Token导航 LogoToken导航TokenDH.com
运维执行命令clawhub未标认证来源可访问clear审计提醒

oz-cli-remote-nodeOZ CLI remote node CLI

Agent Skill

oz-cli-remote-node 用于辅助前端页面、组件、样式和交互逻辑开发,适合在 OpenClaw 中需要维护前端项目、生成组件或检查界面实现时使用。可结合来源仓库、安装命令和原始 README 继续核验具体用法。安装前建议确认权限范围、维护状态,以及是否会触发联网、命令执行或文件读写。

总安装

3,684

周安装

152

GitHub Stars

公开资料未说明

下载量

1,204
OpenClaw

安装说明

本站只整理中文说明和来源信息,不托管安装包,也不代用户安装。

GitHub

来源数

2

许可证

MIT-0

最后核验

2026-05-01

来源状态

来源可访问

安装方式

通过对话安装

复制提示词发给支持本地命令或 Skills 的 AI 助手,先确认命令和权限,再让它执行。

请帮我安装这个 Agent Skill:oz-cli-remote-node(OZ CLI remote node CLI)
来源仓库:https://github.com/eladrave/oz-cli-remote-node
安装命令:
openclaw skills install oz-cli-remote-node
安装前请先检查当前环境是否支持对应 CLI,并向我确认将要执行的命令、安装目录、联网范围和文件读写权限;确认后再执行。

命令行安装

复制命令到本机终端执行。该命令会通过 OpenClaw 从第三方来源获取 Skill;本站只展示命令,不托管安装包,也不自动执行。

ClawHubOpenClaw
openclaw skills install oz-cli-remote-node

简介

通过 CLI 在 Oz Build Node 上执行命令或任务,并在远程节点上提供基于模式的控制和直接 bash 命令支持。

SKILL.md

oz

Description

This skill implements the /oz slash command and acts as a proxy to the oz-cli on a remote node. Use this skill ONLY when the user activates Oz mode with /oz or explicitly asks to use Oz/Warp.

Onboarding & Setup (First-Time Use)

When a user first installs or attempts to use this skill, guide them through the setup process:

  1. Explain the Purpose: Let the user know this skill allows them to run the oz CLI on a *remote* node via the Gateway, acting as a proxy.
  2. Determine the Remote Node: Ask the user for the exact name of the remote node where oz-cli is installed (e.g., "Build Node"). Save this node name into your workspace's TOOLS.md or a local memory file so you remember it for all future uses.
  3. Verify Tool Permissions: Verify that you (the agent) have the nodes tool enabled. If not, instruct the user to grant it (e.g., openclaw config set agents.list[<your-index>].tools.allow '["nodes"]') and restart the gateway.
  4. Sandbox Detection: Check if you are currently running in a Docker sandbox (e.g., by checking if /.dockerenv exists or using cat /proc/1/cgroup). If you are in a sandbox, inform the user that local oz execution is impossible without a custom developer image.
  5. Select Agent Profile: Once the remote node is known, execute oz agent profile list --output-format json on that node using the nodes tool.

- Note: Ignore standard error warnings (like libEGL warning) and parse the JSON array (e.g., [{"id":"Unsynced","name":"Default"},{"id":"...","name":"Claude"}]). - Show the user the list of profile *names* and ask which one they want to use. - Save the selected profile's id (not the name) to your memory alongside the node name.

State Memory (Oz Mode)

  • If the user types /oz, you MUST reply with a clear confirmation: "✅ Oz mode is now ACTIVE. All messages will be sent to the Oz CLI on the remote node. Type /oz off to exit, or use ! <command> to run a direct bash command."
  • If the user types /oz off, you MUST reply: "❌ Oz mode is now OFF. Returning to normal assistant behavior."
  • While in Oz mode, treat EVERY subsequent message from the user as a prompt for the oz CLI on the remote node, UNLESS the message starts with !.

Execution on Remote Node

You must use the nodes tool with action="run" and node="<Saved_Node_Name>" for all execution here. Do NOT use local exec.

  1. Direct Bash Commands (!):

- If the user message starts with ! (e.g. ! ls -la or ! mkdir test): - Strip the ! and run it as a direct shell command on the remote node using the nodes tool. - Command array format: ["bash", "-c", "<command>"] - Use cwd: ~/git (or the last known working directory if changed). - Report the output directly.

  1. Oz CLI Prompts & Run Tracking:

- CRITICAL: ALWAYS use oz agent run. NEVER use oz agent run-cloud. - Run the oz CLI on the remote node using the nodes tool. - Command array format MUST include the saved profile id: ["oz", "agent", "run", "--cwd", "~/git", "--profile", "<Saved_Profile_ID>", "--prompt", "<user_prompt>"] - Run Tracking: When you start an oz agent run, you MUST keep track of it: 1. Monitor the initial output to capture the unique Run ID and the "Open in Oz URL". 2. Create a tracking file in your workspace named oz_run.<id>.md. 3. Include in this file: Date, Time, Run ID, the Original Prompt, the Open in Oz URL, and any other relevant metadata. 4. Keep track of the oz commands that are currently running. - Completion: Once the oz agent run command is done, update the tracking file and send a response back to the user formatted EXACTLY like this:

     oz agent for run ID: <id> done.
     
     <the response from the oz agent>

Supported oz agent run Flags

You can append these flags to the oz agent run command if explicitly requested by the user:

  • --cwd <PATH> (-C): run from a different directory.
  • --name <NAME> (-n): label the run for grouping and traceability.
  • --share: share the session with teammates.
  • --profile <ID>: use a specific agent profile (defaults to the saved ID if not specified).
  • --model <MODEL_ID>: override the default model.
  • --skill <SPEC>: use a skill as the base prompt.
  • --mcp <SPEC>: start one or more MCP servers (UUID, JSON path, or inline JSON). Can be repeated.
  • --environment <ID> (-e): run in a specific cloud environment.
  • --file <PATH> (-f): load run configuration from a YAML or JSON file.

适合场景

01

OpenClaw 用户查找和安装 Skill 时

02

用户想查找某类 Agent Skill 时

03

需要根据任务场景推荐可安装能力包时

04

需要对比不同来源的安装命令和来源信息时

能力概览

能力 1

按任务关键词查找相关 Skills

能力 2

展示可复制的安装命令

能力 3

保留来源站点、仓库和原始说明,方便继续核验

能力 4

补充不同宿主或平台的使用分布数据

能力 5

展示第三方安全扫描或审计结果

安装后应在对应宿主中按原始 README 的触发条件使用;具体调用方式请以来源页面和 README 为准。

平台分布

OpenClaw

73.56%
按下载量换算886

安全审计

VirusTotal

可疑

ClawScan

可疑

Static analysis

通过

权限和风险

执行命令

安装流程涉及命令执行,可能通过 openclaw skills install oz-cli-remote-node 联网下载 Skill 或依赖。用户安装前应确认命令来源、仓库内容和执行环境。

安装前确认

本站仅展示第三方公开信息,不托管安装包,不提供自动安装或运行环境。安装前应自行审查源码、依赖和命令行为。来源安全扫描存在 warning/failed 结果,不能写成本站确认安全。当前只有一个来源,正式发布前建议补源仓库或其他目录站核验。

来源信息

继续浏览同类 Skills