How can i get output result and analyse again the question to answer?

Questions
1

Dear,

I have this diagramm:

The chat is linked with the 1st AI Agent : Goal: when the user ask a question , it check is the question is link to Activity, for example : Have you play football ?
If the Agent manage only Football question so i filter if it’s True or false.

If it’s false, the user try again to ask other question and says on prompt it’s not link to Football
If it’s True, the second AGent check if the question can be answer with the Vector storage.

I hope it’s clear.

Thank you for your help.

CAP_1
CAP_11618×672 119 KB

Information on your n8n setup

  • **n8n version:1.86.1 (Cloud free)
  • Database (default: SQLite):
  • n8n EXECUTIONS_PROCESS setting (default: own, main):
  • Running n8n via (Docker, npm, n8n cloud, desktop app):
  • **Operating system:Windows 10
2

So, what is your question?

If you mean that the user is never told that their question is not related to Football, then you just need to disconnect the Set node in the false branch from AI Agent. There is no point to sending this message to the AI Agent.

Whether you keep the Set node disconnected or need to add a Respond to Webhook node, depends on Chat trigger node configuration.

If the above resolves your question, please mark this post as a :white_check_mark: Solution.

3

How can i configure the 1st Agent AI to control the question if it related to “Football” for example (initialize variable to TRUE or FALSE )? If NO, it ask USER to ask again a question. If it related to Football so the 2nd AGENT answer at the question.

4

Assuming that the workflow you have is able to work with an arbitrary activity and given the amount of information you disclosed, the only advice possible is to adjust the AI Agent’s prompt so that it fits your requirement.