微沙盒 是 AI Skill Hub 本期精选MCP工具之一。已获得 6.3k 颗 GitHub Star,综合评分 8.5 分,整体质量较高。我们强烈推荐将其纳入你的 AI 工具库,帮助提升工作效率。
微沙盒 是一款遵循 MCP(Model Context Protocol)标准协议的 AI 工具扩展。通过 MCP 协议,它可以让 Claude、Cursor 等主流 AI 客户端直接访问和操作外部工具、数据源和服务,实现 AI 能力的无缝扩展。无论是文件操作、数据库查询还是 API 调用,都可以通过自然语言在 AI 对话中直接触发,极大提升生产效率。
微沙盒 是一款遵循 MCP(Model Context Protocol)标准协议的 AI 工具扩展。通过 MCP 协议,它可以让 Claude、Cursor 等主流 AI 客户端直接访问和操作外部工具、数据源和服务,实现 AI 能力的无缝扩展。无论是文件操作、数据库查询还是 API 调用,都可以通过自然语言在 AI 对话中直接触发,极大提升生产效率。
# 方式一:通过 Claude Code CLI 一键安装
claude skill install https://github.com/superradcompany/microsandbox
# 方式二:手动配置 claude_desktop_config.json
{
"mcpServers": {
"---": {
"command": "npx",
"args": ["-y", "microsandbox"]
}
}
}
# 配置文件位置
# macOS: ~/Library/Application Support/Claude/claude_desktop_config.json
# Windows: %APPDATA%/Claude/claude_desktop_config.json
# 安装后在 Claude 对话中直接使用 # 示例: 用户: 请帮我用 微沙盒 执行以下任务... Claude: [自动调用 微沙盒 MCP 工具处理请求] # 查看可用工具列表 # 在 Claude 中输入:"列出所有可用的 MCP 工具"
// claude_desktop_config.json 配置示例
{
"mcpServers": {
"___": {
"command": "npx",
"args": ["-y", "microsandbox"],
"env": {
// "API_KEY": "your-api-key-here"
}
}
}
}
// 保存后重启 Claude Desktop 生效
<br />
<br /> <br />
<br />
Microsandbox runs untrusted workloads inside fast, local microVMs: AI agents, user code, plugins, CI jobs, dev environments, scrapers, and automation.
##
<br />
The SDK lets you create and control sandboxes directly from your application. Sandbox.builder("...").create() boots a microVM as a child process. No infrastructure required.
<details> <summary><b> Rust Example →</b></summary>> import { Sandbox } from "microsandbox"; > > await using sandbox = await Sandbox.builder("my-sandbox") > .image("python") > .cpus(1) > .memory(512) > .create(); > > const output = await sandbox.exec("python", [ > "-c", > "print('Hello from a microVM!')", > ]); > > console.log(output.stdout()); ></details> <details> <summary><b> Python Example →</b></summary>> use microsandbox::Sandbox; > > #[tokio::main] > async fn main() -> Result<(), Box<dyn std::error::Error>> { > let sandbox = Sandbox::builder("my-sandbox") > .image("python") > .cpus(1) > .memory(512) > .create() > .await?; > > let output = sandbox > .exec("python", ["-c", "print('Hello from a microVM!')"]) > .await?; > > println!("{}", output.stdout()?); > > sandbox.stop().await?; > > Ok(()) > } ></details> <details> <summary><b> Ruby Example →</b></summary>> import asyncio > from microsandbox import Sandbox > > async def main(): > sandbox = await Sandbox.create( > "my-sandbox", > image="python", > cpus=1, > memory=512, > ) > > output = await sandbox.exec("python", ["-c", "print('Hello from a microVM!')"]) > > print(output.stdout_text) > > await sandbox.stop() > > asyncio.run(main()) >See the Ruby SDK guide for installation, lifecycle, networking, and backend details. </details> <details> <summary><b> Go Example →</b></summary>> require "microsandbox" > > sandbox = Microsandbox::Sandbox.create( > "my-sandbox", > image: "python", > cpus: 1, > memory: 512, > network: { > allowed_hosts: ["api.openai.com"], > allowed_ports: [443] > }, > secrets: [{ > env: "OPENAI_API_KEY", > value: ENV.fetch("OPENAI_API_KEY"), > allowed_host: "api.openai.com" > }] > ) > > output = sandbox.exec("python", ["-c", "print('Hello from a microVM!')"]) > puts output.stdout > > sandbox.stop ></details>> package main > > import ( > "context" > "fmt" > "log" > > microsandbox "github.com/superradcompany/microsandbox/sdk/go" > ) > > func main() { > ctx := context.Background() > > // Downloads the microsandbox runtime to ~/.microsandbox/ on first run. > if err := microsandbox.EnsureInstalled(ctx); err != nil { > log.Fatal(err) > } > > sandbox, err := microsandbox.CreateSandbox(ctx, "my-sandbox", > microsandbox.WithImage("python"), > microsandbox.WithCPUs(1), > microsandbox.WithMemory(512), > ) > if err != nil { > log.Fatal(err) > } > defer sandbox.Stop(ctx) > > output, err := sandbox.Exec(ctx, "python", []string{"-c", "print('Hello from a microVM!')"}) > if err != nil { > log.Fatal(err) > } > > fmt.Println(output.Stdout()) > } >
The first call to create() pulls the image if it isn't cached locally, so it may take longer depending on your connection. Subsequent runs reuse the cache.
<br />
<a href="https://docs.microsandbox.dev/sdk/overview"><img src="https://img.shields.io/badge/SDK_Docs-%E2%86%92-A770EF?style=flat-square&labelColor=2b2b2b" alt="SDK Docs"></a>
<br />
#### <img height="14" src="https://octicons-col.vercel.app/move-to-bottom/A770EF"> Install the SDK >
> npm i microsandbox # 🟦 TypeScript
> > > > cargo add microsandbox # 🦀 Rust
> > > > uv add microsandbox # 🐍 Python
> > > > go get github.com/superradcompany/microsandbox/sdk/go # 🐹 Go
> #### <img height="14" src="https://octicons-col.vercel.app/download/A770EF"> Install the CLI
Boot a microVM in a single command:## Or install the> npx microsandbox run debian >msbcommand globally:> curl -fsSL https://install.microsandbox.dev | sh # 🍎 macOS / 🐧 Linux ><details> <summary><em> We also support other package managers →</em></summary> ##> irm https://install.microsandbox.dev/windows | iex # 🪟 Windows >> brew install superradcompany/tap/microsandbox >> npm i -g microsandbox >> uv tool install microsandbox ></details> ## Then you can run> cargo install microsandbox >msbdirectly:> msb run debian >
##
Requirements: - <img height="14" src="https://api.iconify.design/simple-icons:apple.svg?color=%23A770EF" alt="macOS"> macOS: Apple Silicon. - <img height="14" src="https://api.iconify.design/simple-icons:linux.svg?color=%23A770EF" alt="Linux"> Linux: KVM enabled. - <img height="14" src="https://api.iconify.design/simple-icons:windows.svg?color=%23A770EF" alt="Windows"> Windows: WHP enabled. Warning: Microsandbox is still beta software. Expect breaking changes, missing features, and rough edges.
<br />
Practical ways to put microsandbox to work:
• <img height="14" src="https://octicons-col.vercel.app/container/A770EF"> Docker in a Sandbox: Run Docker without touching the host daemon.<br /> • <img height="14" src="https://octicons-col.vercel.app/code/A770EF"> OpenCode: Give a coding agent an isolated project workspace.<br /> • <img height="14" src="https://octicons-col.vercel.app/globe/A770EF"> Playwright: Run headless browser jobs inside a microVM.<br /> • <img height="14" src="https://octicons-col.vercel.app/cache/A770EF"> Warm Workers: Snapshot a toolchain and launch clean workers.<br /> • <img height="14" src="https://octicons-col.vercel.app/database/A770EF"> Migration Rehearsal: Test a database migration, then restore the baseline.<br /> • <img height="14" src="https://octicons-col.vercel.app/workflow/A770EF"> GitHub Actions Runner: Run each self-hosted job in a disposable microVM.<br /> • <img height="14" src="https://octicons-col.vercel.app/file/A770EF"> Documents to PDF: Convert untrusted documents in a fresh offline worker.
<br />
<a href="https://docs.microsandbox.dev/examples/overview"><img src="https://img.shields.io/badge/Browse_Examples-%E2%86%92-A770EF?style=flat-square&labelColor=2b2b2b" alt="Browse Examples"></a>
<br />
• <img height="14" src="https://octicons-col.vercel.app/workflow/A770EF"> Eve by Vercel: Agent framework that ships microsandbox as a sandbox backend.<br /> • <img height="14" src="https://octicons-col.vercel.app/organization/A770EF"> Agentic Coding Quickstart by U.S. GSA: From zero to a running AI coding agent with USAi in minutes.<br /> • <img height="14" src="https://octicons-col.vercel.app/package/A770EF"> Condukt and Once by Tuist: Elixir agentic engine, and cacheable actions that run in fresh sandboxes.<br /> • <img height="14" src="https://octicons-col.vercel.app/link/A770EF"> langchain-microsandbox by kenwoodjw: Microsandbox integration for LangChain Deep Agents.<br /> • <img height="14" src="https://octicons-col.vercel.app/history/A770EF"> Smithers by Smithers: Agent workflows with full observability, rewind, fork, and replay.<br /> • <img height="14" src="https://octicons-col.vercel.app/terminal/A770EF"> wrap by Tobi Lütke: Run coding agents and project commands in isolated Arch Linux microVMs.<br /> • <img height="14" src="https://octicons-col.vercel.app/shield-lock/A770EF"> Agent VM by Wiren Board: Run AI agents in safe VMs scoped to a local folder.
• <img height="14" src="https://octicons-col.vercel.app/browser/A770EF"> h5i by h5i: Secure, auditable browser for AI agents, written in pure Rust.<br /> • <img height="14" src="https://octicons-col.vercel.app/cloud/A770EF"> Devsy by Devsy: Deploy devcontainers onto any cloud, Kubernetes cluster, or Docker host.<br /> • <img height="14" src="https://octicons-col.vercel.app/briefcase/A770EF"> OpenWork by Different AI: Open-source Claude Cowork alternative with a microsandbox image.
• <img height="14" src="https://octicons-col.vercel.app/list-unordered/A770EF"> Awesome Microsandbox by ya-luotao: Curated list of SDKs, integrations, tools, and resources.<br /> • <img height="14" src="https://octicons-col.vercel.app/device-desktop/A770EF"> msb-omarchy by ya-luotao: Omarchy desktop with graphics inside a microVM on Apple Silicon.
<br />
<a href="https://discord.gg/T95Y3XnEAK"><img src="https://img.shields.io/badge/Share_a_Project-%E2%86%92-A770EF?style=flat-square&labelColor=2b2b2b" alt="Share a Project"></a>
<br />
microsandbox是一个高质量的AI沙盒工具,具有安全和可编程性
AI Skill Hub 为第三方内容聚合平台,本页面信息基于公开数据整理,不对工具功能和质量作任何法律背书。
建议在沙箱或测试环境中充分验证后,再部署至生产环境,并做好必要的安全评估。
✅ Apache 2.0 — 宽松开源协议,可商用,需保留版权声明和 NOTICE 文件,含专利授权条款。
经综合评估,微沙盒 在MCP工具赛道中表现稳健,质量优秀。如果你已有明确的使用需求,可以直接上手体验;如果还在评估阶段,建议对比同类工具后再做决策。
| 原始名称 | microsandbox |
| 原始描述 | 开源MCP工具:🧱 secure, local and programmable sandboxes for AI agents。⭐6.3k · Rust |
| Topics | AI沙盒Rust容器 |
| GitHub | https://github.com/superradcompany/microsandbox |
| License | Apache-2.0 |
| 语言 | Rust |
收录时间:2026-05-29 · 更新时间:2026-05-30 · License:Apache-2.0 · AI Skill Hub 不对第三方内容的准确性作法律背书。
选择 Agent 类型,复制安装指令后粘贴到对应客户端