The idea is:
To be able to create conditions for AI Agent tools that determine whether to use them or not. This would in essence prohibit n8n core from wiring it up into the low level agent prompt, making it unknown to it.
My use cases:
- skip memory store for non-chat, “instruct” routes like email handling
- skip authentication workflow tool when something like an auth flag is already provided in the flow
I think it would be beneficial to add this because:
We now have to duplicate AI Agents and have to manage both, just to have one with and without memory store
Any resources to support this?
Are you willing to work on this?
About to leave for a long roadtrip, hmmmm…don’t think so