Can any expert answer:
It can be found when viewing the running results in LLM, but when assigning a variable, it always shows that the input is empty and cannot be found. And both the assigned variable and the session variable are of the same type.
Can any expert answer:
It can be found when viewing the running results in LLM, but when assigning a variable, it always shows that the input is empty and cannot be found. And both the assigned variable and the session variable are of the same type.
For example: LLM-1.structured_output.risk_level(string), session variable risk_leve (string)
| Topic | Replies | Views | Activity | |
|---|---|---|---|---|
| LLM和Agent节点的上下文变量怎么设置,为什么是空的 | 1 | 197 | January 12, 2026 | |
| Gemini 模型插件报错:PluginInvokeError | 7 | 299 | January 21, 2026 | |
| 30分钟快速入门示例中的platform_name无法显示 | 1 | 37 | March 1, 2026 | |
| Dify对话运行时报错,显示 运行失败 | 1 | 30 | January 26, 2026 | |
| Dify 1.11.2配置gemini和gpt的模型后工作流报错 | 6 | 98 | January 16, 2026 | |
| [bug]工作流设置图片上传是非必选的,但是报错 | 7 | 187 | December 19, 2025 | |
| 在“代码执行”节点中无法单独关联 input_file, rules_file, stock_file,并怀疑与 userinput.files 有关。请求提供正确映射变量或处理遗留字段的官方方法。 | 7 | 76 | January 26, 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 | 213 | January 22, 2026 | |
| 迭代中有3个失败,导致工作流不往下面跑了 | 0 | 46 | January 5, 2026 | |
| 大模型的上下文不能接收Array[Object]类型吗? | 1 | 54 | February 4, 2026 |
