# SEO Content Machine [Health: Active]

**Category:** 🔎 Search & Data Extraction  
**Repository:** https://github.com/gitTerebi/seo-content-machine-plugin  
**GitHub Stars:** 0  
**Views:** 0  
**Installs:** 0  
**Upvotes:** 0  
**Directory Page:** https://allmcps.com/mcp/seo-content-machine

## Description
Run SEO Content Machine tasks from an MCP client: scraping, crawls, Google Maps, Workspace pages.

## Claude Desktop Quick Installation
Remote MCP endpoint (confidence: high). Install path detected from listing signals. Add as a URL/SSE server in your client:

```json
"mcpServers": {
  "seo-content-machine": {
    "url": "http://localhost:8008/mcp"
  }
}
```

## Documentation

## What SEO Content Machine MCP server does

SEO Content Machine MCP server gives an MCP client access to the SEO Content Machine desktop SEO workspace. It can create and run tasks, then read the resulting data through SCM. The supported workflows cover several forms of research and content administration:

- Scrape articles from search results using keyword lists.
- Crawl a website and identify its internal links.
- Find Google Maps listings, including ratings and reviews.
- Download archived copies of websites from archive.org.
- Extract selected page values into CSV files.
- Add or update pages in an SEO Workspace.

The actual browsing, scraping, task execution, and file output happen inside SEO Content Machine on the local machine.

## How it works

The repository contains a single MCP server file, `mcp/index.mjs`, which has no dependencies. Claude Code starts it with Node.js and communicates with it over standard input. For each request, the server sends the corresponding message to SCM's MCP endpoint, which defaults to `http://localhost:8008/mcp`.

If an API key is configured in SCM, the server sends it as a Bearer authorization header. It does not send requests to another service. When SCM is not running, the server checks the configured address every five seconds and asks the client to refresh its tool list after SCM becomes available.

SEO Content Machine MCP server is therefore dependent on a local SCM installation rather than providing an independent scraping backend. Tasks started through the client use SCM's own output folders and the websites or searches specified in the task.

## Setup and configuration

The requirements are:

- SEO Content Machine installed and running on Windows, macOS, or Linux.
- Node.js 18 or newer.
- An SCM installation with access to the application's local API.

The documented installation method is through Claude Code:

```text
/plugin marketplace add gitTerebi/seo-content-machine-plugin
/plugin install seo-content-machine@seo-content-machine
```

The plugin uses SCM's default settings. Its optional configuration entries are an SCM API key and an SCM API address. The key is needed only when one has been configured under SCM's application settings. The address only needs changing when SCM is not using its default local endpoint. SCM can be opened before or after Claude Code; its tools appear when the application is available.

The same MCP server is also published as the `seo-content-machine-mcp` npm package for use with other MCP clients. The README does not provide client-specific setup instructions for those clients.

## Tools and capabilities

The available capabilities are centered on SCM task execution and result access. An agent can request local search and content research, inspect crawl results, collect local business data, retrieve archived website material, create CSV output from page fields, and manage SEO Workspace pages. Example requests include finding highly rated dentists in a location, checking which crawled pages lack internal links, and adding multiple page ideas to an SEO Workspace.

## Limitations and notes

SCM is a paid desktop application, although its provider offers a 14-day trial without requiring a card. The SCM application must be available on the same machine as the MCP server. Node.js 18 or later is required. The README does not enumerate individual MCP tool names or define the exact input schema for each task, so those details should be checked in the running client. Results are written to SCM's output folders rather than being described as a separate hosted data store.

_Full upstream README: https://allmcps.com/mcp/seo-content-machine/readme_