360 Secure SkillHub
360 安全技能中心

Computer Use Linux

这是Linux专属的计算机使用技能,核心为捆绑的独立MCP服务器,首次启动自动搭建Python虚拟运行环境,完全不依赖本地Claude相关资源。适用于Linux桌面自动化控制场景,通过命令行执行构建、启动指令,运行后提供MCP服务接口实现桌面操作控制。

wimi321
wimi321
热度 55.68
最近更新 2026-05-18 11:34:21
分类:开发依赖:代码执行 / MCP 服务应用场景:自动化脚本 / DevOps

安装技能

复制下面的话术发给你的 Agent,它会通过 secure-skills 检索并安装该技能。

执行命令安装此 skill: npx secure-skills-cli add computer-use-linux
Metadata
name
computer-use-linux
version
0.1.1
description
Top-level Linux computer-use skill with a bundled standalone runtime that bootstraps itself without any local Claude installation, private native modules, or extracted app assets.
tags
["skill","linux","computer-use","automation","mcp"]

Linux Computer-Use Skill

Use this skill when the task needs a portable Linux computer-use skill bundled with its own standalone runtime and MCP server.

What this skill does

  • uses the bundled linux-computer-use-skill project under the installed skill directory
  • builds the standalone MCP server
  • lets the server auto-bootstrap its Python runtime on first launch
  • avoids any dependency on local Claude binaries, .node modules, or extracted app assets
  • stays explicitly Linux-only because the underlying desktop-control backend is Linux-specific

Default bundled project path

After installation, assume the standalone project lives at:

~/.codex/skills/computer-use-linux/project

If the user installed the skill under a custom CODEX_HOME, use that equivalent path instead.

Build

Always build from the bundled project:

cd ~/.codex/skills/computer-use-linux/project
npm install
npm run build

Run

cd ~/.codex/skills/computer-use-linux/project
node dist/cli.js

The first real run will automatically create .runtime/venv and install the public Python dependencies.

Validation notes

  • Version 0.1.1 fixes the shared system-key blocklist and tool capability typings so Linux builds use Linux-specific shortcut protections instead of a broken copied platform branch.
  • This project has been statically validated from macOS for TypeScript build health, bundled project integrity, and Python helper syntax, but still needs end-to-end runtime validation on a real Linux desktop.

Guardrails

  • Treat this host as trusted-local only.
  • Do not tell the user to search their local Claude install for binaries or hidden assets.
  • Be explicit that this runtime is standalone and uses public dependencies only.
  • Mention that the current runtime reports screenshotFiltering: none, so action gating is handled at the MCP layer.