feat(provenance): govern source snapshots and data inputs

scripts/rotate_belgium_building_holdouts.py

@@ -7,6 +7,7 @@ import argparse
 import hashlib
 import json
 import shutil
+import sys
 from collections import Counter
 from pathlib import Path
 from typing import Any
@@ -15,6 +16,19 @@ from pyproj import Transformer
 from shapely.geometry import box
 from shapely.ops import transform as shapely_transform
 
+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 (  # noqa: E402
+    TrainingEligibilityError,
+    assert_frozen_manifest_training_eligible,
+)
+from training_release_manifest import (  # noqa: E402
+    TrainingReleaseError,
+    assert_yolo_summary_bound_to_embedded_training_release,
+)
+
 
 def sha256(path: Path) -> str:
     digest = hashlib.sha256()
@@ -73,6 +87,11 @@ def main() -> int:
     parser.add_argument("--internal-val-samples", required=True)
     parser.add_argument("--version", required=True)
     parser.add_argument("--output-dir", type=Path, required=True)
+    parser.add_argument(
+        "--fixture-mode",
+        action="store_true",
+        help="Accept only an explicitly fixture-only source corpus; never use for operational holdout rotation.",
+    )
     args = parser.parse_args()
 
     role_slugs = {
@@ -86,6 +105,20 @@ def main() -> int:
         raise SystemExit("rotated holdout lists overlap")
 
     manifest = json.loads(args.corpus_manifest.read_text(encoding="utf-8"))
+    try:
+        assert_frozen_manifest_training_eligible(
+            args.corpus_manifest,
+            fixture_mode=args.fixture_mode,
+            verify_live=True,
+        )
+        for path in args.source_summary:
+            assert_yolo_summary_bound_to_embedded_training_release(
+                summary_path=path,
+                corpus_manifest=args.corpus_manifest,
+                fixture_mode=args.fixture_mode,
+            )
+    except (TrainingEligibilityError, TrainingReleaseError) as exc:
+        raise SystemExit(str(exc)) from exc
     source_samples = {item["sample_slug"]: item for item in manifest["samples"]}
     missing = sorted(all_holdouts - source_samples.keys())
     if missing:
@@ -126,6 +159,18 @@ def main() -> int:
     args.output_dir.mkdir(parents=True, exist_ok=True)
     manifest_path = args.output_dir / "operator_samples_manifest.json"
     write_json(manifest_path, rotated_manifest)
+    write_json(
+        args.output_dir / "corpus-freeze.json",
+        {
+            "schema_version": 2,
+            "dataset_version": rotated_manifest["dataset_version"],
+            "manifest_sha256": sha256(manifest_path),
+            "sample_count": len(rotated_manifest["samples"]),
+            "immutable": True,
+            "training_eligibility_policy": rotated_manifest["training_eligibility"]["policy_version"],
+            "fixture_mode": bool(args.fixture_mode),
+        },
+    )
     leakage = audit_spatial_leakage(rotated_manifest["samples"])
     write_json(args.output_dir / "spatial-leakage-audit.json", leakage)
     if leakage["status"] != "ok":
@@ -193,6 +238,11 @@ def main() -> int:
         "fit_samples_in_internal_validation": sorted(
             {tile["sample_slug"] for tile in train_tiles} & {tile["sample_slug"] for tile in internal_val_tiles}
         ),
+        "fixture_mode": bool(args.fixture_mode),
+        "training_eligible": False,
+        "training_eligibility_reason": (
+            "A rotated split needs a new human review, label contract validation and immutable training release."
+        ),
     }
     if evidence["protected_samples_in_training"]:
         raise SystemExit("protected samples leaked into rotated training lists")