This repository is the openai/codex monorepo, which implements a local AI coding agent called Codex (also referred to as Codex CLI). It provides a terminal user interface (TUI), command-line interface (CLI), and a local app-server daemon that manages AI-driven coding sessions. The architecture is client-server: the TUI/CLI communicates with the local app-server via JSON-RPC over Unix sockets or WebSockets, while the exec-server handles sandboxed command execution with platform-specific security policies (macOS Seatbelt, Windows sandbox, Linux bubblewrap). The system is built entirely in Rust, with additional TypeScript and Python SDKs for programmatic access. It supports a rich extension ecosystem including plugins, skills, MCP servers, memories, and a guardian safety layer for approval workflows. The protocol layer is generated from Rust types into TypeScript schemas, ensuring type safety across the boundary. The codebase is well-structured, heavily tested, and follows modern Rust practices (async/await, trait-based abstractions, comprehensive error handling). It is not a data research project; there are no notebooks, CSV files, or ML pipelines.

Whisper is a speech recognition system that provides both a command-line interface and a Python library for transcribing and translating audio. The system processes audio by converting it to log-Mel spectrograms, then uses a Transformer-based encoder-decoder model to predict text tokens. Key components include audio processing, model architecture, decoding strategies, tokenization, and utilities for word-level timestamp alignment. The codebase is modular, well-tested, and actively maintained with recent updates supporting modern Python and PyTorch versions.

OpenAI Gym is a foundational Python library providing a standard API for Reinforcement Learning (RL) environments. It defines the core `Env` interface (`reset`, `step`, `render`) and includes a registry of environments spanning classic control (CartPole), physics simulations (MuJoCo, Box2D), and grid worlds. The architecture relies heavily on a composition-based wrapper pattern (`gym.wrappers`) to modify environment behavior without altering core logic, and includes a vectorization module (`gym.vector`) for parallel execution using shared memory. While the codebase demonstrates strong architectural craftsmanship in its abstraction layers and space definitions, it is in a maintenance phase with development migrated to Gymnasium. It relies on external compiled libraries (MuJoCo, Box2D, Pygame) which introduce platform-specific complexity and dependency drift.

The openai/CLIP repository provides a Python library for loading and using CLIP (Contrastive Language-Image Pre-Training) models. It offers a clean API for model loading, image preprocessing, text tokenization, and computing similarity between images and text. The core implementation includes two vision encoder architectures (ModifiedResNet and VisionTransformer) and a Transformer-based text encoder, all built with PyTorch. The library is designed for research and inference, enabling zero-shot image classification, image-text retrieval, and feature extraction. While primarily a research artifact, the code is well-structured and demonstrates strong engineering practices, with reusable components such as the BPE tokenizer and model architecture definitions. The repository is stable but not actively developed, with dependencies on older PyTorch versions.

The openai-python repository is the official Python SDK for interacting with OpenAI's API services. It provides both synchronous and asynchronous clients for accessing OpenAI's REST and WebSocket APIs, including chat completions, embeddings, image generation, audio processing, fine-tuning, vector stores, and admin organization management. The library is generated from OpenAPI specifications using Castiron, ensuring strong type safety via Pydantic models and TypedDict for request/response objects. Key features include streaming support, webhook verification, workload identity authentication, and modular resource organization. The SDK handles authentication, retry logic, pagination, and error handling transparently, making it suitable for production AI-integrated applications.

The OpenAI Agents SDK (Python) is a production-grade library for building agentic AI applications. It provides a complete agent runtime (`Runner`, `Agent`, `RunConfig`) with support for tool calling, handoffs, guardrails, sandboxed execution (Docker, Modal, E2B, Cloudflare, Vercel, Unix-local), realtime voice/audio pipelines, MCP integration, and multi-provider model routing (OpenAI, LiteLLM, AnyLLM, custom). The codebase is async-first with extensive tracing/observability, session persistence (SQLite, OpenAI Conversations, Responses Compaction), and a rich lifecycle-hook system. It includes a large test suite covering unit, integration, and sandbox security scenarios, plus internal tooling skills for release management and code review. The SDK is actively maintained with a 0.23.x version, release automation via release-please, and comprehensive docs in English and Japanese.

Symphony is an autonomous AI agent orchestration service written in Elixir/Phoenix, designed to manage coding agents (like OpenAI's Codex) that resolve issues from project trackers. In production, it polls Linear, GitHub, GitLab, Jira, or Asana for new work, creates isolated per-issue workspaces, launches Codex agents with provider-native tools, manages retries and backoff, and provides live observability via a terminal dashboard and optional web UI. The architecture follows OTP best practices: a supervisor tree with GenServers for the orchestrator state machine, polling loop, status dashboard, and tracker adapters. Key entry points include the CLI (`SymphonyElixir.CLI.main`), the Phoenix web endpoint, and Mix tasks for code quality checks. Data flows are event-driven: poll tracker -> normalize to Issue struct -> dispatch via AgentRunner -> create workspace -> run Codex turns -> cleanup. The system is explicitly experimental ("engineering preview") but demonstrates strong engineering in its adapter pattern, configuration design, and test infrastructure.

This repository is a lightweight, reference implementation of several standard LLM benchmarks (MMLU, GPQA, MATH, HumanEval, DROP, MGSM, SimpleQA, BrowseComp, HealthBench) developed by OpenAI. It provides a unified evaluation framework based on an `Eval` base class and `SamplerBase` that supports OpenAI Chat Completions API, Anthropic Claude API, OpenAI Responses API, and reasoning models (o1, o3, o4-mini). The main entry point is `simple_evals.py` which accepts command-line arguments to select models and evals, runs them, and outputs HTML reports and JSON metrics. The code is cleanly factored into per-benchmark modules, a shared utility module (`common.py`), and sampler adapters. The framework is designed for reproducibility of OpenAI's published benchmark numbers, with all data loaded from public Azure blob storage URLs. The repository is deprecated as of July 2025 and will not receive updates for new models or benchmarks, though it will continue to host reference implementations for HealthBench, BrowseComp, and SimpleQA.

This is the canonical implementation of OpenAI's Harmony response format, the structured prompt format used by the gpt-oss open-weight model series. The repository provides a high-performance Rust core for rendering conversations into token sequences and parsing model outputs back into structured messages. It supports multi-channel reasoning (analysis, commentary, final), tool/function calling with JSON Schema-defined parameters, structured outputs, and streaming incremental parsing. The library is packaged with first-class Python bindings via PyO3 and a JavaScript/WASM build for browser use. A Next.js demo visualizes the token-level rendering. The codebase demonstrates strong engineering: a custom BPE tokenizer (tiktoken) implementation in Rust, a streaming parser that gracefully handles partial UTF-8 sequences, and a JSON Schema-to-TypeScript converter for tool definitions. Tests are comprehensive with parity between Rust and Python test suites.

ChatKit.js provides TypeScript type definitions and React bindings for the ChatKit web component, which is served from a CDN. The repository does not contain the actual web component implementation but offers ergonomic React integration via the `useChatKit` hook and `ChatKit` component, along with comprehensive type safety. Key functionality includes synchronizing React state with the web component, mapping custom events to React handler props, and providing imperative methods to control the chat UI (e.g., sending messages, managing threads). The library emphasizes stable references for options and event handlers to prevent stale closures in React, ensuring reliable integration with the external web component.

codex-universal is a Docker image that provides a universal development environment with multiple language runtimes and tools. It allows users to configure specific versions of Python, Node.js, Rust, Go, Swift, Ruby, PHP, and Java via environment variables at runtime. The image is built from Ubuntu 24.04 and includes a wide array of development tools. The architecture is monolithic but modular via environment variable-driven configuration. Key entry points are the Docker container's entrypoint script which runs setup_universal.sh and then starts a bash shell. The system relies on external package managers (apt, pyenv, nvm, rustup, etc.) and scripts from external sources for installation.

This repository is a Next.js application demonstrating the use of the OpenAI Realtime API via WebRTC for real-time voice interaction with a 3D solar system scene built in Spline. The application allows users to ask questions about the solar system, moons, the ISS position, and data visualizations, which trigger function calls to update the 3D scene or display charts in the UI. The architecture consists of a frontend built with React and Next.js, featuring components for the 3D scene (using @splinetool/react-spline), controls for managing the WebRTC connection, and a debug log. Two API routes proxy requests to external services: one for fetching the ISS position and another for obtaining ephemeral client secrets from the OpenAI Realtime API. The frontend establishes a WebRTC connection to OpenAI, sends audio input, and processes function call responses to trigger animations in the Spline scene or update the UI with charts. A notable reusable component is the charting system located in `components/charts/chart.tsx`, which provides a set of theme-aware, flexible charting components (Container, Tooltip, Legend) that abstract the complexity of integrating Recharts and Chart.js with Tailwind CSS. This component demonstrates a clean separation of concerns and can be adapted for use in other projects with minimal modification. The rest of the codebase is tailored to the solar system demo but follows modern React and Next.js best practices, including TypeScript strict mode, proper error handling, and modular organization. The application is actively maintained, with recent commits and up-to-date dependencies, and exhibits low technical drift, adhering to current web standards and APIs.