OpenClaw 控制技能

SkillCommunication

Lets your agent send tasks to the OpenClaw AI assistant, check its status, and manage its settings.

Then ask your AI: use the OpenClaw 控制技能 skill

Details

Instructions available. Your AI can read the instructions. Execution depends on the setup they require.

Add it as a step in a pipeline you build on Ahel, or connect Ahel to Claude, ChatGPT, Cursor, Claude Code or Codex and ask your AI to use it.

About this skill

Control and manage the OpenClaw AI assistant. Use it to send messages to OpenClaw, check status, manage configuration, and install skills. Use this skill when the user wants to interact with or manage OpenClaw.

What this skill tells your AI

The instructions your AI receives, as published by rtgs2017/nagaagent in skills/openclaw-control/SKILL.md and read by Ahel’s review.

本技能让你能够控制和管理 OpenClaw AI 助手。

核心能力

1. 发送消息给 OpenClaw

通过 Naga 的 Agent Server API 向 OpenClaw 发送任务:

POST /openclaw/send
{
  "message": "你的消息内容",
  "wake_mode": "now"  // 立即执行
}

2. 检查 OpenClaw 状态

  • GET /openclaw/health - 健康检查
  • GET /openclaw/status - 运行状态
  • GET /openclaw/session - 会话信息

3. 管理 Gateway

  • POST /openclaw/gateway/start - 启动
  • POST /openclaw/gateway/stop - 停止
  • POST /openclaw/gateway/restart - 重启
  • GET /openclaw/gateway/status - 状态

4. 配置管理

  • GET /openclaw/config - 获取配置
  • POST /openclaw/config/model - 设置模型
  • POST /openclaw/config/hooks - 配置 Hooks

使用流程

  1. 检查连接: 先调用健康检查确认 OpenClaw 可用
  2. 发送任务: 使用 /openclaw/send 发送消息
  3. 监控结果: 通过 /openclaw/history 查看执行结果

注意事项

  • OpenClaw 默认运行在 http://127.0.0.1:18789
  • 需要配置 Hooks Token 才能发送消息
  • 建议使用免费的 GLM-4.7 模型进行测试

示例

让 OpenClaw 执行任务

{
  "message": "帮我整理桌面上的文件,按类型分类",
  "wake_mode": "now"
}

查询当前状态

调用 GET /openclaw/status 获取 OpenClaw 当前正在做什么。

Signals

GitHub stars
2k
Forks
165
Last commit
Aug 2026
Advanced
Item type
skill
Key
openclaw-control
Source
github.com/rtgs2017/nagaagent