OpenAI MemoryKey

As many of you might relate, I’ve been using ActivePieces to automate my social media tasks, specifically posting specific content on Twitter weekly. It’s been a fantastic journey so far, but I’ve hit a bit of a snag. Recently, I’ve noticed that OpenAI has been making duplicate posts on my Twitter account, which is not ideal, to say the least.

I’ve heard whispers of a feature called the “Memory Key” that might be the solution to my problem. However, I must admit, I’m a bit lost on how to effectively use this feature. Could anyone kindly shed some light on how this Memory Key works? How can I implement it to prevent these duplicate posts?

Any guidance, tips, or insights would be immensely appreciated.

Thank you in advance for your time and help.


Hello @Mike,

No worries, the name memory key might be a bit scary but it works in a fairly simple way.

You can use memory key in a lot of ways. But I’ll stick to your use case here, to make the GPT remember previous posts you can fill the memory key field with any key, such as twitter-gpt for example, it’s basically a way for the flow to remember where the messages belong.

That way, every new post generated will be saved under twitter-gpt, and can be remembered. If you have other OpenAI actions, you can use the same key to share the memory between them, or create a new key for a new memory.

Also make sure to always check how much context can the model you’re using handle, eventually the memory gets too big and it ‘forgets’ the oldest posts in it over time.

1 Like


I really overcomplicated that. :sweat_smile:

This topic was automatically closed 24 hours after the last reply. New replies are no longer allowed.