diff --git a/core/wren/src/wren/memory/seed_queries.py b/core/wren/src/wren/memory/seed_queries.py
index 69a0f391dc..cfc8a2e26d 100644
--- a/core/wren/src/wren/memory/seed_queries.py
+++ b/core/wren/src/wren/memory/seed_queries.py
@@ -58,7 +58,9 @@ def _model_seeds(
     model: dict, relationship_keys: frozenset[str] = frozenset()
 ) -> list[dict]:
     name = model["name"]
-    columns = model.get("columns", [])
+    columns = [
+        c for c in (model.get("columns") or []) if isinstance(c, dict) and c.get("name")
+    ]
     primary_keys = _primary_key_columns(model)
     pairs = []
 
@@ -74,14 +76,16 @@ def _model_seeds(
     numeric_col = None
     group_col = None
     for col in columns:
-        col_name = col["name"]
-        norm_name = _norm_ident(col_name)
+        col_name = col.get("name")
+        norm_name = _norm_ident(str(col_name))
         col_type = (col.get("type") or "").split("(")[0].lower().strip()
         is_calc = col.get("isCalculated", False)
         is_pk = norm_name in primary_keys
         # Identifiers are numeric by storage but not measures: summing a join
         # key (e.g. SUM(customer_id)) is semantically meaningless.
-        is_identifier = is_pk or norm_name in relationship_keys or _is_id_like(col_name)
+        is_identifier = (
+            is_pk or norm_name in relationship_keys or _is_id_like(str(col_name))
+        )
 
         if (
             col_type in _NUMERIC_TYPES
@@ -117,6 +121,7 @@ def _model_seeds(
         )
 
     for col in columns:
+        col_name = col.get("name")
         accepted_values = _prop_raw(col, "acceptedValues", "accepted_values")
         first_value = _first_accepted_value(accepted_values)
         if not first_value:
@@ -124,8 +129,8 @@ def _model_seeds(
         quoted = first_value.replace("'", "''")
         pairs.append(
             {
-                "nl": f"Show {name} where {col['name']} is {first_value}",
-                "sql": f"SELECT * FROM {name} WHERE {col['name']} = '{quoted}' LIMIT 100",
+                "nl": f"Show {name} where {col_name} is {first_value}",
+                "sql": f"SELECT * FROM {name} WHERE {col_name} = '{quoted}' LIMIT 100",
             }
         )
 
diff --git a/core/wren/tests/unit/test_seed_queries.py b/core/wren/tests/unit/test_seed_queries.py
index 8f56cd71e3..965bf2628a 100644
--- a/core/wren/tests/unit/test_seed_queries.py
+++ b/core/wren/tests/unit/test_seed_queries.py
@@ -792,3 +792,25 @@ def test_relationship_key_case_insensitive_isolated(self):
         sqls = [p["sql"] for p in generate_seed_queries(manifest)]
         assert not any("SUM(CreatedBy)" in s for s in sqls)
         assert "SELECT SUM(word_count) FROM documents" in sqls
+
+
+def test_skips_non_dict_columns():
+    """Partial MDL column arrays must not TypeError the seed generator."""
+    manifest = {
+        "models": [
+            {
+                "name": "orders",
+                "primaryKey": "id",
+                "columns": [
+                    None,
+                    "bad",
+                    {"name": "amount", "type": "int"},
+                    {"type": "varchar"},  # missing name
+                ],
+            }
+        ]
+    }
+    pairs = generate_seed_queries(manifest)
+    sqls = [p["sql"] for p in pairs]
+    assert "SELECT * FROM orders LIMIT 100" in sqls
+    assert "SELECT SUM(amount) FROM orders" in sqls