BerriAI/litellm · error · OCIError

Tool call `function.name` must be a string

Error message

Tool call `function.name` must be a string

What it means

The nested function object of a tool call must have a string 'name'. A missing, None, or non-string name raises OCIError(400) "Tool call `function.name` must be a string" before the request leaves the client. The name is required because OCI's OCIToolCall maps it to the FUNCTION tool invocation name.

Source

Thrown at litellm/llms/oci/chat/generic.py:133

    """Convert an assistant tool-call message to OCI format."""
    tool_calls_formatted: Final = []
    for tool_call in tool_calls:
        if not isinstance(tool_call, dict):
            raise OCIError(status_code=400, message="Each tool call must be a dictionary")
        if tool_call.get("type") != "function":
            raise OCIError(status_code=400, message="OCI only supports function tool calls")

        tool_call_id = tool_call.get("id")
        if not isinstance(tool_call_id, str):
            raise OCIError(status_code=400, message="Tool call `id` must be a string")

        tool_function = tool_call.get("function")
        if not isinstance(tool_function, dict):
            raise OCIError(status_code=400, message="Tool call `function` must be a dictionary")

        function_name = tool_function.get("name")
        if not isinstance(function_name, str):
            raise OCIError(status_code=400, message="Tool call `function.name` must be a string")

        arguments = tool_call["function"].get("arguments", "{}")
        if not isinstance(arguments, str):
            raise OCIError(
                status_code=400,
                message="Tool call `function.arguments` must be a JSON string",
            )

        tool_calls_formatted.append(
            OCIToolCall(
                id=tool_call_id,
                type="FUNCTION",
                name=function_name,
                arguments=arguments,
            )
        )

    return OCIMessage(

View on GitHub (pinned to 6c2dcb801b)

Solutions

  1. Always set function.name to the exact tool name string that was declared in the tools parameter.
  2. Render-check templates so name is never empty.
  3. Pre-flight assert isinstance(tc['function'].get('name'), str) and it's non-empty.

Example fix

# before
{'id':'c1','type':'function','function':{'arguments':'{"city":"SF"}'}}

# after
{'id':'c1','type':'function','function':{'name':'get_weather','arguments':'{"city":"SF"}'}}
Defensive patterns

Strategy: type-guard

Validate before calling

for tc in msg.get('tool_calls') or []:
    fn = tc.get('function') or {}
    assert isinstance(fn.get('name'), str) and fn['name'], f'tool call function.name missing: {tc!r}'

Type guard

def tool_call_has_string_name(tc: object) -> bool:
    return (
        isinstance(tc, dict)
        and isinstance(tc.get('function'), dict)
        and isinstance(tc['function'].get('name'), str)
        and bool(tc['function']['name'])
    )

Prevention

When it happens

Trigger: Sending {'id':'c1','type':'function','function':{'arguments':'{}'}} (name omitted) or with name set to None/numeric in assistant history to an oci/ GENERIC model.

Common situations: Synthetic few-shot tool turns built without names; templated history where the name placeholder failed to render; data pipelines that rename/drop the name key.

Related errors


AI-assisted analysis of BerriAI/litellm@6c2dcb801b (2026-08-15). Data as JSON: /api/errors/bf40b55828135b51. Report an issue: GitHub.