← Back to Skills Marketplace
Agent Deployment
by
mirowangl-ops
· GitHub ↗
· v1.0.0
· MIT-0
107
Downloads
0
Stars
0
Active Installs
1
Versions
Install in OpenClaw
/install xiaozhua-agent-deployment
Description
AI Agent 部署与管理。在 WSL2/Linux 上安装 OpenClaw，配置 GitHub Copilot 认证，启动 Gateway，初始化 workspace（SOUL.md/USER.md/TASKS.md）。支持多 Agent 部署、端口配置、认证排错。使用场景：部署新 Agent、重启 Age...
README (SKILL.md)
Agent 部署
部署新 Agent（WSL2 + OpenClaw）
前置条件
- WSL2 已安装（需管理员权限：
wsl --install） - Ubuntu 发行版
安装步骤
# 1. 安装 nvm + Node.js（绕过 sudo）
curl -o- https://raw.githubusercontent.com/nvm-sh/nvm/v0.40.3/install.sh | bash
source ~/.nvm/nvm.sh
nvm install 24
# 2. 安装 OpenClaw
npm install -g openclaw
# 3. 初始化
openclaw setup
# 4. GitHub Copilot 认证
openclaw login-github-copilot
# 注意：device code 每次超时就失效，可能需要多试几次
# 5. 配置模型和端口
# 编辑 ~/.openclaw/openclaw.json
# gateway.port: \x3CYOUR_PORT>（避免和主 Agent 冲突）
# model: github-copilot/claude-sonnet-4-20250514
# 6. 启动
openclaw gateway
启动已有 Agent
wsl -d Ubuntu
source ~/.nvm/nvm.sh
openclaw gateway
- 端口：\x3CYOUR_PORT>
- Token：
\x3CYOUR_TOKEN> - 浏览器打开 http://127.0.0.1:\x3CYOUR_PORT>
Workspace 初始化
新 Agent 上线后写入：
- SOUL.md — 身份、职责、性格
- USER.md — Lei 的信息和偏好
- TASKS.md — 任务清单
- COLLABORATION.md — 协作协议
常见问题
- device code 超时 — 每次生成新 code，快速在浏览器完成认证
- 端口冲突 — 改 openclaw.json 的 gateway.port
- config overwrite — Gateway 启动时可能重写配置，启动后检查端口是否正确
- 关机后重启 — 需要手动启动 WSL + Gateway，AI 没有管理员权限自动启动
Usage Guidance
This skill appears to do what it says: install and run OpenClaw on WSL2 and guide you through Copilot authentication. Before running the commands: (1) inspect the remote install script instead of blindly piping to bash; view https://raw.githubusercontent.com/nvm-sh/nvm/v0.40.3/install.sh first. (2) Verify the npm package 'openclaw' is the legitimate project you expect (check the package page or project repo). (3) Note that Copilot device-code authentication will produce tokens—confirm where tokens are stored and revoke them if needed. (4) The model string shown ("github-copilot/claude-sonnet-4-20250514") looks unusual (mix of vendor names); verify that this model identifier is correct for your OpenClaw configuration. If any of these checks fail or sources are unknown, avoid running the installer commands.
Capability Analysis
Type: OpenClaw Skill
Name: xiaozhua-agent-deployment
Version: 1.0.0
The skill bundle provides standard, transparent instructions for deploying and configuring the OpenClaw agent on WSL2/Linux. It includes steps for installing Node.js via NVM, installing the OpenClaw CLI, and performing GitHub Copilot authentication. While it utilizes the 'curl | bash' pattern for NVM installation in SKILL.md, this is the official installation method for that tool and does not indicate malicious intent. No evidence of data exfiltration, unauthorized persistence, or prompt injection was found.
Capability Assessment
Purpose & Capability
The name/description (Agent deployment and management on WSL2/Linux) matches the SKILL.md steps: install nvm/Node.js, install openclaw, run setup, perform GitHub Copilot auth, edit openclaw.json and start the gateway. All requested actions are coherent for deploying an agent.
Instruction Scope
Instructions limit actions to installing runtime (nvm/Node), installing/starting OpenClaw, editing ~/.openclaw/openclaw.json and creating workspace files (SOUL.md/USER.md/TASKS.md). There are no instructions to read or exfiltrate unrelated system files or environment variables.
Install Mechanism
This is an instruction-only skill (no install spec). The commands include piping a script from raw.githubusercontent.com into bash (nvm installer) and a global 'npm install -g openclaw'. raw.githubusercontent.com is a common host, but piping remote scripts and installing global npm packages have inherent risks—users should inspect scripts and confirm the npm package's provenance before running.
Credentials
The skill does not declare or require environment variables or external credentials in metadata. The SKILL.md legitimately requires interactive GitHub Copilot device-code authentication (expected for Copilot integration). There are no unrelated credential requests.
Persistence & Privilege
The skill does not request 'always' presence or any elevated persistent privileges. It documents that the user must manually start WSL and the gateway after reboot; no unexpected permanent modifications are instructed.
How to Use
- Make sure OpenClaw is installed (local or Docker)
- Run the install command in chat:
/install xiaozhua-agent-deployment - After installation, invoke the skill by name or use
/xiaozhua-agent-deployment - Provide required inputs per the skill's parameter spec and get structured output
Version History
v1.0.0
Initial release of agent-deployment.
- 提供详细的 AI Agent 部署与管理流程，适用于 WSL2/Linux 环境。
- 支持 OpenClaw 安装、GitHub Copilot 认证与多 Agent 配置，含端口与模型自定义。
- 包含部署新 Agent 和已有 Agent 的启动指引。
- 详述 workspace 初始化文档要求（SOUL.md、USER.md、TASKS.md、COLLABORATION.md）。
- 收录常见部署/认证/配置问题与解决方案。
Metadata
Frequently Asked Questions
What is Agent Deployment?
AI Agent 部署与管理。在 WSL2/Linux 上安装 OpenClaw，配置 GitHub Copilot 认证，启动 Gateway，初始化 workspace（SOUL.md/USER.md/TASKS.md）。支持多 Agent 部署、端口配置、认证排错。使用场景：部署新 Agent、重启 Age... It is an AI Agent Skill for Claude Code / OpenClaw, with 107 downloads so far.
How do I install Agent Deployment?
Run "/install xiaozhua-agent-deployment" in the OpenClaw or Claude Code chat to install it in one step — no extra setup required.
Is Agent Deployment free?
Yes, Agent Deployment is completely free, licensed under MIT-0. You can download, install and use it at no cost.
Which platforms does Agent Deployment support?
Agent Deployment is cross-platform and runs anywhere OpenClaw / Claude Code is available (cross-platform).
Who created Agent Deployment?
It is built and maintained by mirowangl-ops (@mirowangl-ops); the current version is v1.0.0.
More Skills