wordpress-seo-mcp
Connects AI assistants to WordPress sites, enabling automated SEO analysis and fixes for meta tags, schema, content, images, and technical SEO via MCP tools.
README
WordPress SEO MCP Server
An MCP (Model Context Protocol) server that connects AI assistants to WordPress sites for automated SEO optimization. Works with opencode, Claude Code, Cursor, and any MCP-compatible AI.
Features
| Tool | What it does |
|---|---|
manage-sites |
Add/list/remove WordPress sites |
analyze-seo |
Check meta title, description, OG, Twitter tags |
apply-seo |
Fix missing/short/long meta tags |
analyze-schema |
Inspect JSON-LD structured data |
apply-schema |
Inject Article, Organization, BreadcrumbList schema |
analyze-content |
Check headings, word count, keyword usage |
analyze-images |
Check alt text, lazy loading |
analyze-technical |
Check canonical URL, robots, hreflang |
generate-sitemap |
Generate XML sitemap (optionally submit to Google) |
suggest-keywords |
Generate keyword ideas for a topic |
search-analytics |
Google Search Console analytics (if configured) |
rollback-seo |
Undo last SEO changes |
get-seo-history |
View change history |
Architecture
AI Assistant → MCP Protocol → Middleware Pipeline → Plugin System → WordPress REST API
├─ Logging ├─ Meta Plugin
├─ Auth ├─ Schema Plugin
├─ Validation ├─ Content Plugin
├─ Rate Limit ├─ Images Plugin
└─ Cache ├─ Technical Plugin
├─ Sitemap Plugin
├─ Keyword Plugin
└─ Integration Plugin
Plugin-based with SQLite audit log (enables rollback) and optional Google Search Console + PageSpeed integration.
Quick Start
Prerequisites
- Node.js 18+
- A WordPress site with REST API enabled and an Application Password
Install
git clone https://github.com/abnormal-yi/wordpress-seo-mcp.git
cd wordpress-seo-mcp
npm install
npm run build
Configure in opencode.jsonc
{
"mcp": {
"wordpress-seo-mcp": {
"type": "local",
"command": ["node", "/path/to/wordpress-seo-mcp/dist/index.js"],
"enabled": true,
"environment": {}
}
}
}
Restart your AI assistant. The server starts automatically and exposes all tools.
Usage Workflow
Step 1: Add your site
→ manage-sites({ action: "add", name: "my-site", config: { url, username, appPassword } })
Step 2: Analyze a page
→ analyze-seo({ site: "my-site", postId: 42 })
→ analyze-content({ site: "my-site", postId: 42, keyword: "seo tools" })
→ analyze-schema({ site: "my-site", postId: 42 })
Step 3: Apply fixes
→ apply-seo({ site: "my-site", postId: 42, fixes: ["missing-title", "missing-desc"] })
→ apply-schema({ site: "my-site", postId: 42, schemas: ["Article", "Organization"] })
Step 4: Generate sitemap
→ generate-sitemap({ site: "my-site" })
Step 5: Roll back if needed
→ rollback-seo({ site: "my-site", postId: 42 })
Or just tell your AI:
"Optimize this WordPress page for SEO"
The AI runs the full pipeline automatically.
Google Search Console (optional)
Set these environment variables in your MCP config:
"environment": {
"GSC_API_KEY": "your-google-api-key",
"GSC_SITE_URL": "https://yoursite.com"
}
Enables search-analytics tool and sitemap submission to GSC.
Development
npm run dev # Watch mode
npm test # Run tests (70+ tests)
npm run typecheck # TypeScript check
npm run build # Production build
Project Structure
src/
├── index.ts # Entry point — wires plugins + services
├── transport.ts # MCP stdio transport
├── orchestrator.ts # Tool registry + plugin interface
├── middleware/pipeline.ts # Middleware pipeline runner
├── plugins/
│ ├── management.ts # manage-sites tool
│ ├── meta/ # Meta tag analyzer + applier
│ ├── schema/ # JSON-LD analyzer + applier
│ ├── content/ # Content SEO analyzer
│ ├── images/ # Image SEO analyzer
│ ├── technical/ # Technical SEO analyzer
│ ├── sitemap/ # XML sitemap generator
│ ├── keyword/ # Keyword suggestion engine
│ ├── rollback.ts # Rollback + history tools
│ └── integration/ # GSC + PageSpeed tools
├── services/
│ ├── wordpress/ # REST API client + site pool
│ ├── storage/ # SQLite audit log
│ └── external/ # GSC + PageSpeed API adapters
└── types/ # Shared type definitions
推荐服务器
Baidu Map
百度地图核心API现已全面兼容MCP协议,是国内首家兼容MCP协议的地图服务商。
Playwright MCP Server
一个模型上下文协议服务器,它使大型语言模型能够通过结构化的可访问性快照与网页进行交互,而无需视觉模型或屏幕截图。
Magic Component Platform (MCP)
一个由人工智能驱动的工具,可以从自然语言描述生成现代化的用户界面组件,并与流行的集成开发环境(IDE)集成,从而简化用户界面开发流程。
Audiense Insights MCP Server
通过模型上下文协议启用与 Audiense Insights 账户的交互,从而促进营销洞察和受众数据的提取和分析,包括人口统计信息、行为和影响者互动。
VeyraX
一个单一的 MCP 工具,连接你所有喜爱的工具:Gmail、日历以及其他 40 多个工具。
graphlit-mcp-server
模型上下文协议 (MCP) 服务器实现了 MCP 客户端与 Graphlit 服务之间的集成。 除了网络爬取之外,还可以将任何内容(从 Slack 到 Gmail 再到播客订阅源)导入到 Graphlit 项目中,然后从 MCP 客户端检索相关内容。
Kagi MCP Server
一个 MCP 服务器,集成了 Kagi 搜索功能和 Claude AI,使 Claude 能够在回答需要最新信息的问题时执行实时网络搜索。
e2b-mcp-server
使用 MCP 通过 e2b 运行代码。
Neon MCP Server
用于与 Neon 管理 API 和数据库交互的 MCP 服务器
Exa MCP Server
模型上下文协议(MCP)服务器允许像 Claude 这样的 AI 助手使用 Exa AI 搜索 API 进行网络搜索。这种设置允许 AI 模型以安全和受控的方式获取实时的网络信息。