Troubleshooting HTTP tool input schema mismatch with 'Let Model Specify Entire Body' enabled

Can you verify if the JSON it produced functions correctly with an actual request?

Yes, this JSON works with a real request.

image2200×1236 197 KB

You could try constructing the fields using the options provided below. This might help by defining a schema behind the scenes:

Screenshot 2025-03-23 at 1.15.58 PM822×456 21.8 KB

When I set “Value Provided: By Model”, the tool functions as anticipated.

image757×345 6.43 KB

I examined the tool's input and observed that it appeared as follows:

image1920×823 150 KB

{
  "query": {
    "query": "history of #HalaMadrid"
  }
}

However, my expectation was simply:

{
  "query": "history of #HalaMadrid"
}

Therefore, I suspect that when using “Specify Body: Let Model Specify Entire Body”, the anticipated JSON structure might also be enclosed in a similar fashion.

To verify this, I adjusted the LLM's output format as depicted in the LangGraph trace screenshot below, but I still encountered the same error:

Received tool input did not match expected schema

image1829×515 38.4 KB

Great! So that was successful? I'm happy to hear that!

That is peculiar, though.

However, if my response was helpful, I would appreciate it if you could mark it as the solution.

Hey! In my situation, the issue was that the AI Assistant was sending a json object but callin.io was expecting a string!

callin.io expects a string by default after you select Defined automatically by the **model**. In the background, it appears callin.io uses the $fromAI function. The 3rd parameter of the function is the expected type for that property. In my case, that type was always expected to be string. See the gif below. It’s changed to string even if I set it to json.

So, the model sends json object and callin.io is expecting string. Therefore, I was getting Received tool input did not match expected schema.

The solution was to remove the Defined automatically by the **model** and use an expression to set the value of the input as {{ $fromAI('inputName', 'input description', 'json') }}. I spent many hours on this. I hope it helps you save some time.

I'm working on a very similar workflow to the one shared previously: an AI agent utilizing an OpenAI gpt4.1-mini model. This workflow functioned flawlessly on version 1.88 of callin.io.

Subsequently, I updated to the latest version, 1.90.2, and began encountering an error. I then switched to a different OpenAI model (gpt 4.1-nano), and the error was resolved.

Hi~ thanks for sharing your experience! I tried running the workflow I shared above again, and in my current callin.io version (1.90.1), the ‘Let Model Specify Entire Body’ option has already been removed from the Http tool node.

Now there are only ‘Using JSON’ and ‘Using Fields Below’ options left. So, I tried using the ‘Using JSON’ option to let the model construct the entire request body, and so far, it seems to be working properly.

image916×446 19.9 KB

Thanks for discovering this – I spent quite a bit of time trying to resolve this issue!

Could you assist me with configuring a single field named 'data' within the flow module, incorporating the specified parameters? I'm also wondering how the incoming data can be automatically sorted by suffix, or if you could provide a more detailed explanation. Thank you!