diff --git a/docs/config-schema.md b/docs/config-schema.md
index 214b9cd..2b88536 100644
--- a/docs/config-schema.md
+++ b/docs/config-schema.md
@@ -27,6 +27,12 @@ I path relativi sono sempre risolti rispetto alla directory che contiene `datase
 | `dataset.years` | `list[int]` | nessuno |
 | `dataset.tags` | `list[string]` | `[]` |
 | `dataset.category` | `string \| null` | `null` |
+| `dataset.time_coverage` | `{start_year: int, end_year: int} \| null` | `null` |
+
+`time_coverage` dichiara l'intervallo temporale del contenuto (non degli anni di
+run). Esposto dal `dataset_loader` (manifest) e dal layer tool per la
+catalogazione; usato dai candidate per serie storiche (es. `ga-sentenze`:
+2023-2026). Opzionale ma raccomandato per serie storiche.
 
 ## raw
 
diff --git a/tests/test_dataset_loader.py b/tests/test_dataset_loader.py
index ffa0122..18d2c95 100644
--- a/tests/test_dataset_loader.py
+++ b/tests/test_dataset_loader.py
@@ -284,6 +284,37 @@ def test_warns_missing_source_id(self, tmp_path: Path) -> None:
         assert result["ok"] is True
         assert any("source_id" in w for w in result["warnings"])
 
+    def test_warns_missing_tags_and_category(self, tmp_path: Path) -> None:
+        _write_dataset(
+            tmp_path,
+            {
+                "dataset": {"name": "test", "years": [2023], "source_id": "openga"},
+                "raw": {"sources": [{"name": "src1"}]},
+            },
+        )
+        result = validate_config(tmp_path)
+        assert result["ok"] is True
+        assert any("tags" in w for w in result["warnings"])
+        assert any("category" in w for w in result["warnings"])
+
+    def test_no_warn_when_tags_category_present(self, tmp_path: Path) -> None:
+        _write_dataset(
+            tmp_path,
+            {
+                "dataset": {
+                    "name": "test",
+                    "years": [2023],
+                    "source_id": "openga",
+                    "tags": ["giustizia"],
+                    "category": "giustizia",
+                },
+                "raw": {"sources": [{"name": "src1"}]},
+            },
+        )
+        result = validate_config(tmp_path)
+        assert result["ok"] is True
+        assert not any("tags" in w or "category" in w for w in result["warnings"])
+
     def test_invalid_yaml(self, tmp_path: Path) -> None:
         (tmp_path / "dataset.yml").write_text("invalid: [")
         result = validate_config(tmp_path)
diff --git a/toolkit/core/dataset_loader.py b/toolkit/core/dataset_loader.py
index 342fb14..dd5bbea 100644
--- a/toolkit/core/dataset_loader.py
+++ b/toolkit/core/dataset_loader.py
@@ -138,6 +138,15 @@ def validate_config(path: str | Path) -> dict[str, Any]:
     if sources and not manifest.get("source_id"):
         warnings.append("'dataset.source_id' non impostato (utile per catalogazione)")
 
+    if not manifest.get("tags"):
+        warnings.append(
+            "'dataset.tags' non impostato (standard candidate: lista kebab-case descrittiva)"
+        )
+    if not manifest.get("category"):
+        warnings.append(
+            "'dataset.category' non impostato (standard candidate: tema — vedi docs/candidate-standard.md)"
+        )
+
     return {
         "ok": len(errors) == 0,
         "errors": errors,