Friends, I want to create a knowledge base scoring workflow. When the LLM node connects to two knowledge bases simultaneously, the context can only select results from one knowledge base. What should I do? I’d like to specify in the prompt: Knowledge Base A: resultA, Knowledge Base B: resultB, please score them.
Related topics
| Topic | Replies | Views | Activity | |
|---|---|---|---|---|
| Chatflow知识检索步骤中,无法检索到结果 | 0 | 33 | January 27, 2026 | |
| 建立的知识库chatflow,知识检索在测试和运行的时候,没有任何反馈,直接提供空 | 7 | 68 | January 22, 2026 | |
| LLM和Agent节点的上下文变量怎么设置,为什么是空的 | 1 | 117 | January 12, 2026 | |
| 如何将用户上传的文件放入知识库? | 0 | 42 | January 23, 2026 | |
| 多路召回时影响到了每一路的召回。这不合理 | 0 | 25 | January 23, 2026 | |
| Llm节点如何支持多模态输入? | 6 | 224 | January 16, 2026 | |
| How to include metadata, from knowledge retrieval, seen by the llm? | 2 | 83 | January 12, 2026 | |
| 如何dify配置集成openmetadata的工作流, 以便LLM基于openmetadata里的元数据给出精准回答 | 0 | 24 | January 26, 2026 | |
| 大模型的上下文不能接收Array[Object]类型吗? | 1 | 28 | February 4, 2026 | |
| 我生成了两份结果,但是显示混合到了一起 | 2 | 17 | January 15, 2026 |