Claude Skill
digitalknk/openclaw-runbook
OpenClaw Runbook 仓库拥有 574 个星标和 51 个分支。这是一个运行手册/操作指南,最后更新于 2026 年,用于系统文档和流程。
概览
仓库信息
安装这个 Skill
git clone https://github.com/digitalknk/openclaw-runbook.gitRegistry 信息
项目简介
OpenClaw Runbook 是一个 GitHub 仓库,拥有 574 个星标和 51 个分支,最后更新于 2026 年 2 月。它似乎是一个运行手册或操作指南,可能与 OpenClaw 项目或系统相关。
Unofficial OpenClaw runbook for running agents day to day without burning money, exposing your gateway, or trusting random automation.
要点
- 运行手册/操作指南格式
- 拥有 574 个星标的流行仓库
- 近期活跃(2026年更新)
- 51 个分支表明社区使用
使用场景
- 系统操作文档
- 事件响应流程
- 管理任务指导
- 团队知识共享
README 摘要
# OpenClaw Runbook > Checked against OpenClaw at commit `5dccba7405` (`2026-05-25`) > Example config based on a working `2026.5.x` setup This repo is a practical runbook for running OpenClaw day to day without burning money, exposing your gateway, or trusting random automation you did not inspect. It is not official OpenClaw documentation. It is not a universal best setup. It is the way I run OpenClaw after enough broken configs, noisy agents, and avoidable risk to want something predictable. ## What this is - A runbook for people who want OpenClaw to run for weeks, not minutes - Opinionated guidance on models, access, memory, automation, and guardrails - Tailscale-first for dashboard/control access - Skeptical of third-party skills by default - Written from the post-honeymoon phase ## What this is not - A beginner replacement for the official docs - A model leaderboard - A ClawHub install guide - A claim that my exact config is right for everyone ## Access stance My setup uses Tailscale as the normal access path, including when I am local. The Gateway stays loopback-bound, and Tailscale handles remote access. If you do not want to sign up for Tailscale, keep OpenClaw local and use a messaging channel such as Telegram for remote use. Avoid public ports and casual LAN exposure unless you know exactly what boundary you are accepting. ## Skills stance I do not recommend blindly installing third-party skills from ClawHub. Use ClawHub as a discovery and source-reading tool. Find an idea, inspect the source, then ask your agent to rebuild a local skill for your setup. That is slower than clicking install, but it is safer than trusting unknown code, broad tool assumptions, hidden behavior, or token-heavy abstractions. The example config keeps `clawhub` disabled o