fix: stop flagging default-resolved fields as GraphQL N+1

src/debug_toolbar/extras/strawberry/analyzers.py

@@ -31,6 +31,9 @@ def __init__(self, threshold: int = 3) -> None:
     def analyze(self, operations: list[TrackedOperation]) -> list[dict[str, Any]]:
         """Analyze operations for N+1 patterns.
 
+        Default-resolved fields (plain attribute reads with no custom
+        resolver) are excluded since they never perform per-item I/O.
+
         Args:
             operations: List of tracked operations.
 
@@ -53,6 +56,8 @@ def analyze(self, operations: list[TrackedOperation]) -> list[dict[str, Any]]:
             for resolver in operation.resolvers:
                 if resolver.parent_type.startswith("__"):
                     continue
+                if resolver.is_default_resolver:
+                    continue
                 arg_sig = self._get_argument_signature(resolver.arguments)
                 signature = f"{resolver.parent_type}.{resolver.field_name}({arg_sig})"
 

src/debug_toolbar/extras/strawberry/extension.py

@@ -23,12 +23,17 @@
 
 try:
     from strawberry.extensions import SchemaExtension as _SchemaExtension
+    from strawberry.resolvers import is_default_resolver
 
     STRAWBERRY_AVAILABLE = True
 except ImportError:
     STRAWBERRY_AVAILABLE = False
     _SchemaExtension = object  # type: ignore[assignment, misc]
 
+    def is_default_resolver(_func: Callable[..., Any]) -> bool:  # type: ignore[misc]
+        """Fallback when Strawberry is unavailable."""
+        return False
+
 
 class DebugToolbarExtension(_SchemaExtension):  # type: ignore[misc]
     """Strawberry extension for tracking GraphQL operations and resolvers.
@@ -206,6 +211,7 @@ def resolve(
             parent_type=parent_type_name,
             return_type=return_type_str,
             arguments=dict(kwargs),
+            is_default_resolver=is_default_resolver(_next),
         )
 
         if self._capture_stacks:

src/debug_toolbar/extras/strawberry/models.py

@@ -23,6 +23,7 @@ class TrackedResolver:
         duration_ms: Execution duration in milliseconds.
         stack_trace: Call stack trace (if captured).
         is_slow: Whether duration exceeds threshold.
+        is_default_resolver: Whether this is Strawberry's default attribute resolver.
     """
 
     resolver_id: str
@@ -37,6 +38,7 @@ class TrackedResolver:
     duration_ms: float = 0.0
     stack_trace: list[dict[str, Any]] | None = None
     is_slow: bool = False
+    is_default_resolver: bool = False
 
     def to_dict(self) -> dict[str, Any]:
         """Convert to dictionary for serialization.
@@ -57,6 +59,7 @@ def to_dict(self) -> dict[str, Any]:
             "duration_ms": self.duration_ms,
             "stack_trace": self.stack_trace,
             "is_slow": self.is_slow,
+            "is_default_resolver": self.is_default_resolver,
         }
 
 

tests/unit/extras/strawberry/test_analyzers.py

@@ -260,6 +260,64 @@ def test_ignores_introspection_fields(self) -> None:
 
         assert N1Analyzer(threshold=3).analyze([operation]) == []
 
+    def test_ignores_default_resolvers(self) -> None:
+        """Should not flag plain attribute reads with no custom resolver."""
+        resolvers = [
+            TrackedResolver(
+                resolver_id=f"r{i}",
+                field_name="id",
+                field_path=f"Query.posts.{i}.id",
+                resolver_function="Post.id",
+                parent_type="Post",
+                return_type="String",
+                duration_ms=0.01,
+                is_default_resolver=True,
+            )
+            for i in range(5)
+        ]
+        operation = TrackedOperation(
+            operation_id="op1",
+            query="query { posts { id } }",
+            variables={},
+            operation_name=None,
+            operation_type="query",
+            start_time=0.0,
+            resolvers=resolvers,
+        )
+
+        assert N1Analyzer(threshold=3).analyze([operation]) == []
+
+    def test_flags_custom_resolver_called_per_item(self) -> None:
+        """Should still flag a custom (non-default) resolver run once per list item."""
+        resolvers = [
+            TrackedResolver(
+                resolver_id=f"r{i}",
+                field_name="author",
+                field_path=f"Query.posts.{i}.author",
+                resolver_function="Post.author",
+                parent_type="Post",
+                return_type="User",
+                duration_ms=10.0,
+                is_default_resolver=False,
+            )
+            for i in range(3)
+        ]
+        operation = TrackedOperation(
+            operation_id="op1",
+            query="query { posts { author { name } } }",
+            variables={},
+            operation_name=None,
+            operation_type="query",
+            start_time=0.0,
+            resolvers=resolvers,
+        )
+
+        patterns = N1Analyzer(threshold=3).analyze([operation])
+
+        assert len(patterns) == 1
+        assert patterns[0]["field_name"] == "author"
+        assert patterns[0]["count"] == 3
+
 
 class TestDuplicateDetector:
     """Tests for DuplicateDetector."""