Struct StreamingFunctionCallingAgent 

pub struct StreamingFunctionCallingAgent { /* private fields */ }

Streaming Function Calling Agent

Streams LLM text (token by token), then emits FinalAnswer at the end. Tool-call state is exposed via AgentStreamEvent::ToolCall. Works with any LLM provider implementing BaseChatModel.

Implementations

impl StreamingFunctionCallingAgent

pub fn new<L>(llm: L) -> Self

where L: BaseChatModel + Send + Sync + 'static, L::Error: Into<ProviderError>,

Creates a new Streaming Function Calling Agent

Backward compatibility

Old code StreamingFunctionCallingAgent::new(openai_chat) still works.

pub fn from_arc( llm: Arc<dyn BaseChatModel<Error = ProviderError> + Send + Sync>, ) -> Self

Creates an agent from a wrapped Arc<dyn BaseChatModel>

pub async fn invoke_stream( &self, input: String, ) -> Pin<Box<dyn Stream<Item = AgentStreamEvent> + Send>>

Streams execution: returns an event stream