$ ~/registry/skill/microsoft-plugin-skills-azure-quotas

SKILL

azure-quotas

Name: azure-quotas
Author: microsoft

帮助你检查和管理 Azure 配额与用量，评估部署容量和区域可用性。

星标

★ 232

来源

GitHub

更新于

2026-07-13

// 安全评估低风险

仅提示词，不执行代码
开源可审计
社区验证· 222

总评

该技能材料显示其本质上是一个开源的提示型 Azure 配额管理指南，本身未声明需要密钥、远程端点或本地执行能力，整体风险较低。需注意其文档强指令式措辞与建议安装 Azure CLI 扩展，但基于现有材料未见明确数据外发、越权或恶意红旗。

凭证密钥低风险

材料与客观检查项均显示无需密钥或环境变量；未见要求用户提供 API token、账号密码或将凭证发送到第三方的描述，凭证暴露面较低。

网络外发低风险

未声明任何远程端点，且系统标注为 prompt-only；文档虽提到 Azure Portal、REST API 与 Azure CLI 用途，但未显示该技能自身会向第三方主机发送用户数据。

代码执行低风险

系统检查项表明其为纯提示/文档型内容，而非可执行 MCP 工具；虽然 README 建议使用 `az quota` 和安装 CLI 扩展，但这是使用说明，不代表该技能自身具备本机执行权限。

数据访问低风险

材料未声明可读写本地文件、访问系统目录或直接操作用户数据；仅从说明看，其作用是提供 Azure 配额检查/申请流程指引，未见超出声明目的的数据访问范围。

来源供应链低风险

来源为 GitHub 上的微软相关开源仓库，且有 222 star，具备一定社区信任与可审计性；许可证和维护状态未明确，属于轻微信息缺口，但不足以构成高风险红旗。

安全建议

使用前核对仓库最近提交、Issue/PR 活跃度与实际维护者身份。
若按文档安装 Azure CLI 扩展，优先从官方源获取并审查扩展权限与变更记录。
在生产订阅中执行配额查询或申请前，遵循最小权限原则，仅授予 Reader 或必要的配额相关角色。
将该技能视为操作指南而非自动执行组件，避免盲从文档中的强指令式表述。

审计模型: gpt-5.4 · 2026-06-17

// 安装

复制安装指令，让 AI 自动完成配置 · 推荐新手

请帮我安装 askskill 上的 "azure-quotas" 技能：
1. 下载 https://raw.githubusercontent.com/microsoft/GitHub-Copilot-for-Azure/main/plugin/skills/azure-quotas/SKILL.md
2. 保存为 ~/.claude/skills/azure-quotas/SKILL.md
3. 装好后重载技能，告诉我可以用了

// 下载

下载 SKILL.md机读安装清单 ↗

// 用法示例

检查订阅 vCPU 配额

输入

请帮我检查我的 Azure 订阅在 East US 区域的 vCPU 配额、当前使用量和剩余可用量，并指出是否足够部署 20 台标准虚拟机。

预期产出

返回该区域的 vCPU 配额、已用量、剩余额度，以及是否满足目标部署规模的判断。

评估区域部署容量

输入

我计划把应用部署到 West Europe、North Europe 和 East US，请比较这几个区域的相关配额与容量限制，并推荐最适合上线的区域。

预期产出

给出各区域配额和容量差异、潜在限制，以及基于可用性的区域建议。

判断是否需要申请提额

输入

我们的部署因为 Azure 配额限制失败了。请根据当前使用情况和目标资源需求，判断需要申请哪些配额提升，并说明原因。

预期产出

列出需要提升的配额项、当前缺口和申请原因，帮助准备提额请求。

// 文档

Azure Quotas - Service Limits & Capacity Management

AUTHORITATIVE GUIDANCE — Follow these instructions exactly for quota management and capacity validation.

Overview

What are Azure Quotas?

Azure quotas (also called service limits) are the maximum number of resources you can deploy in a subscription. Quotas:

Prevent accidental over-provisioning
Ensure fair resource distribution across Azure
Represent available capacity in each region
Can be increased (adjustable quotas) or are fixed (non-adjustable)

Key Concept: Quotas = Resource Availability

If you don't have quota, you cannot deploy resources. Always check quotas when planning deployments or selecting regions.

When to Use This Skill

Invoke this skill when:

Planning a new deployment - Validate capacity before deployment
Selecting an Azure region - Compare quota availability across regions
Troubleshooting quota exceeded errors - Check current usage vs limits
Requesting quota increases - Submit increase requests via CLI or Portal
Comparing regional capacity - Find regions with available quota
Validating provisioning limits - Ensure deployment won't exceed quotas

Quick Reference

Property	Details
Primary Tool	Azure CLI (`az quota`) - USE THIS FIRST, ALWAYS
Extension Required	`az extension add --name quota` (MUST install first)
Key Commands	`az quota list`, `az quota show`, `az quota usage list`, `az quota usage show`
Complete CLI Reference	commands.md
Azure Portal	My quotas - Use only as fallback
REST API	Microsoft.Quota provider - Unreliable, do NOT use first
MCP Server	`azure-quota` MCP server — NEVER use this. It is unreliable. Always use `az quota` CLI instead.
Required Permission	Reader (view) or Quota Request Operator (manage)

⚠️ ALWAYS USE CLI FIRST

REST API and Portal can show misleading "No Limit" values — this does not mean unlimited capacity. It means the quota API doesn't support that resource type. Always start with az quota commands; fall back to Azure service limits docs if CLI returns BadRequest.

For complete CLI reference, see commands.md.

Quota Types

Type	Adjustability	Approval	Examples
Adjustable	Can increase via Portal/CLI/API	Usually auto-approved	VM vCPUs, Public IPs, Storage accounts
Non-adjustable	Fixed limits	Cannot be changed	Subscription-wide hard limits

Important: Requesting quota increases is free. You only pay for resources you actually use, not for quota allocation.

Understanding Resource Name Mapping

⚠️ CRITICAL: There is NO 1:1 mapping between ARM resource types and quota resource names.

Example Mappings

ARM Resource Type	Quota Resource Name
`Microsoft.App/managedEnvironments`	`ManagedEnvironmentCount`
`Microsoft.Compute/virtualMachines`	`standardDSv3Family`, `cores`, `virtualMachines`
`Microsoft.Network/publicIPAddresses`	`PublicIPAddresses`, `IPv4StandardSkuPublicIpAddresses`

Discovery Workflow

Never assume the quota resource name from the ARM type. Always use this workflow:

List all quotas for the resource provider:

az quota list --scope /subscriptions/<id>/providers/<ProviderNamespace>/locations/<region>

Match by localizedValue (human-readable description) to find the relevant quota
Use the name field (not ARM resource type) in subsequent commands:

…

查看完整文档 ↗

引导你逐步自定义 Azure OpenAI 模型部署参数与高级选项。

microsoft装→

$ loading_