codex

中文版 | English

English

Sandbox runtime for the OpenAI Codex CLI agent, supporting AI-driven code generation and execution.

Environment

This template comes preinstalled with the following tools and dependencies:

Category	Details
Base image	python:3.12-slim
Python	Python 3 + venv (at `/opt/venv`)
Node.js	Node.js 22.x + npm
Python packages	`openai>=1.0`, `rich`, `httpx`
Node packages	`@openai/codex`
System tools	git, curl, wget, build-essential, ripgrep, jq
Resources	2 CPU / 4096 MB memory

Installation

Install from a local path:

abx template install ./codex --registry-type local

Install from GitHub:

abx template install alibaba-agent-sandbox/awesome-templates//codex

Usage

Create a sandbox instance and pass in the API key:

abx create --template codex --env OPENAI_API_KEY=sk-xxx

Run a code-generation task inside the sandbox:

abx exec <sandbox-id> -- codex "Write a quicksort algorithm"

Using the Python SDK:

from agent_sandbox import Sandbox

sandbox = Sandbox.create_sync(
    template="codex",
    env={"OPENAI_API_KEY": "sk-xxx"}
)

result = sandbox.commands.run("codex 'Generate a Flask REST API'")
print(result.stdout)

Configuration

Environment Variables

Variable	Description	Required
`OPENAI_API_KEY`	OpenAI API key	Yes
`CODEX_SANDBOX_MODE`	Sandbox mode marker, defaults to `true`	No
`WORKSPACE`	Working directory path, defaults to `/workspace`	No

Custom Configuration

To install extra dependencies on top of this template, run the following after the sandbox is created:

abx exec <sandbox-id> -- pip install <package-name>
abx exec <sandbox-id> -- npm install -g <package-name>

Notes

API key security: Never hardcode OPENAI_API_KEY into the template; always inject it via the --env option or environment variables.
Network access: The Codex agent needs to reach the OpenAI API — make sure the sandbox has outbound internet connectivity.
Resource usage: AI code-generation tasks may consume significant memory; keeping 4096 MB or more is recommended.
Virtual environment: Python packages are installed in /opt/venv; no manual activation is needed since PATH is preconfigured.

中文

OpenAI Codex CLI Agent 的沙箱运行环境，支持 AI 驱动的代码生成和执行。

环境说明

本模板预装以下工具和依赖：

类别	内容
基础镜像	python:3.12-slim
Python	Python 3 + venv 虚拟环境（位于 `/opt/venv`）
Node.js	Node.js 22.x + npm
Python 包	`openai>=1.0`、`rich`、`httpx`
Node 包	`@openai/codex`
系统工具	git、curl、wget、build-essential、ripgrep、jq
资源配置	2 CPU / 4096 MB 内存

安装方式

从本地安装：

abx template install ./codex --registry-type local

从 GitHub 安装：

abx template install alibaba-agent-sandbox/awesome-templates//codex

使用示例

创建沙箱实例并传入 API Key：

abx create --template codex --env OPENAI_API_KEY=sk-xxx

在沙箱中执行代码生成任务：

abx exec <sandbox-id> -- codex "编写一个快速排序算法"

使用 Python SDK：

from agent_sandbox import Sandbox

sandbox = Sandbox.create_sync(
    template="codex",
    env={"OPENAI_API_KEY": "sk-xxx"}
)

result = sandbox.commands.run("codex '生成一个 Flask REST API'")
print(result.stdout)

配置说明

环境变量

变量名	说明	必填
`OPENAI_API_KEY`	OpenAI API 密钥	是
`CODEX_SANDBOX_MODE`	沙箱模式标识，默认 `true`	否
`WORKSPACE`	工作目录路径，默认 `/workspace`	否

自定义配置

如需在模板基础上安装额外依赖，可在创建沙箱后执行：

abx exec <sandbox-id> -- pip install <package-name>
abx exec <sandbox-id> -- npm install -g <package-name>

注意事项

API Key 安全：请勿将 OPENAI_API_KEY 硬编码到模板中，始终通过 --env 参数或环境变量注入。
网络访问：Codex Agent 需要访问 OpenAI API，确保沙箱具备外网连接能力。
资源消耗：AI 代码生成任务可能消耗较多内存，建议保持 4096 MB 以上的内存配置。
虚拟环境：Python 包安装在 /opt/venv 中，使用时无需手动激活，PATH 已自动配置。

Deploy and local e2e

Install and register this template:

abx template install alibaba-agent-sandbox/awesome-templates//codex

From this directory (build, push, and register):

abx deploy

From the repository root, with agent-sandbox-server and httpx installed:

python3 scripts/run_local_e2e.py

That starts python3 commands.py in this directory and runs e2e_test.py.

Sandbox runtime for the OpenAI Codex CLI…