#@Angelo @Anashel even for a agent with
1 messages · Page 1 of 1 (latest)
It results in the following
CALL TRANSCRIPT
Agent: Hello! How can I help you today?
User: What is the price of a cat?
Workflow transitioned to Workflow Test price subagent
Agent: It's 50 Euro.
User: What are your opening hours on Monday?
Workflow transitioned to Workflow Test Main
Workflow transitioned to Workflow Test opening hours subagent
Agent: The opening hours are 8 a.m. to 5 p.m.
User: What is the price of a dog?
Workflow transitioned to Workflow Test Main
Workflow transitioned to Workflow Test opening hours subagent
Agent: It's 50 Euro.
User: What are your opening hours on Tuesday?
Workflow transitioned to Workflow Test Main
Agent: The opening hours are 8 a.m. to 5 p.m.
User: What is the price of a cow?
Workflow transitioned to Workflow Test price subagent
Agent: It's 50 Euro.
User: What are your opening hours on Wednesday?
Agent: The opening hours are 8 a.m. to 5 p.m.
User: What is the price of a lion?
Agent: It's 50 Euro.
User: What are your opening hours on Thursday?
Agent: The opening hours are 8 a.m. to 5 p.m.
Is this just for testing usage of workflows? Since you could easily handle this in a single agent? - I am curious about when to single agent and when to "workflow" though
@crisp orbit yes, this was just for the purpose to test whether workflows are working. This is why this was the most easiest task I came up with, and it never worked out properly. So I have a hard time convincing myself to go all in on that (workflow style agent) with a production agent.
Which LLM are you using?
What I am understanding is that your subagent completes the task related to it but if user asks something which is not related to the subagent, your task does not get routed back to the main agent and call ends. Am I correct?
I dont know if you saw but all workflow are now in the agent details object with all nested node logic. So you can now update, backup and change entirely via the API.