Call Log ID: 019e18c3-c1e3-7003-8183-dc4812d5da09 My assistant is calling my tool and I see in the call log that it is triggering but its executing for too long and then the call ends. Its not a complex tool so it shouldnt take so long to process. Also, Im never getting the request on my server so its never hitting my webhook.

This usually points to a timeout or routing issue rather than tool complexity. I can help you fix this. From what you described, the tool is being triggered but the request is likely failing before it reaches your webhook (or getting stuck in the provider layer), which is why the call eventually ends. We should first verify the tool execution timeout settings and confirm whether the request is actually being sent out from the assistant platform. I can help you trace the call flow and pinpoint where it’s dropping so we can stabilize the webhook delivery. @Devin

@Rita Heritage The assistant having problems with this tool is part of a handoff sequence. When i call the assistant by itself it has no problem with the tool

If the tool works when the assistant runs alone but fails during handoff, the issue is almost certainly in the handoff context (not the tool itself). In many cases, the payload or session state passed during handoff changes slightly, which can break tool execution or delay the request until timeout.. I can help you fix this by reviewing what data is being passed during the handoff and ensuring the tool call context stays intact and is not getting truncated or delayed. @Devin

Hi @User, I’m running into an issue where a tool is being triggered in the call log, but the webhook request is never actually reaching my server. The strange part is that the tool works correctly when I call the assistant directly. However, when that same assistant is entered through a handoff sequence, the call log shows the tool being triggered, but it appears to hang/execute for too long until the call eventually ends. I checked my server logs, and no request is received at all, so this does not seem to be an issue with my webhook response time or server processing. I duplicated the destination assistant, reconnected the handoff to the new duplicated assistant, and the tool started working again immediately. That makes it seem like the original assistant had a stale or corrupted runtime/tool configuration specifically when reached through the handoff path. Could your team look into why an assistant can show a tool call as triggered in the logs but never actually dispatch the HTTP request to the configured webhook when used after a handoff? It would also be helpful if the logs clearly showed whether the webhook request was actually dispatched, what server URL was resolved at runtime, and whether Vapi failed before making the outbound request. Right now, the log makes it look like the tool executed, but from the server side, the webhook was never hit. This is difficult to debug because the assistant works when called directly, but fails only when reached through the handoff sequence.

Hi, I’ve investigated call 019e18c3-c1e3-7003-8183-dc4812d5da09. Your analysis is correct that the tool isn't completing, but the issue is actually occurring on the Make.com side.