Skip to main content

zeroone-openstack

Inspect and operate ZeroOne OpenStack resources using the local cloud configuration. Use for flavors, quotas, worker capacity, networking, floating IPs, and OpenTofu deployment questions.

来源信息

仓库
ilude/dotfiles
最近来源活动
2026年10月25日 14:46
检测到的 SKILL.md 语言
英语
星标
4
Forks
1

安装方式

默认使用会先检查来源的 Prompt;你也可以切换为直接命令,或下载本地副本。

检查来源文件

决定是否安装前,请先阅读 SKILL.md,以及 SkillsMP 当前展示的配套文件。

文件资源管理器
3 个文件

正在显示 SKILL.md

SKILL.md
来源说明 · 只读预览
name
zeroone-openstack
description
Inspect and operate ZeroOne OpenStack resources using the local cloud configuration. Use for flavors, quotas, worker capacity, networking, floating IPs, and OpenTofu deployment questions.
# ZeroOne OpenStack Read [local access and commands](references/local-access.md) for the working WSL CLI setup and read-only discovery commands. Read [networking](references/networking.md) when evaluating public access or floating-IP port forwarding. ## Workflow - Use the configured OpenStack API for account-specific facts. Discover flavors, quota usage, and network capabilities rather than inferring them from generic OpenStack documentation or flavor names. - Establish the selected cloud, region, and project context before interpreting results. Keep credentials in the external cloud configuration; do not print or copy its contents into chat, scripts, or version-controlled files. - Distinguish quota, current allocation, physical capacity, and price. Flavor visibility and quota headroom do not establish that a launch will succeed or what it costs. Read-only inventory does not authorize billable resource creation. - For deployment changes, inspect the repository's current OpenTofu configuration and state ownership. Preserve existing changes and use a fresh plan to identify the actual resource changes before applying an authorized deployment. - Treat workers as shared hosts capable of running multiple game services. Keep service lifecycles, persistent data, ports, and resource settings independent. Adding a game does not imply stopping another game or removing its ingress. Assess aggregate workloads and host headroom when sizing; configured memory limits are ceilings, not actual consumption. - Keep build/configuration work separate from live infrastructure changes. A blocked deployment choice need not block independent server definitions or client-pack artifacts. - Report observations with their scope and date when retaining them. Store changing addresses, resource IDs, quotas, and deployment status in current project configuration/documentation, not as permanent skill rules.
在 GitHub 查看