跳到技能内容
词映 AI导航
发现工具选型指南Skills 技能AI 快讯AI图片AI视频AI文档AI办公AI助手AI搜索AI编程AI音频AI设计AI智能体模型与开发AI学习模型评测AI提示词内容检测AI金融AI 3D扫描与识别其他工具跨境电商 AI
词映 AI导航
工具库Skills 技能
我的收藏
词映 AI导航
发现工具选型指南Skills 技能AI 快讯AI图片AI视频AI文档AI办公AI助手AI搜索AI编程AI音频AI设计AI智能体模型与开发AI学习模型评测AI提示词内容检测AI金融AI 3D扫描与识别其他工具跨境电商 AI
Skills directory/jeremylongshore/claude-code-plugins-plus-skills/optimizing-cloud-costs

Claude Code Cloud Usage Skill for Reviewing Idle Capacity

Compare billing categories with workload timing to identify which resources merit a closer utilization review.

中文

Install in Claude Code

The offline exercise uses a synthetic billing summary and utilization table. Live data collection in the upstream skill requires an authenticated cloud CLI with read-only billing and monitoring access.

  1. Get the complete `plugins/devops/cloud-cost-optimizer/skills/optimizing-cloud-costs` directory from the pinned GitHub checkout.
  2. Copy it under the target project's `.claude/skills/`. Confirm `.claude/skills/optimizing-cloud-costs/SKILL.md`, open Claude Code at the project root, and call `/optimizing-cloud-costs`.

Example task

Two test instances average four percent CPU. One has a scheduled daily peak and the other remains nearly flat. Explain which needs review first and what context is missing.

What to check

The report separates periodic usage from sustained low utilization and names the observation window and dependencies to verify.

Original SKILL.md on GitHub · Updated 2026-11-13

Frequently asked question

Is a low average enough to call a resource idle?

Review the peaks, workload schedule and service role before drawing that conclusion.

Browse Claude Code skills in the Chinese directory
词映 AI 导航Skills 技能
工具分类全部工具已本地化技能选型指南AI 快讯