Skip to main content

I have set up the ChatGPT integration with messenger and it works well in terms of running conversation. In the Assistant Instructions I have put a detailed training prompt which in my understanding would be used as basis for the conversation. 

The problem is that the answers I am getting from GPT are not aligned witht he guideline I have provided. 

I have changed the training description 15 times but it still turns back to a specific option which I have clearly prohibited to even mention. 

Hi @somate!

Have you tried the prompt in ChatGPT to test how well it works? If I want to set up a Zap using Open AI or ChatGPT, I spend time working directly in ChatGPT to craft the prompt before I put it into a Zap. 

Sometimes asking chatgpt what you prompt to use to achieve what you need can get you closer to your desired end result. 

I hope that helps!


Reply