tool calling is not working
# support
m
looks like the make.com scenario is correctly running and returning results but the AI assistant keeps saying "Unable to fetch" call id: 717fc86b-c536-40cd-8f23-35556003679a tool id: 6ff5f455-649b-4679-89d5-7e7eef8e1788 org id: 02294105-0630-44fa-8367-52d7f212e4b2 https://cdn.discordapp.com/attachments/1385705147484864724/1385705147790921748/image.png?ex=685709fc&is=6855b87c&hm=413a325b09183780846cc44343e5085048363173a8401b3ce156e78e9ebd6dbe&
a
Copy code
{
        "role": "tool_calls",
        "time": 1750418227144,
        "message": "",
        "toolCalls": [
          {
            "id": "call_6SDpBAGUd8pWGfQ2wzTxlIyX",
            "type": "function",
            "function": {"name": "questions_to_ask", "arguments": "{}"}
          }
        ],
        "secondsFromStart": 10.471
      },
      {
        "role": "bot",
        "time": 1750418228049.999,
        "source": "",
        "endTime": 1750418228690,
        "message": "Just a sec.",
        "duration": 640.0009765625,
        "secondsFromStart": 11.639999
      },
      {
        "name": "questions_to_ask",
        "role": "tool_call_result",
        "time": 1750418228158,
        "result": "No result returned.",
        "toolCallId": "call_6SDpBAGUd8pWGfQ2wzTxlIyX",
        "secondsFromStart": 11.485
      },
From the logs, it is either not formatted correctly or requires parameters to receive the values. Make sure when you return the values that there are no line breaks in the result so the LLM can process it correctly.
h
Hi Kyle, i've got this exact problem too, except my tool call result returns "Success" instead of the actual result (free slots on calendar)
v
i found that if i turn async off it works properly
a
While I wait for @mj to respond, @hj.ong9 could you please provide a call id? Also please try turning async off like @vahidkowsari suggested.
m
increased the timeout from 1 second to 10 seconds. this fixed the issue
a
Glad the issue was resolved. Please reach out to us if you have any other questions regarding this issue
v
Is there somewhere in the docs that says what we have to do to code async tools?
ok i see in other posts that there is no code difference...it just doesnt speak the result immediately.
a
Let us know if you need additional assistance