I’ve got a scenario where I’m sending a sequence of prompts to Claude (Anthropic), so I’ve got a bunch of Claude modules in a row (see picture below). How I do add a conversation thread ID so that Claude knows these prompts are all part of the same conversation?
In the ChatGPT module, this is easy because there’s a field for it (see picture below), but the Claude module doesn’t have that. And I really want to use Claude because I like the output better than ChatGPT.
I want to have a thread ID in the Claude modules so that each prompt I send to Claude is going into the same conversation. My prompt sequence is LOOOONG with a lot of prompts and a lot of outputs from Claude, so I’m hoping I don’t have to manually map in previous responses or something like that.
Claude (Anthropic) does not have Threads, as such, you will need to pass in an array of previous prompts if necessary.
samliew – request private consultation
Join the Make Fans Discord server to chat with other makers!
1 Like
This is not true. In Zapier it’s called Memory Key - I’m also looking for this possibility in Make.com since I’m desperate to switch over from Zapier to Make.
Welcome to the Make community!
The equivalent to a memory key in Make, is a Data Store.
For more information, see:
I have successfully implemented the data store as memory key for AI threads in many occasions.
Hope this helps! Let me know if there are any further questions or issues.
Note: I see hundreds of posts, notifications, and messages daily on this forum, so if I missed your reply, please message me to look at your reply.
— @samliew
P.S.: Did you know, the concepts of about 70% of questions asked on this forum are already covered in the Make Academy. Investing some effort into it will save you lots of time and frustration using Make later!
@samliew Thanks for your reply! I’ve never used an array in Make before. I looked for any module that says “array” and I can’t find one. Which module do you use for the array?
@samliew I couldn’t figure out the array thing. I found a solution but it really sucks to set up. I’m still looking for a better way.
Here it is step by step:
Set the first prompt:
Use an HTTP module to connect to Claude:
Parse the JSON response from Claude:
Delete the line breaks in Claude’s response so that it can be used for the next HTTP module:
Set prompt #2:
Use an HTTP module to connect to Claude and send Prompt #1, Response #1 and Prompt #2:
And then keep repeating those steps. In each new HTTP module, include all of the previous prompts and responses, plus the new prompt.
It works, but it’s so tedious to set up.
