feat: extend operational investigation workflows

app/infra/integrations/operational_data.py

@@ -5,6 +5,7 @@ import hashlib
 import json
 from datetime import datetime, timezone
 from pathlib import Path
+from typing import Literal
 
 import httpx
 from pydantic import Field
@@ -33,6 +34,23 @@ class TaskDataError(Exception):
         self.detail = detail
 
 
+def peripheral_records_origin(items: object) -> Literal["system_data", "synthetic"]:
+    """Classify a record collection; mixed collections retain per-item provenance."""
+    if not isinstance(items, list) or any(
+        not isinstance(item, dict) or item.get("origin") not in {"system_data", "synthetic"}
+        for item in items
+    ):
+        raise TaskDataError(502, "外围记录缺少明确来源，不能作为已核实证据")
+    return "synthetic" if items and all(item["origin"] == "synthetic" for item in items) else "system_data"
+
+
+def normalize_peripheral_work_orders(items: object) -> list[dict[str, object]]:
+    """Expose peripheral work orders only as project-owned system data."""
+    if not isinstance(items, list) or any(not isinstance(item, dict) for item in items):
+        raise TaskDataError(502, "外围工单格式无效，不能作为已核实证据")
+    return [{**item, "origin": "system_data"} for item in items]
+
+
 def _matches_project(value: object, project_id: str) -> bool:
     if isinstance(value, list):
         return all(_matches_project(item, project_id) for item in value)
@@ -147,7 +165,14 @@ async def read_task_data(
         data = _select_demo_observations(data, query, str(context.project_id))
     elif not _matches_project(data, str(context.project_id)):
         raise TaskDataError(502, "数据源返回了其他项目的数据")
-    origin = "synthetic" if peripheral else "system_data"
+    origin = "system_data"
+    if query.source == "peripheral_work_orders":
+        items = data.get("items") if isinstance(data, dict) else None
+        data = {**data, "items": normalize_peripheral_work_orders(items)}
+    elif query.source == "peripheral_test_data":
+        items = data.get("items") if isinstance(data, dict) else None
+        origin = peripheral_records_origin(items)
+        data = {**data, "provenance_scope": "per_record", "source_origins": sorted({item["origin"] for item in items})}
     historical_replay = False
     if query.source in {"peripheral_points", "peripheral_observations"}:
         source_origin = data.get("origin") if isinstance(data, dict) else None