Crate prompty_openai 

OpenAI provider for Prompty.

This crate provides the OpenAI executor and processor implementations, registered under the key "openai".

Usage

prompty_openai::register();
// Now invoke/turn will use OpenAI for agents with provider="openai"

Re-exports

pub use executor::OpenAIExecutor;

pub use models::list_models;

pub use models::list_models_async;

pub use processor::OpenAIProcessor;

pub use processor::extract_tool_calls;

pub use processor::process_response;

pub use wire::build_chat_args;

pub use wire::build_embedding_args;

pub use wire::build_image_args;

pub use wire::build_responses_args;

pub use wire::format_tool_messages;

pub use wire::message_to_wire;

pub use wire::tools_to_wire;

Modules

executor

OpenAI executor — sends requests to the OpenAI Chat Completions API.

models

Model discovery for OpenAI — calls GET /v1/models and enriches results with known context-window and modality data.

processor

OpenAI processor — extracts results from OpenAI API responses.

wire

Wire format conversion for the OpenAI Chat Completions API.

Functions

register

Register the OpenAI executor and processor in the global registry.