Crate sim_lib_openai_server 

OpenAI-shaped gateway routes for SIM.

This crate provides the OpenAI gateway health route, OpenAI JSON transcript codec, model discovery, gateway object helpers, plan parsing and fixture atom execution through POST /v1/responses and POST /v1/chat/completions, embeddings through POST /v1/embeddings, SSE streaming projection, memory-backed response retrieval by id, and stored response replay/fork inspection routes. File, audio, image, and vector-store routes are SIM JSON fixture/subset surfaces: they use OpenAI-shaped paths and response objects, while request bodies use the explicit JSON fields handled by these route modules rather than multipart upload or provider media protocols. The model list route advertises models from registered ModelCard records, so loadable local placement sites appear beside fixture and remote runner entries when installed.

Re-exports

pub use capabilities::AI_RUNNER_CACHE_CAPABILITY;

pub use capabilities::NETWORK_CAPABILITY;

pub use capabilities::OPENAI_GATEWAY_ADMIN_CAPABILITY;

pub use capabilities::OPENAI_GATEWAY_FEDERATE_CAPABILITY;

pub use capabilities::OPENAI_GATEWAY_PLAN_CAPABILITY;

pub use capabilities::OPENAI_GATEWAY_PLAN_REMOTE_CAPABILITY;

pub use capabilities::OPENAI_GATEWAY_SERVE_CAPABILITY;

pub use capabilities::OPENAI_GATEWAY_TOOLS_CAPABILITY;

pub use capabilities::WEBHOOK_SERVE_CAPABILITY;

pub use capabilities::ai_runner_cache_capability;

pub use capabilities::network_capability;

pub use capabilities::openai_gateway_admin_capability;

pub use capabilities::openai_gateway_federate_capability;

pub use capabilities::openai_gateway_plan_capability;

pub use capabilities::openai_gateway_plan_remote_capability;

pub use capabilities::openai_gateway_serve_capabilities;

pub use capabilities::openai_gateway_serve_capability;

pub use capabilities::openai_gateway_tools_capability;

pub use capabilities::require_openai_gateway_serve_capabilities;

pub use capabilities::webhook_serve_capability;

pub use codec_openai::ChatTranscript;

pub use codec_openai::GatewayEventData;

pub use codec_openai::OpenAiSseSurface;

pub use codec_openai::StreamSink;

pub use codec_openai::encode_gateway_events_sse;

pub use codec_openai::encode_openai_responses_response;

pub use codec_openai::gateway_event_data_from_packet;

pub use codec_openai::gateway_event_data_kind;

pub use codec_openai::gateway_event_data_packets;

pub use content_id::VOLATILE_REQUEST_FIELDS;

pub use content_id::content_id_for_expr;

pub use content_id::normalize_request_expr;

pub use content_id::request_content_id;

pub use content_id::request_expr_content_id;

pub use ids::GatewayIdGenerator;

pub use install::install_openai_gateway_lib;

pub use manifest::OpenAiGatewayLib;

pub use manifest::manifest_name;

pub use objects::GATEWAY_EVENT_OBJECT;

pub use objects::GATEWAY_REQUEST_OBJECT;

pub use objects::GATEWAY_RESPONSE_OBJECT;

pub use objects::GATEWAY_RUN_OBJECT;

pub use objects::GatewayEvent;

pub use objects::GatewayRequest;

pub use objects::GatewayResponse;

pub use objects::GatewayResponseValue;

pub use objects::GatewayRun;

pub use objects::content_id_expr;

pub use objects::content_id_hex;

pub use ops::OpenAiGatewayFunction;

pub use ops::cache_stats_symbol;

pub use ops::capability_report_symbol;

pub use ops::events_symbol;

pub use ops::fabric_symbol;

pub use ops::health_symbol;

pub use ops::key_add_symbol;

pub use ops::key_list_symbol;

pub use ops::model_health_symbol;

pub use ops::models_symbol;

pub use ops::plan_check_symbol;

pub use ops::plan_combinators_symbol;

pub use ops::plan_explain_symbol;

pub use ops::plan_parse_symbol;

pub use ops::plan_run_symbol;

pub use ops::run_get_symbol;

pub use ops::runs_symbol;

pub use ops::serve_symbol;

pub use ops::storage_stats_symbol;

pub use plan::BackendDescriptor;

pub use plan::PlanCombinator;

pub use plan::PlanEvalEvent;

pub use plan::PlanEvalReport;

pub use plan::PlanLimits;

pub use plan::check_plan;

pub use plan::check_plan_with_limits;

pub use plan::eval_plan;

pub use plan::eval_plan_report;

pub use plan::eval_plan_report_with_cache;

pub use plan::eval_plan_report_with_cache_and_runners;

pub use plan::eval_plan_report_with_cache_runners_and_federation;

pub use plan::eval_plan_report_with_federation;

pub use plan::explain_plan;

pub use plan::parse_plan;

pub use plan::plan_combinators;

pub use plan::plan_combinators_expr;

pub use plan::plan_symbol;

pub use plan::provider_prefixes;

pub use plan::resolve_atom_address;

pub use routes::admin::ADMIN_CACHE_STATS_PATH;

pub use routes::admin::ADMIN_CAPABILITY_REPORT_PATH;

pub use routes::admin::ADMIN_EVENTS_PATH;

pub use routes::admin::ADMIN_MODEL_HEALTH_PATH;

pub use routes::admin::ADMIN_RUN_RETRIEVAL_PREFIX;

pub use routes::admin::ADMIN_RUN_RETRIEVAL_ROUTE;

pub use routes::admin::ADMIN_RUNS_PATH;

pub use routes::admin::ADMIN_STORAGE_STATS_PATH;

pub use routes::audio::AUDIO_SPEECH_PATH;

pub use routes::audio::AUDIO_TRANSCRIPTIONS_PATH;

pub use routes::audio::handle_audio_speech;

pub use routes::audio::handle_audio_transcriptions;

pub use routes::batches::BATCH_CANCEL_ROUTE;

pub use routes::batches::BATCH_RETRIEVAL_PREFIX;

pub use routes::batches::BATCH_RETRIEVAL_ROUTE;

pub use routes::batches::BATCHES_PATH;

pub use routes::batches::handle_batch_cancel;

pub use routes::batches::handle_batch_retrieval;

pub use routes::batches::handle_batches;

pub use routes::batches::retrieve_batch;

pub use routes::chat_completions::CHAT_COMPLETIONS_PATH;

pub use routes::chat_completions::ChatCompletionExecution;

pub use routes::chat_completions::execute_chat_completion_request;

pub use routes::chat_completions::execute_chat_completion_request_with_runners;

pub use routes::chat_completions::execute_chat_completion_request_with_runners_and_federation;

pub use routes::chat_completions::handle_chat_completions;

pub use routes::embeddings::EMBEDDINGS_PATH;

pub use routes::embeddings::EmbeddingExecution;

pub use routes::embeddings::EmbeddingIdGenerators;

pub use routes::embeddings::TENSOR_F64_SMALL_EMBEDDING_MODEL;

pub use routes::embeddings::execute_embedding_request;

pub use routes::embeddings::handle_embeddings;

pub use routes::files::FILE_RETRIEVAL_PREFIX;

pub use routes::files::FILE_RETRIEVAL_ROUTE;

pub use routes::files::FILES_PATH;

pub use routes::files::handle_file_retrieval;

pub use routes::files::handle_files;

pub use routes::files::retrieve_file;

pub use routes::health::HEALTH_BODY;

pub use routes::health::HEALTH_PATH;

pub use routes::health::health_response;

pub use routes::images::IMAGES_GENERATIONS_PATH;

pub use routes::images::handle_image_generations;

pub use routes::models::MODELS_PATH;

pub use routes::models::ModelCatalog;

pub use routes::models::OpenAiModel;

pub use routes::models::models_response;

pub use routes::models::models_response_for_catalog;

pub use routes::models::models_response_for_runner_args;

pub use routes::replay::RESPONSE_EVENTS_ROUTE;

pub use routes::replay::RESPONSE_EVENTS_SUFFIX;

pub use routes::replay::RESPONSE_SIM_ROUTE;

pub use routes::replay::RESPONSE_SIM_SUFFIX;

pub use routes::replay::SIM_EXTENSION_CAPABILITY;

pub use routes::replay::SIM_FORK_PATH;

pub use routes::replay::SIM_INSPECTION_CAPABILITY;

pub use routes::replay::SIM_REPLAY_PATH;

pub use routes::replay::handle_response_events;

pub use routes::replay::handle_response_sim;

pub use routes::replay::handle_sim_fork;

pub use routes::replay::handle_sim_replay;

pub use routes::replay::response_events;

pub use routes::replay::response_sim;

pub use routes::responses::RESPONSE_RETRIEVAL_PREFIX;

pub use routes::responses::RESPONSE_RETRIEVAL_ROUTE;

pub use routes::responses::RESPONSES_PATH;

pub use routes::responses::ResponseExecution;

pub use routes::responses::ResponseIdGenerators;

pub use routes::responses::ResponseRuntimeTargets;

pub use routes::responses::execute_response_request;

pub use routes::responses::execute_response_request_with_cache;

pub use routes::responses::execute_response_request_with_cache_and_runners;

pub use routes::responses::execute_response_request_with_cache_runners_and_federation;

pub use routes::responses::execute_response_request_with_runners;

pub use routes::responses::handle_response_retrieval;

pub use routes::responses::handle_responses;

pub use routes::responses::retrieve_response;

pub use routes::threads::THREAD_MESSAGES_ROUTE;

pub use routes::threads::THREAD_RETRIEVAL_PREFIX;

pub use routes::threads::THREAD_RETRIEVAL_ROUTE;

pub use routes::threads::THREADS_PATH;

pub use routes::threads::handle_thread_get;

pub use routes::threads::handle_thread_post;

pub use routes::threads::handle_threads;

pub use routes::threads::list_messages;

pub use routes::threads::retrieve_thread;

pub use routes::vector_stores::VECTOR_STORE_SEARCH_PREFIX;

pub use routes::vector_stores::VECTOR_STORE_SEARCH_ROUTE;

pub use routes::vector_stores::VECTOR_STORE_SEARCH_SUFFIX;

pub use routes::vector_stores::VECTOR_STORES_PATH;

pub use routes::vector_stores::handle_vector_store_search;

pub use routes::vector_stores::handle_vector_stores;

pub use runtime::OPENAI_GATEWAY_KEY_OBJECT;

pub use runtime::OpenAiFederatedGateway;

pub use runtime::OpenAiFederation;

pub use runtime::OpenAiFederationPolicy;

pub use runtime::OpenAiGatewayFabric;

pub use runtime::OpenAiGatewayKey;

pub use runtime::OpenAiKeyTable;

pub use runtime::OpenAiPlanCache;

pub use runtime::OpenAiRunnerRegistry;

pub use runtime::PlanCacheKey;

pub use runtime::PlanCacheMode;

pub use runtime::PlanCacheWriteTarget;

pub use runtime::ToolLoopConfig;

pub use runtime::global_openai_key_table;

pub use runtime::key_hash;

pub use runtime::redacted_gateway_request;

pub use runtime::run_tool_loop_with_cache;

pub use runtime::run_tool_loop_with_registry;

pub use server::GatewayRouteState;

pub use server::GatewayRoutes;

pub use server::GatewayRoutesValue;

pub use server::configure_routes;

pub use server::configure_routes_with_state;

pub use storage::GATEWAY_BATCH_KIND;

pub use storage::GATEWAY_FILE_KIND;

pub use storage::GATEWAY_THREAD_KIND;

pub use storage::GATEWAY_THREAD_MESSAGE_KIND;

pub use storage::GATEWAY_VECTOR_STORE_ITEM_KIND;

pub use storage::GATEWAY_VECTOR_STORE_KIND;

pub use storage::GatewayBatch;

pub use storage::GatewayBatchCounts;

pub use storage::GatewayBatchStatus;

pub use storage::GatewayFile;

pub use storage::GatewayFileStorageRef;

pub use storage::GatewayResponseObjectStore;

pub use storage::GatewayStateStore;

pub use storage::GatewayStore;

pub use storage::GatewayThread;

pub use storage::GatewayThreadMessage;

pub use storage::GatewayVectorStore;

pub use storage::GatewayVectorStoreItem;

pub use storage::MemoryGatewayStore;

pub use storage::StoredGatewayResponse;

pub use storage::TableGatewayStore;

pub use translate::tools::OpenAiTool;

pub use translate::tools::OpenAiToolCall;

pub use translate::tools::OpenAiToolRegistry;

pub use translate::tools::OpenAiToolResult;

pub use translate::tools::openai_name_to_symbol;

pub use translate::tools::symbol_from_text;

Modules

capabilities

Capability identifiers and constructors that gate gateway operations.

citizen

Citizen descriptor types bridging gateway objects into the SIM object graph.

clock

Compatibility names for the canonical server wall-clock contract. Compatibility names for the canonical server wall-clock contract.

codec_openai

OpenAI JSON codec surface: decode, encode, streaming, and shared shapes.

content_id

Content-id helpers that normalize and hash gateway request expressions.

ids

Identifier generation for gateway requests, responses, and runs.

install

Installation glue that registers the gateway library with a runtime.

manifest

Library manifest naming and export wiring for the gateway.

objects

Core gateway object records: requests, responses, runs, and events.

ops

Gateway operations exposed to the runtime as callable functions.

plan

Plan parsing, checking, evaluation, and combinator surface. Plan subsystem: parsing, checking, and evaluating SIM plan expressions that orchestrate model atoms through combinators such as race, fallback, chain, verify, debate, and cache.

routes

HTTP route handlers mapping OpenAI endpoints onto SIM eval and agents.

runtime

Runtime support: key tables, plan cache, federation, runners, and fabric.

server

Server wiring that registers gateway routes and holds route state.

storage

Persistent and in-memory stores for gateway records and content objects. Storage backends for the OpenAI-shaped gateway.

translate

Translation between OpenAI tool/function shapes and SIM symbols.

Structs

DeterministicGatewayClock

Thread-safe WallClock that advances by a fixed step on each observation.

DeterministicWallClock

Thread-safe WallClock that advances by a fixed step on each observation.

OpenAiCodec

Runtime codec for OpenAI-compatible chat-completion JSON.

OpenAiCodecOptions

Options for OpenAI-compatible provider request JSON generation.

SystemGatewayClock

WallClock backed by the host system clock.

SystemWallClock

WallClock backed by the host system clock.

WallTimestamp

One observed host wall-clock instant, represented as milliseconds since the Unix epoch.

Statics

RECIPES

Cookbook recipes for this lib, embedded at build time.

Traits

GatewayClock

Object-safe source of host wall-clock observations.

WallClock

Object-safe source of host wall-clock observations.

Functions

decode_openai_request

Decodes OpenAI request JSON into a validated chat-transcript expression.

decode_openai_response

Decodes an OpenAI chat-completion response body into a model-response transcript, optionally embedding the raw provider JSON.

encode_openai_request

Encodes a model-request transcript into OpenAI chat-completion JSON.

encode_openai_response

Encodes a model-response transcript into OpenAI chat-completion JSON.

openai_codec_symbol

Returns the codec symbol codec:openai.

Type Aliases

OpenAiRequestOptions

Alias for OpenAiCodecOptions when used for provider request generation.