Struct LlmModelExecutor 

pub struct LlmModelExecutor { /* private fields */ }

The default ModelExecutor, wrapping a general salvor_llm::Client.

This is not consumer-specific: it wraps the general model transport, so any host that reaches a provider through salvor-llm gets a working executor by constructing a Client and handing it here. salvor serve builds one from its environment client-construction path.

Implementations

impl LlmModelExecutor

pub fn new(client: Client) -> Self

Wraps client as a model executor.

Trait Implementations

impl ModelExecutor for LlmModelExecutor

fn execute<'life0, 'async_trait>( &'life0 self, request: Value, ) -> Pin<Box<dyn Future<Output = Result<MessageResponse, String>> + Send + 'async_trait>>

where Self: 'async_trait, 'life0: 'async_trait,

Perform a single (non-streaming) model call and return the response.

fn open_stream<'life0, 'async_trait>( &'life0 self, request: Value, ) -> Pin<Box<dyn Future<Output = Result<Box<dyn ModelStream>, String>> + Send + 'async_trait>>

where Self: 'async_trait, 'life0: 'async_trait,

Open a streaming model call, returning a ModelStream of provider events the server pumps for the live ticker.