Agents react to my own message blocks
I am using cards with buttons. They're connected to message blocks and then lead to my main agent.
My main agent often treats the context from these message blocks as if it is a user input and sometimes loops this input with further responses.
How can I avoid that my agent reacts to my own messages in the workflow as if it is a user input?
My main agent often treats the context from these message blocks as if it is a user input and sometimes loops this input with further responses.
How can I avoid that my agent reacts to my own messages in the workflow as if it is a user input?
