PR #521: address CodeRabbit review — consistent 401 payload; mask key prefix; accurate reranking_applied; remove unused import

python/src/server/api_routes/knowledge_api.py

@@ -33,7 +33,6 @@ from ..services.storage import DocumentStorageService
 from ..utils import get_supabase_client
 from ..services.embeddings.embedding_exceptions import (
     EmbeddingAuthenticationError,
-    EmbeddingQuotaExhaustedError,
 )
 from ..utils.document_processing import extract_text_from_document
 
@@ -756,7 +755,7 @@ async def perform_rag_query(request: RagQueryRequest):
         raise
     except EmbeddingAuthenticationError as e:
         safe_logfire_error(f"Authentication error in RAG query: {str(e)}")
-        raise HTTPException(status_code=401, detail="Invalid API key")
+        raise HTTPException(status_code=401, detail={"error": "Invalid API key"})
     except Exception as e:
         safe_logfire_error(
             f"RAG query failed | error={str(e)} | query={request.query[:50]} | source={request.source}"
@@ -793,7 +792,7 @@ async def search_code_examples(request: RagQueryRequest):
         raise
     except EmbeddingAuthenticationError as e:
         safe_logfire_error(f"Authentication error in code examples search: {str(e)}")
-        raise HTTPException(status_code=401, detail="Invalid API key")
+        raise HTTPException(status_code=401, detail={"error": "Invalid API key"})
     except Exception as e:
         safe_logfire_error(
             f"Code examples search failed | error={str(e)} | query={request.query[:50]} | source={request.source}"

python/src/server/services/embeddings/embedding_exceptions.py

@@ -93,9 +93,10 @@ class EmbeddingAuthenticationError(EmbeddingError):
 
     def __init__(self, message: str, api_key_prefix: str | None = None, **kwargs):
         super().__init__(message, **kwargs)
-        self.api_key_prefix = api_key_prefix
-        if api_key_prefix:
-            self.metadata["api_key_prefix"] = api_key_prefix
+        masked = f"{api_key_prefix[:4]}…" if api_key_prefix else None
+        self.api_key_prefix = masked
+        if masked:
+            self.metadata["api_key_prefix"] = masked
 
 
 class EmbeddingAPIError(EmbeddingError):

python/src/server/services/search/rag_service.py

@@ -339,11 +339,13 @@ class RAGService:
                     )
 
                 # Apply reranking if we have a strategy
+                reranking_applied = False
                 if self.reranking_strategy is not None and results:
                     try:
                         results = await self.reranking_strategy.rerank_results(
                             query, results, content_key="content"
                         )
+                        reranking_applied = True
                     except Exception as e:
                         logger.warning(f"Code reranking failed: {e}")
 
@@ -367,14 +369,14 @@ class RAGService:
                     "query": query,
                     "source_filter": source_id,
                     "search_mode": "hybrid" if use_hybrid_search else "vector",
-                    "reranking_applied": self.reranking_strategy is not None,
+                    "reranking_applied": reranking_applied,
                     "results": formatted_results,
                     "count": len(formatted_results),
                 }
 
                 span.set_attribute("results_found", len(formatted_results))
                 span.set_attribute("hybrid_used", use_hybrid_search)
-                span.set_attribute("reranking_used", self.reranking_strategy is not None)
+                span.set_attribute("reranking_used", reranking_applied)
 
                 return True, response_data