claude-code

中文版 | English

English

Sandbox runtime for Anthropic Claude Code, preinstalled with the Claude Code CLI to support AI-driven coding, debugging, and refactoring.

Environment

This template comes preinstalled with the following tools and dependencies:

Installation

Install from a local path:

abx template install ./claude-code --registry-type local

Install from GitHub:

abx template install alibaba-agent-sandbox/awesome-templates//claude-code

Usage

Create a sandbox instance and pass in the API key:

abx create --template claude-code --env ANTHROPIC_API_KEY=sk-ant-xxx

Launch Claude Code inside the sandbox:

abx exec <sandbox-id> -- claude "Write a Python HTTP server"

Run a task in non-interactive mode:

abx exec <sandbox-id> -- claude -p "Refactor this function for better performance" --allowedTools "Edit,Read,Write"

Using the Python SDK:

from agent_sandbox import Sandbox

sandbox = Sandbox.create_sync(
    template="claude-code",
    env={"ANTHROPIC_API_KEY": "sk-ant-xxx"}
)

result = sandbox.commands.run("claude -p 'Generate a REST API scaffold'")
print(result.stdout)

Configuration

Environment Variables

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 ANTHROPIC_API_KEY into the template; always inject it via the --env option or environment variables.
• Network access: Claude Code needs to reach the Anthropic API — make sure the sandbox has outbound internet connectivity.
• Node.js is the core runtime: The Claude Code CLI runs on Node.js; Node.js 22 is a core dependency.
• 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

Anthropic Claude Code 的沙箱运行环境，预装 Claude Code CLI，支持 AI 驱动的代码编写、调试和重构。

环境说明

本模板预装以下工具和依赖：

安装方式

从本地安装：

abx template install ./claude-code --registry-type local

从 GitHub 安装：

abx template install alibaba-agent-sandbox/awesome-templates//claude-code

使用示例

创建沙箱实例并传入 API Key：

abx create --template claude-code --env ANTHROPIC_API_KEY=sk-ant-xxx

在沙箱中启动 Claude Code：

abx exec <sandbox-id> -- claude "编写一个 Python HTTP 服务器"

非交互模式执行任务：

abx exec <sandbox-id> -- claude -p "重构这个函数以提高性能" --allowedTools "Edit,Read,Write"

使用 Python SDK：

from agent_sandbox import Sandbox

sandbox = Sandbox.create_sync(
    template="claude-code",
    env={"ANTHROPIC_API_KEY": "sk-ant-xxx"}
)

result = sandbox.commands.run("claude -p '生成一个 REST API 框架'")
print(result.stdout)

配置说明

环境变量

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 ANTHROPIC_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：Claude Code 需要访问 Anthropic API，确保沙箱具备外网连接能力。
• Node.js 核心：Claude Code CLI 基于 Node.js 运行，Node.js 22 为核心依赖。
• 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//claude-code

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.