I'm looking to test out the benefits of transient assistant for us for our inbound agents/assistants. When I use make.com and return a webhook response that's simple, e.g.: { "assistant": { "firstMessage": "Thanks for calling XYZ Company, this is Casey, what can I do for you?", "model": { "provider": "openai", "model": "gpt-4o", "messages": [ { "role": "system", "content": "You're a helpful assistant." } ] } } } It works. But the configurations of our actual performing agents are much more complex. E.g. the system prompt is ~6k words. And testing even without the system prompt, with all the other configurations that I build via swagger via Create Call api call, it still ends with the same error. Please advise. I'm trying to pinpoint exactly what the challenge and having an understanding of the expected response will be helpful. Could you guys clarify please?

Hey george.l when you send wrong response we send debugging artifacts to your server can you try looking into your server logs? FYI: I cannot check debugging artifacts.

Was able to resolve it, thank you for your help