[Feature]AI 对话节点的上下文输出变量{history}

[lulanghero]

MaxKB Version

1.10.1

Please describe your needs or suggestions for improvements

能否将每个AI模型在问答对话中产生的结果作为一个单独的输出的历史记录呢，作为一个{AI.history} 以及向量库检索可以按照分段标题进行检索吗，当我们处理了标准知识库，标题将直接对应分段，根据标题进行向量检索将更好的提高检索结果的效果

Please describe the solution you suggest

No response

Additional Information

No response

[baixin513]

感谢反馈，1、AI 对话节点的上下文可以考虑一下输出变量； 2、分段标题和内容都在同一表中，只检索分段标题和检索整个分段效率是一样的。现在只要命中标题就会召回分段，不会有影响，暂不考虑。

[shaohuzhang1]

Bot detected the issue body's language is not English, translate it automatically. 👯👭🏻🧑‍🤝‍🧑👫🧑🏿‍🤝‍🧑🏻👩🏾‍🤝‍👨🏿👬🏿

---

Thanks for feedback, 1. The context of the AI ​​dialogue node can consider the output variables; 2. The segment title and content are in the same table. Only searching the segment title is the same as searching the entire segment. Now, as long as the title is hit, the segment will be recalled, and there will be no impact, so I will not consider it for the time being.

[baixin513]

请更新至v2.3.0，已支持该功能。

[shaohuzhang1]

Bot detected the issue body's language is not English, translate it automatically. 👯👭🏻🧑‍🤝‍🧑👫🧑🏿‍🤝‍🧑🏻👩🏾‍🤝‍👨🏿👬🏿

---

Please update to v2.3.0, this feature is already supported.