Any strategies on how to pass the output of one tool as the input to another tool? is it possible to control or tune the query generation for the tools?
@James Walls that's what I ended up doing. I manually orchestrated agents (as tools). This gave me more control over sequence and what each agent received as a prompt. When using the AI Agent (Tools Agent) the prompts it generates for the tools are not the best. I need to dive deeper into LangChain to understand what's happening under N8N