mcpField
I
infrastructure-documenter

infrastructure-documenter

github.com/jmsktm云与 DevOps10205 次安装40 周前更新
代码仓库

关于 infrastructure-documenter

面向工程与运维团队的基础设施文档编写技能,提供架构说明、运行手册、API 文档和决策记录等模板与实践,并涵盖文档即代码。

主要功能

  1. 01架构文档与图表
  2. 02事件运行手册
  3. 03API 与环境文档
  4. 04架构决策记录
  5. 05文档即代码实践

适用场景

  1. 01记录系统组件与数据流
  2. 02编写故障排查和恢复步骤
  3. 03建立 API 与环境参考文档
  4. 04规划文档目录和自动生成流程

安装

复制下面的命令,在终端中运行。

$npx skills add https://github.com/jmsktm/claude-settings --skill infrastructure-documenter

安装后在 Claude Code 中输入/infrastructure-documenter确认已加载。

详细介绍

基础设施文档编写技能

概述

此技能可帮助你创建清晰、易维护的基础设施文档,涵盖架构图、运行手册、系统文档、运维流程以及文档即代码实践。

文档理念

原则
  1. 持续更新的文档:确保文档与实际情况同步
  2. 关注受众:为不同读者编写不同文档
  3. 可执行:每份文档都应帮助读者完成某项工作
  4. 版本控制:使用代码变更追踪文档变更
文档类型
类型受众用途
架构工程师了解系统设计
运行手册运维/SRE处理事故
API 文档开发者与系统集成
入职指南新员工快速熟悉工作
决策记录未来的你了解决策原因

架构文档

系统架构概览
# System Architecture

## Overview

[Project Name] is a [type] application that [purpose].

## High-Level Architecture

┌─────────────────────────────────────────────────────────────┐ │ Users │ └─────────────────────────────────────────────────────────────┘ │ ▼ ┌─────────────────────────────────────────────────────────────┐ │ Vercel Edge │ │ ┌─────────────────┐ ┌─────────────────┐ │ │ │ Next.js App │ │ Edge Functions │ │ │ └─────────────────┘ └─────────────────┘ │ └─────────────────────────────────────────────────────────────┘ │ ┌───────────────┼───────────────┐ ▼ ▼ ▼ ┌─────────────────┐ ┌─────────────────┐ ┌─────────────────┐ │ Supabase │ │ Redis │ │ Stripe │ │ - PostgreSQL │ │ - Session │ │ - Payments │ │ - Auth │ │ - Cache │ │ - Webhooks │ │ - Realtime │ │ │ │ │ │ - Storage │ │ │ │ │ └─────────────────┘ └─────────────────┘ └─────────────────┘


## Components

### Frontend (Next.js App)
- **Location**: Vercel Edge Network
- **Framework**: Next.js 14 (App Router)
- **Styling**: Tailwind CSS + shadcn/ui
- **State**: Zustand + React Query

### Backend Services
| Service | Provider | Purpose |
|---------|----------|---------|
| Database | Supabase | PostgreSQL with RLS |
| Auth | Supabase Auth | User authentication |
| Storage | Supabase Storage | File uploads |
| Cache | Upstash Redis | Session & API cache |
| Payments | Stripe | Subscriptions |
| Email | Resend | Transactional emails |

### Data Flow

1. User request → Vercel Edge
2. SSR/API Route processes request
3. Database queries via Supabase client
4. Response cached at edge (when applicable)
5. Response returned to user

## Security

### Authentication Flow
1. User signs in via Supabase Auth
2. JWT token issued and stored in cookie
3. Server validates token on each request
4. RLS policies enforce data access

### Data Protection
- All data encrypted at rest (AES-256)
- TLS 1.3 for data in transit
- Secrets stored in Vercel environment
- PII fields encrypted in database
Mermaid 图表
## Request Flow

```mermaid
sequenceDiagram
    participant U as User
    participant V as Vercel
    participant N as Next.js
    participant S as Supabase
    participant R as Redis

    U->>V: HTTPS Request
    V->>N: Route to App

    alt Cached Response
        N->>R: Check Cache
        R-->>N: Cache Hit
        N-->>U: Return Cached
    else Cache Miss
        N->>S: Query Database
        S-->>N: Data
        N->>R: Store in Cache
        N-->>U: Return Response
    end

数据库架构

erDiagram
    users ||--o{ projects : owns
    users {
        uuid id PK
        text email
        text name
        timestamp created_at
    }
    projects ||--o{ tasks : contains
    projects {
        uuid id PK
        uuid user_id FK
        text name
        text status
    }
    tasks {
        uuid id PK
        uuid project_id FK
        text title
        boolean completed
    }

## Runbooks

### Runbook Template

```markdown
# Runbook: [Service Name] - [Issue Type]

## Overview
Brief description of the issue and when this runbook applies.

## Severity
- **P1 (Critical)**: Complete outage
- **P2 (High)**: Degraded service
- **P3 (Medium)**: Minor impact
- **P4 (Low)**: No user impact

## Detection
How this issue is typically detected:
- [ ] Alert from [monitoring system]
- [ ] User report
- [ ] Automated check failure

## Impact Assessment
- **Users affected**: All / Segment / None
- **Data at risk**: Yes / No
- **Revenue impact**: High / Medium / Low / None

## Prerequisites
- [ ] Access to [system/dashboard]
- [ ] Credentials for [service]
- [ ] Contact info for [team/person]

## Resolution Steps

### Step 1: Verify the Issue
```bash
# Check service status
curl -I https://api.example.com/health

# Check logs
vercel logs --follow
步骤 2:确定根本原因

常见原因:

  • 数据库连接池耗尽
  • 达到内存上限
  • 外部服务宕机
  • 部署不当
步骤 3:应用修复
如果是数据库问题:
# Check connection count
SELECT count(*) FROM pg_stat_activity;

# Kill idle connections
SELECT pg_terminate_backend(pid)
FROM pg_stat_activity
WHERE state = 'idle' AND query_start < now() - interval '1 hour';
如果是部署不当:
# Rollback to previous deployment
vercel rollback
步骤 4:验证修复
# Check service health
curl https://api.example.com/health

# Monitor error rates for 15 minutes

升级处理

如果 30 分钟内无法解决:

  1. 呼叫值班工程师:[contact]
  2. 在 #incidents 中通知相关方
  3. 更新状态页

事故后续

  • 创建事故报告
  • 安排事后复盘(仅限 P1/P2)
  • 如有需要,更新此运行手册

相关链接


### Database Runbooks

```markdown
# Runbook: Database Performance Issues

## Symptoms
- Slow API responses (>1s)
- Timeout errors in logs
- High database CPU in dashboard

## Quick Checks

### 1. Check Active Connections
```sql
SELECT
  state,
  count(*),
  max(now() - query_start) as max_duration
FROM pg_stat_activity
GROUP BY state;
2. 查找长时间运行的查询
SELECT
  pid,
  now() - query_start AS duration,
  query
FROM pg_stat_activity
WHERE state = 'active'
  AND now() - query_start > interval '30 seconds'
ORDER BY duration DESC;
3. 检查表大小
SELECT
  schemaname,
  tablename,
  pg_size_pretty(pg_total_relation_size(schemaname || '.' || tablename)) as size
FROM pg_tables
WHERE schemaname = 'public'
ORDER BY pg_total_relation_size(schemaname || '.' || tablename) DESC
LIMIT 10;
4. 检查缺失的索引
SELECT
  relname,
  seq_scan,
  idx_scan,
  seq_scan - idx_scan AS difference
FROM pg_stat_user_tables
WHERE seq_scan > idx_scan
ORDER BY difference DESC;

解决方案

终止问题查询
SELECT pg_terminate_backend(pid)
FROM pg_stat_activity
WHERE pid = [PID_FROM_ABOVE];
添加缺失的索引
CREATE INDEX CONCURRENTLY idx_table_column
ON table_name (column_name);

## Decision Records (ADRs)

### ADR Template

```markdown
# ADR-001: Choose Supabase for Database

## Status
Accepted

## Context
We need a database solution for [Project Name] that supports:
- PostgreSQL compatibility
- Real-time subscriptions
- Built-in authentication
- Easy local development
- Generous free tier

## Decision
We will use Supabase as our primary database and auth provider.

## Alternatives Considered

### PlanetScale
**Pros:**
- Excellent scaling
- Branching for schema changes
- MySQL compatible

**Cons:**
- No built-in auth
- No real-time subscriptions
- Additional services needed

### Firebase
**Pros:**
- Real-time built-in
- Mature platform
- Good mobile SDKs

**Cons:**
- NoSQL (not ideal for our use case)
- Vendor lock-in concerns
- Complex security rules

## Consequences

### Positive
- Single provider for DB + Auth + Storage
- Great developer experience
- Row Level Security for data protection
- Local development with supabase CLI

### Negative
- PostgreSQL-specific features tie us to provider
- Supabase still maturing (some rough edges)
- Limited to their managed offering

### Risks
- Supabase scaling limitations at high traffic
- Migration cost if we need to move

## References
- [Supabase Documentation](https://supabase.com/docs)
- [Comparison: Supabase vs Firebase](https://...)

API 文档

端点文档
# API Reference

## Base URL

生产环境:https://api.example.com/v1 预发布环境:https://staging-api.example.com/v1


## Authentication

All API requests require authentication via Bearer token.

```bash
curl -H "Authorization: Bearer YOUR_TOKEN" \
  https://api.example.com/v1/users

端点

用户
获取当前用户
GET /users/me

响应:

{
  "id": "usr_123",
  "email": "[email protected]",
  "name": "John Doe",
  "created_at": "2024-01-01T00:00:00Z"
}
更新用户
PATCH /users/me

请求正文:

字段类型必填描述
namestring否显示名称
avatar_urlstring否个人资料图片 URL

示例:

curl -X PATCH \
  -H "Authorization: Bearer YOUR_TOKEN" \
  -H "Content-Type: application/json" \
  -d '{"name": "Jane Doe"}' \
  https://api.example.com/v1/users/me
错误响应
状态代码描述
400BAD_REQUEST请求正文无效
401UNAUTHORIZED缺少令牌或令牌无效
403FORBIDDEN权限不足
404NOT_FOUND未找到资源
429RATE_LIMITED请求过多
500INTERNAL_ERROR服务错误

错误响应格式:

{
  "error": {
    "code": "NOT_FOUND",
    "message": "User not found"
  }
}

## Environment Documentation

### Environment Matrix

```markdown
# Environments

## Overview

| Environment | URL | Purpose | Deploy |
|-------------|-----|---------|--------|
| Production | https://myapp.com | Live users | Manual (main) |
| Staging | https://staging.myapp.com | Pre-release testing | Auto (main) |
| Preview | https://pr-*.vercel.app | PR review | Auto (PR) |
| Development | http://localhost:3000 | Local dev | Manual |

## Configuration

### Production
```env
NODE_ENV=production
DATABASE_URL=[Supabase Production]
NEXT_PUBLIC_APP_URL=https://myapp.com
预发布环境
NODE_ENV=production
DATABASE_URL=[Supabase Staging Branch]
NEXT_PUBLIC_APP_URL=https://staging.myapp.com
开发环境
NODE_ENV=development
DATABASE_URL=[Local Supabase]
NEXT_PUBLIC_APP_URL=http://localhost:3000

访问权限

生产环境
  • Vercel:仅限管理员
  • 数据库:开发人员只读,管理员可写
  • 日志:所有工程师
预发布环境
  • Vercel:所有工程师
  • 数据库:所有工程师
  • 日志:所有工程师

密钥轮换

密钥轮换周期上次轮换时间
数据库密码90 天2024-01-15
API 密钥90 天2024-01-15
JWT 密钥从不初始设置

## Documentation-as-Code

### Documentation Structure

docs/ ├── README.md # 文档索引 ├── architecture/ │ ├── overview.md # 系统架构 │ ├── data-flow.md # 数据流图 │ └── decisions/ # ADR │ ├── 001-database.md │ └── 002-hosting.md ├── runbooks/ │ ├── README.md # 运维手册索引 │ ├── database.md # 数据库问题 │ ├── deployment.md # 部署问题 │ └── outage.md # 服务中断 ├── api/ │ └── reference.md # API 文档 └── onboarding/ ├── setup.md # 本地设置 └── contributing.md # 如何贡献


### Auto-Generated Documentation

```yaml
# .github/workflows/docs.yml
name: Generate Docs

on:
  push:
    branches: [main]
    paths:
      - 'src/**'
      - 'docs/**'

jobs:
  generate-docs:
    runs-on: ubuntu-latest
    steps:
      - uses: actions/checkout@v4

      - name: Generate API docs from OpenAPI
        run: |
          npx @redocly/cli build-docs openapi.yaml \
            --output docs/api/index.html

      - name: Generate TypeDoc
        run: npx typedoc --out docs/api/typescript

      - name: Deploy to GitHub Pages
        uses: peaceiris/actions-gh-pages@v3
        with:
          github_token: ${{ secrets.GITHUB_TOKEN }}
          publish_dir: ./docs

文档检查清单

架构文档
  • 系统概览图
  • 组件说明
  • 数据流文档
  • 安全架构
  • 技术决策(ADR)
运维文档
  • 常见问题的运维手册
  • 部署流程
  • 监控与告警
  • 事件响应计划
  • 值班流程
开发者文档
  • 本地设置指南
  • API 参考
  • 贡献指南
  • 代码规范
  • 测试指南
维护
  • 文档审查计划
  • 指定负责人
  • 定义变更流程
  • 版本管理策略

何时使用此技能

在以下情况下调用此技能:

  • 创建架构文档
  • 编写运维手册
  • 记录决策理由(ADR)
  • 设置文档结构
  • 创建入职材料
  • 构建自动化文档
  • 规划事件响应流程

评论

支持 Markdown