ReviewWebsite

You run a Model Context Protocol (MCP) server that connects AI assistants to the ReviewWebsite API to create and manage website reviews, extract data, convert URLs to Markdown, and more. This server provides a flexible, production-ready way to interact with ReviewWebsite.com through both local (stdio) and remote (HTTP) transports, with configurable AI models, data extraction, SEO insights, and automated content generation workflows.

How to use

You can connect to the ReviewWebsite MCP server using either the stdio transport for CLI-native usage or the Streamable HTTP transport for web-based clients. Provide your API key when required, and choose the action you want to perform, such as creating a review, converting a URL to Markdown, or extracting structured data. You can also run SEO analyses, scrape pages, and summarize websites to drive AI-assisted workflows.

How to install

Prerequisites: ensure you have Node.js (version 18 or newer) and Git installed on your machine.

Step 1 – Install dependencies for the MCP server project.

# Clone the repository
git clone https://github.com/mrgoonie/reviewwebsite-mcp-server.git
cd reviewwebsite-mcp-server

# Install dependencies
npm install

Step 2 – Run the MCP server in development mode using stdio transport (default). This starts the server with hot-reloading and the MCP Inspector.

npm run dev:server

Step 3 – Alternatively, run the MCP server with the Streamable HTTP transport for web clients.

npm run dev:server:http

Step 4 – When the HTTP transport is used, the server will be available at the default MCP URL. You can customize host, port, and path using environment variables if needed.

Configuration and usage notes

HTTP transport configuration uses an MCP HTTP endpoint. The following values are commonly set via environment variables to tailor the HTTP server:

MCP_HTTP_HOST=127.0.0.1
MCP_HTTP_PORT=8080
MCP_HTTP_PATH=/mcp

The HTTP configuration snippet for an MCP client looks like this (remote server):

json
{
  "mcpServers": {
    "reviewwebsite": {
      "type": "http",
      "url": "http://localhost:8080/mcp",
      "args": []
    }
  }
}

If you prefer running the server locally via stdio, use this configuration snippet. It runs the server binary directly via Node, pointing to the built entry script.

json
{
  "mcpServers": {
    "reviewwebsite": {
      "command": "node",
      "args": ["/path/to/reviewwebsite-mcp-server/dist/index.js"],
      "transportType": "stdio"
    }
  }
}

Environment variables shown for HTTP transport include MCP_HTTP_HOST, MCP_HTTP_PORT, and MCP_HTTP_PATH. You can also define a key for the ReviewWebsite API usage, for example REVIEWWEBSITE_API_KEY, to pass credentials to the API via your server or tools.

Security and best practices

Keep your API keys secret. Use environment variable management to avoid exposing secrets in code. When exposing the MCP endpoint, consider network access controls and, if using HTTP transport, enable appropriate authentication mechanisms on the client side to ensure only authorized clients can access the MCP server.

Examples of common actions you can perform

• Create, read, update, and delete website reviews for a given URL with custom instructions.

• Get available AI models to tailor responses to your needs.

• Convert a URL to Markdown using a selected model.

• Extract structured data from a URL with specified instructions.

• Scrape a URL and extract its content for analysis.

• Extract links from a website and analyze internal/external link structure.

• Summarize a URL or an entire website to obtain concise insights.

• Perform SEO insights such as keyword ideas, keyword difficulty, traffic analysis, and backlinks for a domain.

Available tools

get_ai_models

Retrieve the list of available AI models that can be used by the MCP server.

create_review

Create a new website review for a given URL with specified instructions.

get_review

Fetch a review by its identifier.

list_reviews

List reviews with pagination options.

update_review

Update an existing review with new URL or instructions.

delete_review

Delete a review by its identifier.

convert_to_markdown

Convert a URL to Markdown using a chosen AI model.

extract_data

Extract structured data from a URL using provided instructions.

scrape_url

Scrape a URL to retrieve its content for analysis.

extract_links

Extract links from a URL or website.

summarize_url

Summarize the content of a URL using an AI model.

seo_keyword_ideas

Generate keyword ideas for SEO for a given term and region.

seo_keyword_difficulty

Assess keyword difficulty for SEO targeting.

seo_traffic

Analyze website traffic for a domain or URL.

seo_backlinks

Retrieve backlinks data for a domain.