(id=scheduled_scan.id).exists()
+
+    def test_cleanup_does_not_affect_different_provider(
+        self, tenants_fixture, providers_fixture
+    ):
+        """Test that cleanup only affects scans for the specified provider."""
+        tenant = tenants_fixture[0]
+        provider1 = providers_fixture[0]
+        provider2 = providers_fixture[1]
+        periodic_task1 = self._create_periodic_task(provider1.id, tenant.id)
+        periodic_task2 = self._create_periodic_task(provider2.id, tenant.id)
+
+        # Create orphan scenario for provider1
+        orphan_scan_p1 = Scan.objects.create(
+            tenant_id=tenant.id,
+            provider=provider1,
+            name="Daily scheduled scan",
+            trigger=Scan.TriggerChoices.SCHEDULED,
+            state=StateChoices.AVAILABLE,
+            scheduler_task_id=periodic_task1.id,
+        )
+        scheduled_scan_p1 = Scan.objects.create(
+            tenant_id=tenant.id,
+            provider=provider1,
+            name="Daily scheduled scan",
+            trigger=Scan.TriggerChoices.SCHEDULED,
+            state=StateChoices.SCHEDULED,
+            scheduler_task_id=periodic_task1.id,
+        )
+
+        # Create AVAILABLE scan for provider2 (should not be affected)
+        available_scan_p2 = Scan.objects.create(
+            tenant_id=tenant.id,
+            provider=provider2,
+            name="Daily scheduled scan",
+            trigger=Scan.TriggerChoices.SCHEDULED,
+            state=StateChoices.AVAILABLE,
+            scheduler_task_id=periodic_task2.id,
+        )
+
+        # Execute cleanup for provider1 only
+        deleted_count = _cleanup_orphan_scheduled_scans(
+            tenant_id=str(tenant.id),
+            provider_id=str(provider1.id),
+            scheduler_task_id=periodic_task1.id,
+        )
+
+        # Verify only provider1's orphan was deleted
+        assert deleted_count == 1
+        assert not Scan.objects.filter(id=orphan_scan_p1.id).exists()
+        assert Scan.objects.filter(id=scheduled_scan_p1.id).exists()
+        assert Scan.objects.filter(id=available_scan_p2.id).exists()
+
+    def test_cleanup_does_not_affect_manual_scans(
+        self, tenants_fixture, providers_fixture
+    ):
+        """Test that cleanup only affects SCHEDULED trigger scans, not MANUAL."""
+        tenant = tenants_fixture[0]
+        provider = providers_fixture[0]
+        periodic_task = self._create_periodic_task(provider.id, tenant.id)
+
+        # Create orphan AVAILABLE scheduled scan
+        orphan_scan = Scan.objects.create(
+            tenant_id=tenant.id,
+            provider=provider,
+            name="Daily scheduled scan",
+            trigger=Scan.TriggerChoices.SCHEDULED,
+            state=StateChoices.AVAILABLE,
+            scheduler_task_id=periodic_task.id,
+        )
+
+        # Create SCHEDULED scan
+        scheduled_scan = Scan.objects.create(
+            tenant_id=tenant.id,
+            provider=provider,
+            name="Daily scheduled scan",
+            trigger=Scan.TriggerChoices.SCHEDULED,
+            state=StateChoices.SCHEDULED,
+            scheduler_task_id=periodic_task.id,
+        )
+
+        # Create AVAILABLE manual scan (should not be affected)
+        manual_scan = Scan.objects.create(
+            tenant_id=tenant.id,
+            provider=provider,
+            name="Manual scan",
+            trigger=Scan.TriggerChoices.MANUAL,
+            state=StateChoices.AVAILABLE,
+        )
+
+        # Execute cleanup
+        deleted_count = _cleanup_orphan_scheduled_scans(
+            tenant_id=str(tenant.id),
+            provider_id=str(provider.id),
+            scheduler_task_id=periodic_task.id,
+        )
+
+        # Verify only scheduled orphan was deleted
+        assert deleted_count == 1
+        assert not Scan.objects.filter(id=orphan_scan.id).exists()
+        assert Scan.objects.filter(id=scheduled_scan.id).exists()
+        assert Scan.objects.filter(id=manual_scan.id).exists()
+
+    def test_cleanup_does_not_affect_different_scheduler_task(
+        self, tenants_fixture, providers_fixture
+    ):
+        """Test that cleanup only affects scans with the specified scheduler_task_id."""
+        tenant = tenants_fixture[0]
+        provider = providers_fixture[0]
+        periodic_task1 = self._create_periodic_task(provider.id, tenant.id)
+
+        # Create another periodic task
+        interval, _ = IntervalSchedule.objects.get_or_create(every=24, period="hours")
+        periodic_task2 = PeriodicTask.objects.create(
+            name=f"scan-perform-scheduled-other-{provider.id}",
+            task="scan-perform-scheduled",
+            interval=interval,
+            kwargs=f'{{"tenant_id": "{tenant.id}", "provider_id": "{provider.id}"}}',
+            enabled=True,
+        )
+
+        # Create orphan scenario for periodic_task1
+        orphan_scan = Scan.objects.create(
+            tenant_id=tenant.id,
+            provider=provider,
+            name="Daily scheduled scan",
+            trigger=Scan.TriggerChoices.SCHEDULED,
+            state=StateChoices.AVAILABLE,
+            scheduler_task_id=periodic_task1.id,
+        )
+        scheduled_scan = Scan.objects.create(
+            tenant_id=tenant.id,
+            provider=provider,
+            name="Daily scheduled scan",
+            trigger=Scan.TriggerChoices.SCHEDULED,
+            state=StateChoices.SCHEDULED,
+            scheduler_task_id=periodic_task1.id,
+        )
+
+        # Create AVAILABLE scan for periodic_task2 (should not be affected)
+        available_scan_other_task = Scan.objects.create(
+            tenant_id=tenant.id,
+            provider=provider,
+            name="Daily scheduled scan",
+            trigger=Scan.TriggerChoices.SCHEDULED,
+            state=StateChoices.AVAILABLE,
+            scheduler_task_id=periodic_task2.id,
+        )
+
+        # Execute cleanup for periodic_task1 only
+        deleted_count = _cleanup_orphan_scheduled_scans(
+            tenant_id=str(tenant.id),
+            provider_id=str(provider.id),
+            scheduler_task_id=periodic_task1.id,
+        )
+
+        # Verify only periodic_task1's orphan was deleted
+        assert deleted_count == 1
+        assert not Scan.objects.filter(id=orphan_scan.id).exists()
+        assert Scan.objects.filter(id=scheduled_scan.id).exists()
+        assert Scan.objects.filter(id=available_scan_other_task.id).exists()