Build Agentic workflows, RAG pipelines, with rich AI model and tool support on one collaborative workspace. Deploy on cloud, VPC, or self-hosted, so teams move from prototype to production without rebuilding the stack.
Python agentagentic-aiagentic-frameworkagentic-workflowaiautomationclaudedeepseekgenaigptllmlow-codemcpnextjsno-codeopenaipythonskillsworkflow
fix(api): honor explicit zero top_k in dataset retrieval paths (#42980)
yetuge7904c8acb285
diff --git a/api/core/rag/retrieval/dataset_retrieval.py b/api/core/rag/retrieval/dataset_retrieval.pyindex ee4a4c7284d..6b266800c16 100644--- a/api/core/rag/retrieval/dataset_retrieval.py+++ b/api/core/rag/retrieval/dataset_retrieval.py@@ -1207,12 +1207,16 @@ class DatasetRetrieval:all_documents.extend(documents)else:if top_k > 0:+ # top_k=0 is a legal explicit value (the SINGLE-retrieval+ # tool path reads the stored top_k as-is); only a missing+ # entry falls back to the default.+ dataset_top_k = retrieval_model.get("top_k")# retrieval sourcedocuments = RetrievalService.retrieve(retrieval_method=retrieval_model["search_method"],dataset_id=dataset.id,query=query,- top_k=retrieval_model.get("top_k") or 4,+ top_k=dataset_top_k if dataset_top_k is not None else 4,score_threshold=retrieval_model.get("score_threshold", 0.0)if retrieval_model["score_threshold_enabled"]else 0.0,@@ -1397,7 +1401,9 @@ class DatasetRetrieval:tool = DatasetMultiRetrieverTool.from_dataset(dataset_ids=[dataset.id for dataset in available_datasets],tenant_id=tenant_id,- top_k=retrieve_config.top_k or 4,+ # top_k=0 is a legal explicit value (the non-tool MULTIPLE path+ # honors it); only None falls back to the default.+ top_k=retrieve_config.top_k if retrieve_config.top_k is not None else 4,score_threshold=retrieve_config.score_threshold,hit_callbacks=[hit_callback],return_resource=return_resource,diff --git a/api/core/tools/utils/dataset_retriever/dataset_multi_retriever_tool.py b/api/core/tools/utils/dataset_retriever/dataset_multi_retriever_tool.pyindex 909381f8804..693eaaeed48 100644--- a/api/core/tools/utils/dataset_retriever/dataset_multi_retriever_tool.py+++ b/api/core/tools/utils/dataset_retriever/dataset_multi_retriever_tool.py@@ -172,6 +172,7 @@ class DatasetMultiRetrieverTool(DatasetRetrieverBaseTool):# get retrieval model , if the model is not setting , using defaultretrieval_model = dataset.retrieval_model or default_retrieval_model+ dataset_top_k = retrieval_model.get("top_k")if dataset.indexing_technique == IndexTechniqueType.ECONOMY:# use keyword table query@@ -179,7 +180,8 @@ class DatasetMultiRetrieverTool(DatasetRetrieverBaseTool):retrieval_method=RetrievalMethod.KEYWORD_SEARCH,dataset_id=dataset.id,query=query,- top_k=retrieval_model.get("top_k") or 4,+ # top_k=0 is a legal explicit value; only None falls back to 4.+ top_k=dataset_top_k if dataset_top_k is not None else 4,)if documents:all_documents.extend(documents)@@ -190,7 +192,8 @@ class DatasetMultiRetrieverTool(DatasetRetrieverBaseTool):retrieval_method=retrieval_model["search_method"],dataset_id=dataset.id,query=query,- top_k=retrieval_model.get("top_k") or 4,+ # top_k=0 is a legal explicit value; only None falls back to 4.+ top_k=dataset_top_k if dataset_top_k is not None else 4,score_threshold=retrieval_model.get("score_threshold", 0.0)if retrieval_model["score_threshold_enabled"]else 0.0,