You can define a function tool in LiveKit that accepts a list as an attribute by using the

@function_tool

decorator along with Python type hints. Here is an example:

from livekit.agents import function_tool, RunContext

@function_tool
def process_numbers(context: RunContext, numbers: list[float]) -> str:
    """Sums a list of numbers and returns the result."""
    return str(sum(numbers))

You can specify the expected list type—such as

list[str]

for a list of strings,

list[int]

for integers, etc. The type hints ensure the schema sent to the LLM correctly identifies argument types, letting your tool receive lists as expected. For more complex structures (such as lists of objects/dictionaries), you can use a raw schema in your function tool definition. Refer to this example:

raw_schema = {
    "type": "function",
    "name": "process_user_list",
    "parameters": {
        "type": "object",
        "properties": {
            "users": {
                "type": "array",
                "items": {
                    "type": "object",
                    "properties": {
                        "name": {"type": "string"},
                        "age": {"type": "integer"}
                    },
                    "required": ["name", "age"]
                }
            }
        },
        "required": ["users"]
    }
}

@function_tool(raw_schema=raw_schema)
def process_user_list(raw_arguments, context: RunContext):
    users = raw_arguments["users"]
    return f"Processed {len(users)} users"

Type hints and docstrings are important for making the expected data structure clear. Tool return values are auto-converted to strings, so handle your function output accordingly. See the complete guide and more examples in the LiveKit knowledge base if you need to handle advanced use-cases like tool factories or lists of nested objects. • Type hint the list in function arguments for standard lists • Use

raw_schema

for complex list/object inputs Sources: Tool definition and use | LiveKit Docs | realtime/openai-realtime-tools.py