feat(provenance): govern source snapshots and data inputs

scripts/prepare_operator_real_data_samples.py

@@ -16,6 +16,12 @@ from dataclasses import dataclass, replace
 from pathlib import Path
 from typing import Any
 
+SCRIPT_DIR = Path(__file__).resolve().parent
+if str(SCRIPT_DIR) not in sys.path:
+    sys.path.insert(0, str(SCRIPT_DIR))
+
+from training_dataset_eligibility import TRAINING_ELIGIBILITY_POLICY_VERSION  # noqa: E402
+
 
 WMS_URL = "https://geo.api.vlaanderen.be/omwrgbmrvl/wms"
 GRB_GBG_URL = "https://geo.api.vlaanderen.be/GRB/ogc/features/v1/collections/GBG/items"
@@ -803,6 +809,14 @@ def main() -> int:
     manifest = {
         "schema_version": 2,
         "description": "GeoIntel operator real-data samples for configured-YOLO QA validation.",
+        "training_eligibility": {
+            "policy_version": TRAINING_ELIGIBILITY_POLICY_VERSION,
+            "status": "not_eligible",
+            "reason": (
+                "Direct provider downloads are QA-only until re-ingested through the governed "
+                "dataset source registry, contract validator and provenance snapshot flow."
+            ),
+        },
         "output_dir": str(output_dir),
         "sample_width": args.width,
         "sample_height": args.height,