### Install Dependencies and Run Web Demo

Source: https://developers.openai.com/cookbook/examples/building_w_rt_mini/building_w_rt_mini

Installs necessary dependencies and starts the Realtime Agents Next.js web demo.

```bash
npm install @openai/agents zod@3
pnpm examples:realtime-next
```

--------------------------------

### Install and Start Supabase Local Backend

Source: https://developers.openai.com/cookbook/examples/voice_solutions/running_realtime_api_speech_on_esp32_arduino_edge_runtime_elatoai

Installs the Supabase CLI and starts a local Supabase server with default migrations and seed data.

```bash
brew install supabase/tap/supabase
supabase start # Starts your local Supabase server with the default migrations and seed data.
```

--------------------------------

### windowsSandbox/setupStart

Source: https://developers.openai.com/codex/app-server

Start Windows sandbox setup for `elevated` or `unelevated` mode; returns quickly and later emits `windowsSandbox/setupCompleted`.

```APIDOC
## windowsSandbox/setupStart

### Description
Start Windows sandbox setup for `elevated` or `unelevated` mode.

### Request Body
- **mode** (string) - Required - The setup mode: `"elevated"` or `"unelevated"`.
```

--------------------------------

### Example Output of SDK Installation

Source: https://developers.openai.com/cookbook/examples/gpt-5/gpt-5_new_params_and_tools

Shows the expected output after successfully installing or upgrading the OpenAI and Pandas libraries. Verify these versions or newer for compatibility.

```text
openai 1.99.2
pandas 2.3.1
```

--------------------------------

### Install Dependencies and Run Next.js Development Server

Source: https://developers.openai.com/cookbook/examples/voice_solutions/running_realtime_api_speech_on_esp32_arduino_edge_runtime_elatoai

Navigates to the frontend directory, installs npm dependencies, and starts the Next.js development server.

```bash
cd frontend-nextjs
npm install

# Run the development server
npm run dev
```

--------------------------------

### Install or Upgrade Dependencies

Source: https://developers.openai.com/cookbook/examples/context_summarization_with_realtime_api

Installs or upgrades the necessary Python packages for the Realtime API example. Run this once to ensure all dependencies are met.

```python
# Run once to install or upgrade dependencies (comment out if already installed)
# !pip install --upgrade openai websockets sounddevice simpleaudio

```

--------------------------------

### Example Goal: Prototype Creation

Source: https://developers.openai.com/codex/use-cases/follow-goals

An example of using `/goal` to guide Codex in creating a prototype based on a `PLAN.md` file, with testing and verification.

```CLI
/goal Implement PLAN.md, creating tests for each milestone and verifying the output with playwright interactive. [include reference screens as needed]
```

--------------------------------

### Start React Chat UI Development Server

Source: https://developers.openai.com/cookbook/examples/mcp/databricks_mcp_cookbook

Instructions to navigate to the UI directory, install dependencies, and start the development server for the React chat application.

```bash
cd ui
npm install
npm run dev
```

--------------------------------

### Install dependencies and tools with setup scripts

Source: https://developers.openai.com/codex/cloud/environments

Use setup scripts to install specific tools like type checkers and to manage project dependencies using package managers like poetry, pnpm, or pip. Ensure environment variables are configured in settings or ~/.bashrc to persist across agent phases.

```bash
# Install type checker
pip install pyright

# Install dependencies
poetry install --with test
pnpm install
```

--------------------------------

### Form Upload and Server Start Output

Source: https://developers.openai.com/cookbook/examples/agents_sdk/computer_use_with_daytona/computer_use_with_daytona

Example output confirming the successful upload of the form and the start of the HTTP server within the sandbox.

```text
Form uploaded to /home/daytona/form/index.html
HTTP server started on port 8080

```

--------------------------------

### Install Dependencies and Build TypeScript Project

Source: https://developers.openai.com/codex/app/local-environments

Use this setup script to install dependencies and perform an initial build for a TypeScript project when a new worktree is created.

```bash
npm install
npm run build
```

--------------------------------

### Install Required Python Packages

Source: https://developers.openai.com/cookbook/examples/fine-tuned_qa/reinforcement_finetuning_healthbench

Install the necessary Python libraries, including `openai`, `evals`, `matplotlib`, `tqdm`, and `rich`, for running the fine-tuning example.

```python
%pip install openai evals matplotlib tqdm rich --upgrade --quiet
```

--------------------------------

### JavaScript Repo Example Execution Command

Source: https://developers.openai.com/blog/skills-agents-sdk

This command is used in the JavaScript repository to start all examples in auto mode, supporting per-example logging and rerun functionality.

```bash
pnpm examples:start-all
```

--------------------------------

### Clone OpenAI Agents JS Repository

Source: https://developers.openai.com/cookbook/examples/building_w_rt_mini/building_w_rt_mini

Clones the `openai-agents-js` repository to get started with the example application.

```bash
git clone https://github.com/openai/openai-agents-js/tree/main
```

--------------------------------

### Run LiveKit Realtime Translation Demo

Source: https://developers.openai.com/cookbook/examples/voice_solutions/realtime_translation_guide

Commands to navigate to the demo project, install dependencies, and start the development server for the LiveKit real-time translation example.

```Shell
cd examples/voice_solutions/realtime_translation_guide/livekit-translation-demo
pnpm install
pnpm dev
```

--------------------------------

### Get OpenAI SDK File Path in Python

Source: https://developers.openai.com/cookbook/examples/evaluation/use-cases/responses-evaluation

Retrieves the file path of the installed OpenAI SDK using the `os` module, useful for locating example code or internal files within the SDK.

```python
openai_sdk_file_path = os.path.dirname(openai.__file__)
```

--------------------------------

### Setup Paths and Directories

Source: https://developers.openai.com/cookbook/examples/codex/build_iterative_repair_loops_with_codex

Configures necessary paths for example directories and temporary run outputs. It also includes error handling for missing data directories and ensures the runs directory is created.

```python
import concurrent.futures
import json
import os
import shlex
import shutil
import subprocess
import tempfile
from pathlib import Path
from typing import Any

CANDIDATE_EXAMPLE_DIRS = [Path("."), Path("examples/codex")]
EXAMPLE_DIR = next((base for base in CANDIDATE_EXAMPLE_DIRS if (base / "data" / "docs").exists()), None)

if EXAMPLE_DIR is None:
    raise RuntimeError(
        "This notebook needs its companion sample notebooks. "
        "Download the data folder that ships with this example and place it next to "
        "this notebook as ./data/docs, or run from a checkout where examples/codex/data/docs exists."
    )

DATA_DIR = EXAMPLE_DIR / "data" / "docs"
DEFAULT_RUNS_DIR = Path(tempfile.gettempdir()) / "codex_iterative_repair_loop_outputs"
RUNS_DIR = Path(os.getenv("CODEX_REPAIR_RUNS_DIR", str(DEFAULT_RUNS_DIR))).expanduser()
RUNS_DIR.mkdir(parents=True, exist_ok=True)
```

--------------------------------

### windowsSandbox/setupStart

Source: https://developers.openai.com/codex/app-server

Triggers the asynchronous setup process for the Windows sandbox, allowing custom clients to avoid blocking on startup checks.

```APIDOC
## windowsSandbox/setupStart

### Description
Triggers the asynchronous setup process for the Windows sandbox, allowing custom clients to avoid blocking on startup checks.

### Method
windowsSandbox/setupStart

### Parameters
#### Request Body
- **mode** (string) - Required - The setup mode to run. Can be `elevated` (run elevated setup path) or `unelevated` (run legacy setup/preflight path).

### Request Example
{
  "method": "windowsSandbox/setupStart",
  "id": 53,
  "params": { "mode": "elevated" }
}

### Response
#### Success Response (200)
- **started** (boolean) - Indicates if the setup process has successfully started.

#### Response Example
{
  "id": 53,
  "result": { "started": true }
}
```

--------------------------------

### Example: Add Context7 MCP Server via CLI

Source: https://developers.openai.com/codex/mcp

This example demonstrates adding the Context7 MCP server, a free server for developer documentation, using `npx`.

```bash
codex mcp add context7 -- npx -y @upstash/context7-mcp
```

--------------------------------

### Install Python Dependencies

Source: https://developers.openai.com/cookbook/examples/agents_sdk/agent_improvement_loop

Installs the necessary Python packages for the agent improvement loop example. Ensure you have Node.js and npx installed for Promptfoo.

```bash
python -m venv .venv
source .venv/bin/activate
pip install openai openai-agents halo-engine
```

--------------------------------

### Example Output: Vector Store Creation and Upload Start

Source: https://developers.openai.com/cookbook/examples/file_search_responses

Shows the console output after successfully creating a vector store and before the parallel PDF upload process begins, including the store ID and initial file count.

```text
Vector store created: {'id': 'vs_67d06b9b9a9c8191bafd456cf2364ce3', 'name': 'openai_blog_store', 'created_at': 1741712283, 'file_count': 0}
21 PDF files to process. Uploading in parallel...

```

--------------------------------

### Install Python Dependencies

Source: https://developers.openai.com/cookbook/examples/agents_sdk/computer_use_with_daytona/computer_use_with_daytona

Install the required Python packages listed in the requirements.txt file for this example using pip.

```bash
%pip install -r requirements.txt --quiet
```

--------------------------------

### Install Required Python Dependencies

Source: https://developers.openai.com/cookbook/examples/multimodal/document_and_multimodal_understanding_tips

Installs or upgrades the `openai` and `pillow` Python packages, necessary for running the notebook examples.

```bash
pip install --upgrade openai pillow
```

--------------------------------

### Setup and Configuration

Source: https://developers.openai.com/cookbook/examples/agents_sdk/agent_improvement_loop

Initializes the OpenAI client, sets up artifact directories, and defines model configurations. Ensures API keys and necessary tools are available.

```python
from __future__ import annotations

import asyncio
import hashlib
import json
import os
import re
import shutil
import subprocess
import sys
import tempfile
import time
import textwrap
import threading
from contextlib import contextmanager
from dataclasses import asdict, dataclass, field
from datetime import datetime, timezone
from importlib.metadata import version
from pathlib import Path
from typing import Any, Iterable, Iterator, Mapping

from IPython.display import Markdown, display
from openai import OpenAI

def find_project_root(start: Path | None = None) -> Path:
    current = (start or Path.cwd()).resolve()
    for candidate in [current, *current.parents]:
        if (candidate / "registry.yaml").exists():
            return candidate
    return current


PROJECT_ROOT = find_project_root()

if not os.getenv("OPENAI_API_KEY"):
    raise RuntimeError("Set OPENAI_API_KEY before running this live notebook.")
if shutil.which("npx") is None:
    raise RuntimeError("Install Node.js with npx before running the Promptfoo eval gate.")

# Edit these in one place if you want to use lower-cost models for part of the loop.
AGENT_MODEL = os.getenv("OPENAI_AGENT_MODEL", "gpt-5.5")
ANALYSIS_MODEL = os.getenv("OPENAI_ANALYSIS_MODEL", "gpt-5.5")
EVAL_GENERATION_MODEL = os.getenv("OPENAI_EVAL_GENERATION_MODEL", ANALYSIS_MODEL)
JUDGE_MODEL = os.getenv("OPENAI_JUDGE_MODEL", ANALYSIS_MODEL)
HALO_MODEL = os.getenv("OPENAI_HALO_MODEL", ANALYSIS_MODEL)
PROMPTFOO_VERSION = os.getenv("PROMPTFOO_VERSION", "0.121.9")

client = OpenAI()


def format_duration(seconds: float) -> str:
    minutes, remainder = divmod(int(round(seconds)), 60)
    return f"{minutes}m {remainder:02d}s" if minutes else f"{remainder}s"


ARTIFACT_DIR = PROJECT_ROOT / "examples" / "agents_sdk" / "agent_improvement_loop_artifacts"
TRACE_DIR = ARTIFACT_DIR / "traces"
HALO_TRACE_PATH = ARTIFACT_DIR / "halo_traces" / "traces.jsonl"
if ARTIFACT_DIR.exists():
    shutil.rmtree(ARTIFACT_DIR)
ARTIFACT_DIR.mkdir(exist_ok=True)
TRACE_DIR.mkdir(exist_ok=True)
HALO_TRACE_PATH.parent.mkdir(exist_ok=True)

print("Project root detected.")
print("Models:", {
    "agent": AGENT_MODEL,
    "analysis": ANALYSIS_MODEL,
    "eval_generation": EVAL_GENERATION_MODEL,
    "judge": JUDGE_MODEL,
    "halo": HALO_MODEL,
    "promptfoo": PROMPTFOO_VERSION,
})
```

```text
Project root detected.
Models: {'agent': 'gpt-5.5', 'analysis': 'gpt-5.5', 'eval_generation': 'gpt-5.5', 'judge': 'gpt-5.5', 'halo': 'gpt-5.5', 'promptfoo': '0.121.9'}
```

--------------------------------

### Install OpenAI Library and Initialize Client

Source: https://developers.openai.com/cookbook/examples/structured_outputs_intro

Install the necessary Python library and set up the OpenAI client with the specified model for API calls.

```bash
%pip install openai -U
```

```python
import json
from textwrap import dedent
from openai import OpenAI
client = OpenAI()
```

```python
MODEL = "gpt-4o-2024-08-06"
```

--------------------------------

### Install or Update Codex using Homebrew

Source: https://developers.openai.com/cookbook/examples/codex/using_goals_in_codex

Use these commands to install or update the Codex CLI via Homebrew and verify the installed version. Goals are available starting in Codex 0.128.0.

```bash
brew update
brew upgrade --cask codex
codex --version
```

--------------------------------

### Initiate Windows Sandbox Setup with windowsSandbox/setupStart (JSON-RPC)

Source: https://developers.openai.com/codex/app-server

Asynchronously trigger the Windows sandbox setup process, specifying the desired mode. A "windowsSandbox/setupCompleted" notification will be emitted upon completion.

```json
{ "method": "windowsSandbox/setupStart", "id": 53, "params": { "mode": "elevated" } }
{ "id": 53, "result": { "started": true } }
```

```json
{
  "method": "windowsSandbox/setupCompleted",
  "params": { "mode": "elevated", "success": true, "error": null }
}
```

--------------------------------

### Install or Update Codex using npm

Source: https://developers.openai.com/cookbook/examples/codex/using_goals_in_codex

Use these commands to install or update the Codex CLI globally via npm and verify the installed version. Goals are available starting in Codex 0.128.0.

```bash
npm install -g @openai/codex@latest
codex --version
```

--------------------------------

### Example `evals/setup-demo-app.prompts.csv` for Skill Evaluation

Source: https://developers.openai.com/blog/eval-skills

This CSV defines a small set of prompts for evaluating the `setup-demo-app` skill, including cases for explicit, implicit, contextual invocation, and a negative control to catch false positives.

```csv
id,should_trigger,prompt
test-01,true,"Create a demo app named `devday-demo` using the $setup-demo-app skill"
test-02,true,"Set up a minimal React demo app with Tailwind for quick UI experiments"
test-03,true,"Create a small demo app to showcase the Responses API"
test-04,false,"Add Tailwind styling to my existing React app"
```

--------------------------------

### Set up working directory and API key

Source: https://developers.openai.com/codex/guides/agents-sdk

Create a new project directory and securely store your OpenAI API key in a `.env` file for local development.

```bash
mkdir codex-workflows
cd codex-workflows
printf "OPENAI_API_KEY=sk-..." > .env
```

--------------------------------

### Prompt Guidelines and Examples for Reservation Tool

Source: https://developers.openai.com/cookbook/examples/gpt-5/gpt-5-1_prompting_guide

This snippet provides prompt instructions for using the `create_reservation` tool, including when to call it, how to handle missing information, and example interactions demonstrating tool calls and assistant responses.

```text
<reservation_tool_usage_rules>
- When the user asks to book, reserve, or schedule a table, you MUST call `create_reservation`.
- Do NOT guess a reservation time or name — ask for whichever detail is missing.
- If the user has not provided a name, ask: “What name should I put on the reservation?”
- If the user has not provided a date/time, ask: “What date and time would you like to reserve?”
- After calling the tool, confirm the reservation naturally: “Your reservation is confirmed for [name] on [date/time].”
</tool_usage_rules>

<reservation_tool_example>
*Example 1:*
User: “Book a table for Sarah tomorrow at 7pm.”
Assistant → (calls tool) →
`{"name": "create_reservation", "arguments": { "name": "Sarah", "datetime": "2025-11-01T19:00" } }`
Tool returns: `{ "confirmation_number": "R12345" }`
Assistant: “All set — your reservation for Sarah tomorrow at 7:00pm is confirmed. Your confirmation number is R12345.”

**Example 2:**
User: “I want to make a reservation.”
Assistant: “Sure! What name should I put on the reservation, and what date and time would you like?”

*Example 3:*
User: “Reserve a table under Daniel at 6 tonight.”
Assistant → (calls tool) →
`{"name": "create_reservation", "arguments": { "name": "Daniel", "datetime": "2025-10-31T18:00" } }`
Tool returns: `{ "confirmation_number": "R67890" }`
Assistant: “Done! Your reservation for Daniel at 6:00pm tonight is confirmed. The confirmation number is R67890.”
</reservation_tool_example>
```

--------------------------------

### Install Required Python Packages

Source: https://developers.openai.com/cookbook/examples/vector_databases/tair/getting_started_with_tair_and_openai

Install the necessary Python libraries, including openai, redis, tair, pandas, and wget, to run the examples in this notebook.

```bash
! pip install openai redis tair pandas wget
```

--------------------------------

### Install Python Libraries for Speech Transcription

Source: https://developers.openai.com/cookbook/examples/speech_transcription_methods

Installs required packages like `openai`, `websockets`, and audio processing tools to run the speech-to-text examples.

```bash
!pip install --upgrade -q openai openai-agents websockets sounddevice pyaudio nest_asyncio resampy httpx websocket-client

```

--------------------------------

### Initialize shadcn/ui Project

Source: https://developers.openai.com/cookbook/examples/build_a_coding_agent_with_gpt-5.1

Initializes shadcn/ui in the current project, setting up configuration files.

```bash
npx shadcn@latest init
```

--------------------------------

### Install Node SDK and MCP Apps Helpers

Source: https://developers.openai.com/apps-sdk/quickstart

Use this command to install the necessary Node SDK, MCP Apps helpers, and Zod for building an MCP server.

```bash
npm install @modelcontextprotocol/sdk @modelcontextprotocol/ext-apps zod
```