Struct RouterProvider 

pub struct RouterProvider { /* private fields */ }

Routes each turn to a backend that still has budget.

Implementations

impl RouterProvider

pub fn new( backends: Vec<Backend>, strategy: Box<dyn RoutingStrategy>, ) -> Result<Self, RunnerError>

Build a router over at least one backend.

Errors

Returns an error when no backends are given — a router over nothing has no honest behaviour, and failing at construction beats failing on the first athlete’s turn.

pub fn with_store(self, store: Box<dyn QuotaStore>) -> Self

Swap the state store — the seam a shared implementation plugs into.

The default is InMemoryQuotaStore, correct for one process. Several processes spending one account pass RedisQuotaStore (feature quota-redis); routers in one process can share an Arc<InMemoryQuotaStore>. Neither changes routing or the strategy.

pub const fn with_threshold(self, percent: f32) -> Self

Set the share of a window at which the router steps aside.

pub const fn with_refresh_interval(self, interval: Duration) -> Self

Set how long a quota reading is reused before refreshing.

pub fn active_index(&self) -> usize

Index of the backend currently serving.

pub fn active_runner(&self) -> Option<Arc<dyn HeadlessTurnProvider>>

The concrete ACP runner behind the backend that is currently answering.

None when the live backend is not a Copilot Headless one — which is the correct answer, not a missing feature. The platform’s dispatch reads this to decide between the SDK tool-calling loop and the CLI text loop, and a stale Some would run the wrong loop against the wrong provider.

Trait Implementations

impl LlmProvider for RouterProvider

fn name(&self) -> &'static str

Unique provider identifier (e.g., claude_code, copilot)

fn display_name(&self) -> &str

Human-readable display name for the provider

fn capabilities(&self) -> LlmCapabilities

Provider capabilities (streaming, function calling, etc.)

fn default_model(&self) -> &str

Default model to use if not specified in request

fn available_models(&self) -> &[String]

Available models for this provider

fn complete<'life0, 'life1, 'async_trait>( &'life0 self, request: &'life1 ChatRequest, ) -> Pin<Box<dyn Future<Output = Result<ChatResponse, RunnerError>> + Send + 'async_trait>>

where Self: 'async_trait, 'life0: 'async_trait, 'life1: 'async_trait,

Perform a chat completion (non-streaming)

fn complete_stream<'life0, 'life1, 'async_trait>( &'life0 self, request: &'life1 ChatRequest, ) -> Pin<Box<dyn Future<Output = Result<ChatStream, RunnerError>> + Send + 'async_trait>>

where Self: 'async_trait, 'life0: 'async_trait, 'life1: 'async_trait,

Perform a streaming chat completion

fn complete_stream_with_tools<'life0, 'life1, 'async_trait>( &'life0 self, request: &'life1 ChatRequest, ) -> Pin<Box<dyn Future<Output = Result<ToolStream, RunnerError>> + Send + 'async_trait>>

where Self: 'async_trait, 'life0: 'async_trait, 'life1: 'async_trait,

Perform a chat completion that may call tools, streaming its text.

fn health_check<'life0, 'async_trait>( &'life0 self, ) -> Pin<Box<dyn Future<Output = Result<bool, RunnerError>> + Send + 'async_trait>>

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

Check if the provider is healthy and ready to serve requests