# 安装 io.actionlayer/actionlayer

- 类型: MCP 工具
- 说明: Concierge for AI agents: takes action in the physical world — book, order, register, resolve.

## 方式一 · 一键代装（复制提示词让 AI 代劳）

### Claude Code
```
请帮我安装 askskill 上的 "io.actionlayer/actionlayer" MCP 服务：
执行：claude mcp add 'io-actionlayer-actionlayer' -- uvx actionlayer
```

### Codex
```
请帮我安装 askskill 上的 "io.actionlayer/actionlayer" MCP 服务：
在 ~/.codex/config.toml 中添加：
[mcp_servers.io-actionlayer-actionlayer]
command = "uvx"
args = ["actionlayer"]
然后重启 Codex 生效
```

### Cursor
```
请帮我安装 askskill 上的 "io.actionlayer/actionlayer" MCP 服务：
在 .cursor/mcp.json 的 "mcpServers" 中添加：
"io-actionlayer-actionlayer": { "command": "uvx", "args": ["actionlayer"] }
然后在 Cursor 设置 > MCP 中启用
```

## 方式二 · 命令行安装（原生命令与配置）

### Claude Code
```bash
claude mcp add --env ACTIONLAYER_API_KEY=YOUR_VALUE --env ACTIONLAYER_API_URL=YOUR_VALUE --transport stdio io-actionlayer-actionlayer -- uvx actionlayer
```
需要环境变量：ACTIONLAYER_API_KEY, ACTIONLAYER_API_URL

### Codex
```bash
codex mcp add --env ACTIONLAYER_API_KEY=YOUR_VALUE --env ACTIONLAYER_API_URL=YOUR_VALUE io-actionlayer-actionlayer -- uvx actionlayer
```
配置文件 ~/.codex/config.toml：
```toml
[mcp_servers.io-actionlayer-actionlayer]
command = "uvx"
args = ["actionlayer"]

[mcp_servers.io-actionlayer-actionlayer.env]
ACTIONLAYER_API_KEY = "YOUR_VALUE"
ACTIONLAYER_API_URL = "YOUR_VALUE"
```
需要环境变量：ACTIONLAYER_API_KEY, ACTIONLAYER_API_URL

### Cursor
配置文件 .cursor/mcp.json：
```json
{
  "mcpServers": {
    "io-actionlayer-actionlayer": {
      "command": "uvx",
      "args": [
        "actionlayer"
      ],
      "env": {
        "ACTIONLAYER_API_KEY": "YOUR_VALUE",
        "ACTIONLAYER_API_URL": "YOUR_VALUE"
      }
    }
  }
}
```
一键安装：cursor://anysphere.cursor-deeplink/mcp/install?name=io-actionlayer-actionlayer&config=eyJjb21tYW5kIjoidXZ4IiwiYXJncyI6WyJhY3Rpb25sYXllciJdLCJlbnYiOnsiQUNUSU9OTEFZRVJfQVBJX0tFWSI6IllPVVJfVkFMVUUiLCJBQ1RJT05MQVlFUl9BUElfVVJMIjoiWU9VUl9WQUxVRSJ9fQ
需要环境变量：ACTIONLAYER_API_KEY, ACTIONLAYER_API_URL