Struct OpenAIProtocolChatDriver 

pub struct OpenAIProtocolChatDriver { /* private fields */ }

OpenAI Protocol Chat Driver

Base implementation of ChatDriver for OpenAI-compatible APIs. Supports streaming responses and tool calls.

Rate limit handling: On 429 errors, automatically retries with exponential backoff, respecting x-ratelimit-reset-* and retry-after headers.

This is the base protocol driver used in examples and for OpenAI-compatible endpoints. For production use with OpenAI, consider using OpenAIChatDriver from the everruns-openai crate.

Example

use everruns_core::OpenAIProtocolChatDriver;

let driver = OpenAIProtocolChatDriver::new();
// Endpoint and authentication are configured on a runtime Provider.
// Retry policy remains a wire-protocol concern.
let driver = OpenAIProtocolChatDriver::new()
    .with_retry_config(LlmRetryConfig::aggressive());

Implementations

impl OpenAIProtocolChatDriver

pub fn new() -> Self

Create a wire-only OpenAI Chat Completions protocol driver.

pub fn with_retry_config(self, config: LlmRetryConfig) -> Self

Configure retry behavior for rate limit errors

pub fn client(&self) -> &Client

Get the HTTP client (for subclass access)

Trait Implementations

impl ChatDriver for OpenAIProtocolChatDriver

fn supports_parallel_tool_calls(&self, _model: &str) -> bool

OpenAI-compatible Chat Completions accept the top-level parallel_tool_calls boolean, so the preference maps directly onto the wire for every model served through this protocol.

fn chat_completion_stream<'life0, 'life1, 'life2, 'async_trait>( &'life0 self, endpoint: &'life1 ProviderEndpoint, messages: Vec<LlmMessage>, config: &'life2 LlmCallConfig, ) -> Pin<Box<dyn Future<Output = Result<LlmResponseStream>> + Send + 'async_trait>>

where Self: 'async_trait, 'life0: 'async_trait, 'life1: 'async_trait, 'life2: 'async_trait,

Call the LLM with streaming response

fn chat_completion<'life0, 'life1, 'life2, 'async_trait>( &'life0 self, endpoint: &'life1 ProviderEndpoint, messages: Vec<LlmMessage>, config: &'life2 LlmCallConfig, ) -> Pin<Box<dyn Future<Output = Result<LlmResponse>> + Send + 'async_trait>>

where Self: 'async_trait, 'life0: 'async_trait, 'life1: 'async_trait, 'life2: 'async_trait,

Call the LLM without streaming (convenience method)

fn list_models<'life0, 'life1, 'async_trait>( &'life0 self, _endpoint: &'life1 ProviderEndpoint, ) -> Pin<Box<dyn Future<Output = Result<Option<Vec<DiscoveredModel>>>> + Send + 'async_trait>>

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

List available models from the provider

fn supports_compact(&self) -> bool

Check if this driver supports the compact endpoint

fn supports_stateful_responses(&self) -> bool

Whether this driver persists Responses API state and can resolve tool calls that are reachable only through previous_response_id.

fn effective_context_window(&self, _model: &str) -> Option<usize>

Effective context window for model, when the driver has authoritative model metadata that is not represented by Everruns’ built-in profiles.

fn compact<'life0, 'life1, 'async_trait>( &'life0 self, _endpoint: &'life1 ProviderEndpoint, _request: CompactRequest, ) -> Pin<Box<dyn Future<Output = Result<Option<CompactResponse>>> + Send + 'async_trait>>

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

Compact a conversation to reduce context size

impl Clone for OpenAIProtocolChatDriver

fn clone(&self) -> OpenAIProtocolChatDriver

Returns a duplicate of the value.

fn clone_from(&mut self, source: &Self)

Performs copy-assignment from source.

impl Debug for OpenAIProtocolChatDriver

fn fmt(&self, f: &mut Formatter<'_>) -> Result

Formats the value using the given formatter.

impl Default for OpenAIProtocolChatDriver

fn default() -> Self

Returns the “default value” for a type.