huggingface/smolagents · error · TypeError

Input '{input_name}': type must be a string or list of strin

Error message

Input '{input_name}': type must be a string or list of strings, got {type(input_content['type']).__name__}

What it means

During Tool initialization, smolagents validates the 'inputs' attribute: each input's 'type' must be either a string (e.g. 'string') or a list of strings. If type is any other Python type (int, dict, None, etc.), a TypeError is raised naming the offending type. This guards downstream JSON-schema generation for the LLM.

Source

Thrown at src/smolagents/tools.py:188

            )
        # Validate inputs
        for input_name, input_content in self.inputs.items():
            assert isinstance(input_content, dict), f"Input '{input_name}' should be a dictionary."
            assert "type" in input_content and "description" in input_content, (
                f"Input '{input_name}' should have keys 'type' and 'description', has only {list(input_content.keys())}."
            )
            # Get input_types as a list, whether from a string or list
            if isinstance(input_content["type"], str):
                input_types = [input_content["type"]]
            elif isinstance(input_content["type"], list):
                input_types = input_content["type"]
                # Check if all elements are strings
                if not all(isinstance(t, str) for t in input_types):
                    raise TypeError(
                        f"Input '{input_name}': when type is a list, all elements must be strings, got {input_content['type']}"
                    )
            else:
                raise TypeError(
                    f"Input '{input_name}': type must be a string or list of strings, got {type(input_content['type']).__name__}"
                )
            # Check all types are authorized
            invalid_types = [t for t in input_types if t not in AUTHORIZED_TYPES]
            if invalid_types:
                raise ValueError(f"Input '{input_name}': types {invalid_types} must be one of {AUTHORIZED_TYPES}")
        # Validate output type
        assert getattr(self, "output_type", None) in AUTHORIZED_TYPES

        # Validate forward function signature, except for Tools that use a "generic" signature (PipelineTool, SpaceToolWrapper, LangChainToolWrapper)
        if not (
            hasattr(self, "skip_forward_signature_validation")
            and getattr(self, "skip_forward_signature_validation") is True
        ):
            signature = inspect.signature(self.forward)
            actual_keys = set(key for key in signature.parameters.keys() if key != "self")
            expected_keys = set(self.inputs.keys())
            if actual_keys != expected_keys:

View on GitHub (pinned to 30bb116109)

Solutions

1. Set the input's type to a string literal like "type": "string"
2. If multiple types are allowed, use a list of strings: "type": ["string", "integer"]
3. Ensure every entry in self.inputs has a 'type' key at all (a missing key raises KeyError instead)

Example fix

# before
self.inputs = {"text": {"type": None, "description": "..."}}
# after
self.inputs = {"text": {"type": "string", "description": "..."}}

def valid_input_schema(inputs):
    for name, spec in inputs.items():
        t = spec.get("type")
        if not isinstance(t, str) and not (isinstance(t, list) and all(isinstance(x, str) for x in t)):
            return False
    return True

assert valid_input_schema(MyTool.inputs)

from typing import Union
def is_valid_type(t: object) -> bool:
    return isinstance(t, str) or (isinstance(t, list) and t and all(isinstance(x, str) for x in t))

When it happens

Trigger: Defining a Tool subclass (or @tool-decorated function's inputs dict) where an input entry has 'type': None, 'type': 3, or 'type': {'type': 'string'} instead of 'type': 'string' or 'type': ['string', 'integer'].

Common situations: Copying Gradio API descriptions or JSON schemas verbatim into inputs, forgetting quotes around type names, or using numpy/python types instead of string literals.

Related errors

• Input '{input_name}': when type is a list, all elements must
• step_callbacks must be a list or a dict
• You must set an attribute {attr}.
• Attribute {attr} should have type {expected_type.__name__},
• Attribute output_schema should have type dict, got {type(out

AI-assisted analysis of huggingface/smolagents@30bb116109 (2026-08-28). Data as JSON: /api/errors/6fdba57808aba75b. Report an issue: GitHub.