Show retrieved knowledge chunks in audit logs

ai_knowledge_base_v2/apps/backend/app/api/admin_records.py

@@ -1,8 +1,10 @@
 from __future__ import annotations
 
 import csv
+import json
 from datetime import datetime
 from io import StringIO
+from typing import Any
 
 from fastapi import APIRouter, Depends, Query
 from fastapi.responses import Response
@@ -236,6 +238,7 @@ def _ai_log_dict(log: AiRequestLog, *, include_prompt: bool = False) -> dict:
         "modelName": log.model_name,
         "knowledgeIds": log.knowledge_ids,
         "retrieveCount": log.retrieve_count,
+        "retrievedChunks": _parse_retrieved_chunks(log.retrieved_chunks),
         "inputToken": log.input_token,
         "outputToken": log.output_token,
         "totalToken": log.total_token,
@@ -247,3 +250,29 @@ def _ai_log_dict(log: AiRequestLog, *, include_prompt: bool = False) -> dict:
     if include_prompt:
         data["prompt"] = log.prompt
     return data
+
+
+def _parse_retrieved_chunks(raw: str | None) -> list[dict[str, Any]]:
+    if not raw:
+        return []
+    try:
+        chunks = json.loads(raw)
+    except json.JSONDecodeError:
+        return []
+    if not isinstance(chunks, list):
+        return []
+    normalized: list[dict[str, Any]] = []
+    for index, item in enumerate(chunks, start=1):
+        if not isinstance(item, dict):
+            continue
+        normalized.append(
+            {
+                "index": item.get("index") or index,
+                "knowledgeId": item.get("knowledgeId"),
+                "knowledgeName": item.get("knowledgeName") or "",
+                "title": item.get("title") or "",
+                "content": item.get("content") or "",
+                "sourceUrl": item.get("sourceUrl") or None,
+            }
+        )
+    return normalized