Understanding chat memory and variable use

I'm curious if I'm "doing it right" regarding chat memory. I spun up a redis instance in docker for testing and it works successfully but what I'm wondering about is the session ID / key.

If I use it as the pre-defined chat Json and try to connect subsequent agent interactions to it, it errors out as it won't get the initial key generated in the first one. I couldn't figure out how to pass that same key to other agents.

What I've done as a workaround is change it to a pre-defined key and it works but then it's way too static if that makes sense. I'm used to other workflow tools where I could just declare that session ID as a variable to use later, is that not something I can do in n8n?

Thanks for the help (anyone) and

@Nate Herk

for establishing this community!

0 comments