feat(provenance): govern source snapshots and data inputs

backend/tests/test_sprint13_yolo_preflight.py

@@ -1,5 +1,6 @@
 from __future__ import annotations
 
+from hashlib import sha256
 import json
 import subprocess
 import sys
@@ -9,6 +10,7 @@ from fastapi.testclient import TestClient
 
 from app.core.config import Settings
 from app.main import app
+from app.services.runtime_model_provenance_service import RuntimeModelProvenanceService
 from app.services.yolo_preflight_service import YoloPreflightService
 
 
@@ -57,6 +59,43 @@ def _manifest(tmp_path: Path, tile_count: int = 1) -> Path:
     return manifest_path
 
 
+def _write_model_sidecar(model_path: Path, settings: Settings) -> None:
+    model_sha256 = sha256(model_path.read_bytes()).hexdigest()
+    payload = {
+        "schema_version": RuntimeModelProvenanceService.MANIFEST_SCHEMA_VERSION,
+        "data_contract": {"key": "geointel.model.pytorch", "version": "1.0.0"},
+        "model": {
+            "model_id": settings.yolo_model_id,
+            "task_type": "object_detection",
+            "sha256": model_sha256,
+            "model_format": "pytorch",
+            "framework": "ultralytics/pytorch",
+            "class_mapping": {"0": "building"},
+            "source_version": settings.yolo_model_version or "test-v1",
+        },
+        "source": {
+            "source_registry_id": "11111111-1111-4111-8111-111111111111",
+            "source_snapshot_id": "22222222-2222-4222-8222-222222222222",
+            "source_registry_key": "model",
+            "source_snapshot_checksum_sha256": model_sha256,
+        },
+        "lineage": {
+            "upstream_asset_ids": ["test-training-corpus"],
+            "upstream_checksums_sha256": ["a" * 64],
+            "transformations": [
+                {"name": "test-training", "version": "1.0.0", "checksum_sha256": "b" * 64}
+            ],
+        },
+        "metadata": {"training_manifest_sha256": "c" * 64},
+        "imported_at": "2026-08-01T10:00:00+00:00",
+    }
+    payload["metadata"]["runtime_manifest_sha256"] = RuntimeModelProvenanceService.manifest_self_checksum(payload)
+    RuntimeModelProvenanceService.manifest_path_for_model(model_path).write_text(
+        json.dumps(payload, sort_keys=True),
+        encoding="utf-8",
+    )
+
+
 def test_yolo_preflight_reports_disabled_without_loading_model(tmp_path: Path, monkeypatch) -> None:
     monkeypatch.setenv("YOLO_CONFIG_DIR", str(tmp_path / "ultralytics"))
 
@@ -99,9 +138,11 @@ def test_yolo_preflight_validates_model_and_manifest_without_importing_yolo(tmp_
     model_path = tmp_path / "model.pt"
     model_path.write_bytes(b"weights")
     manifest_path = _manifest(tmp_path, tile_count=2)
+    settings = Settings(yolo_enabled=True, yolo_model_path=str(model_path), yolo_max_tiles=4)
+    _write_model_sidecar(model_path, settings)
 
     result = YoloPreflightService.run(
-        settings=Settings(yolo_enabled=True, yolo_model_path=str(model_path), yolo_max_tiles=4),
+        settings=settings,
         tile_manifest_path=str(manifest_path),
         yolo_adapter_class=AvailableAdapter,
     )
@@ -109,6 +150,7 @@ def test_yolo_preflight_validates_model_and_manifest_without_importing_yolo(tmp_
     assert result["status"] == "ready"
     assert result["checks"]["dependencies_available"] is True
     assert result["checks"]["model_file_exists"] is True
+    assert result["checks"]["model_provenance_valid"] is True
     assert result["checks"]["manifest_valid"] is True
     assert result["tile_count"] == 2
     assert result["will_download_models"] is False
@@ -120,9 +162,11 @@ def test_yolo_preflight_marks_assumed_dependencies_in_runtime_details(tmp_path:
     model_path = tmp_path / "model.pt"
     model_path.write_bytes(b"weights")
     manifest_path = _manifest(tmp_path, tile_count=1)
+    settings = Settings(yolo_enabled=True, yolo_model_path=str(model_path), yolo_max_tiles=4)
+    _write_model_sidecar(model_path, settings)
 
     result = YoloPreflightService.run(
-        settings=Settings(yolo_enabled=True, yolo_model_path=str(model_path), yolo_max_tiles=4),
+        settings=settings,
         tile_manifest_path=str(manifest_path),
         yolo_adapter_class=MissingDependencyAdapter,
         assume_dependencies=True,
@@ -138,9 +182,11 @@ def test_yolo_preflight_can_explicitly_smoke_load_local_model(tmp_path: Path) ->
     model_path = tmp_path / "model.pt"
     model_path.write_bytes(b"weights")
     manifest_path = _manifest(tmp_path)
+    settings = Settings(yolo_enabled=True, yolo_model_path=str(model_path), yolo_max_tiles=4)
+    _write_model_sidecar(model_path, settings)
 
     result = YoloPreflightService.run(
-        settings=Settings(yolo_enabled=True, yolo_model_path=str(model_path), yolo_max_tiles=4),
+        settings=settings,
         tile_manifest_path=str(manifest_path),
         yolo_adapter_class=AvailableAdapter,
         check_model_load=True,
@@ -156,9 +202,11 @@ def test_yolo_preflight_can_explicitly_smoke_load_local_model(tmp_path: Path) ->
 def test_yolo_preflight_reports_explicit_model_load_failure(tmp_path: Path) -> None:
     model_path = tmp_path / "model.pt"
     model_path.write_bytes(b"weights")
+    settings = Settings(yolo_enabled=True, yolo_model_path=str(model_path), yolo_max_tiles=4)
+    _write_model_sidecar(model_path, settings)
 
     result = YoloPreflightService.run(
-        settings=Settings(yolo_enabled=True, yolo_model_path=str(model_path), yolo_max_tiles=4),
+        settings=settings,
         tile_manifest_path=str(_manifest(tmp_path)),
         yolo_adapter_class=FailingLoadAdapter,
         check_model_load=True,
@@ -173,6 +221,7 @@ def test_yolo_preflight_script_outputs_json(tmp_path: Path) -> None:
     model_path = tmp_path / "model.pt"
     model_path.write_bytes(b"weights")
     manifest_path = _manifest(tmp_path)
+    _write_model_sidecar(model_path, Settings(yolo_enabled=True, yolo_model_path=str(model_path)))
 
     result = subprocess.run(
         [
@@ -201,6 +250,7 @@ def test_yolo_preflight_script_uses_environment_configuration(tmp_path: Path, mo
     model_path = tmp_path / "model.pt"
     model_path.write_bytes(b"weights")
     manifest_path = _manifest(tmp_path)
+    _write_model_sidecar(model_path, Settings(yolo_enabled=True, yolo_model_path=str(model_path), yolo_max_tiles=4))
     monkeypatch.setenv("YOLO_ENABLED", "true")
     monkeypatch.setenv("YOLO_MODEL_PATH", str(model_path))
     monkeypatch.setenv("YOLO_MAX_TILES", "4")
@@ -227,6 +277,21 @@ def test_yolo_preflight_script_uses_environment_configuration(tmp_path: Path, mo
     assert payload["max_tiles"] == 4
 
 
+def test_yolo_preflight_refuses_unmanifested_local_weights(tmp_path: Path) -> None:
+    model_path = tmp_path / "model.pt"
+    model_path.write_bytes(b"unmanifested weights")
+
+    result = YoloPreflightService.run(
+        settings=Settings(yolo_enabled=True, yolo_model_path=str(model_path)),
+        tile_manifest_path=str(_manifest(tmp_path)),
+        yolo_adapter_class=AvailableAdapter,
+    )
+
+    assert result["status"] == "contract_incomplete"
+    assert result["checks"]["model_provenance_valid"] is False
+    assert result["error_code"] == "MODEL_PROVENANCE_MANIFEST_MISSING"
+
+
 def test_yolo_preflight_script_rejects_assumed_dependencies_for_model_load(tmp_path: Path) -> None:
     result = subprocess.run(
         [