Hey Vapi team, I’m running into an issue with my assistant for scheduling appointments. I’ve got it connected to Make.com through a webhook. In chat everything works fine — the assistant collects details, recaps them, and only triggers the webhook after the patient confirms. But on voice calls, the webhook fires almost immediately (like after the first word), before any confirmation happens. I’ve already set the tools to strict, added a Confirm flag in the schema, and updated the prompt so it only calls after confirmation, but it still fires too early. Is there a way to make sure the webhook to Make.com only triggers once the scheduling details are fully confirmed on a voice call? Thanks! https://cdn.discordapp.com/attachments/1413247038623191101/1413247038937759745/image.png?ex=68bb3c5e&is=68b9eade&hm=1599375bf60823576916d56b7b0f9930ed594684cccd67193143bafa010a6069& https://cdn.discordapp.com/attachments/1413247038623191101/1413247039604523128/image.png?ex=68bb3c5f&is=68b9eadf&hm=9805b66dcd74c8c29d1a4b718a42136cf74d2fbe332b03dbe87db3a110398f6b&

Hey @VAPIUSER999 what is the name of your function? This sounds silly, but have you provided it with a good description + proper insutrctions on how to use it in the system prompt?

Hi VAPIUSER999, If you're experiencing issues with your voice assistant scheduling webhooks firing too early using Make.com, it may be related to the workflow configuration or timing settings in your assistant's setup. Here are some steps to help troubleshoot and potentially resolve the issue: 1. **Verify Webhook Timing Configuration**: Check your assistant's configuration to ensure the timing for triggering webhooks is set appropriately. If there's a delay parameter, adjust it to align with your expected timing. 2. **Review Assistant Hooks**: Utilize assistant hooks to manage specific events that should trigger webhooks. You can include conditions to control when hooks trigger, ensuring they only execute at the right time. Detailed information on assistant hooks can be found [here](https://docs.vapi.ai/assistants/assistant-hooks). 3. **Log Execution Flows**: Use logging to track when and why a webhook is triggered. This could help you identify any unexpected early firings. 4. **Test in a Controlled Environment**: Utilize the Vapi dashboard to run tests and check how individual components of your assistant function, helping you isolate the issue without external variables interfering. By verifying these components, you should be able to narrow down the cause and adjust the timing accordingly. If you need detailed assistance on any specific part of this process, feel free to ask.