selfhosted -template

This MCP server provides a remote, full-stack SEO auditing capability powered by the FastMCP framework. It exposes tools that analyze on-page SEO, technical health, and social metadata across a target webpage, returning structured insights and actionable recommendations. You can deploy it as a remote MCP server (via Docker or hosting platforms) and connect clients like Claude Desktop, Cursor, or Windsurf to run comprehensive audits or quick checks. Available tools include analyze_seo for comprehensive site-wide analysis, seo_quick_check for fast health assessments, and seo_meta_tags_check for focused meta and social metadata evaluation. Use the provided mcpServers configuration to register the server in your MCP client and invoke the appropriate commands or prompts to perform analyses on URLs you specify.

Prerequisites:

• Docker installed on your host or platform (for remote deployment)
• Git for cloning the repository (optional if using a pre-built image)
• Python (if you prefer local Python-based setup) and virtual environment tooling (optional)

Installation steps (remote/docker-based deployment):

1. Prepare your environment and obtain access to a host with Docker installed.
2. Deploy the MCP server template (for example via the Hostinger 1-click deploy or by cloning the repository and running Docker Compose as instructed in the repository).
3. Ensure the server is exposed on port 8080 (as indicated in the deployment docs).
4. Register the server in your MCP client using the provided mcpServers entry (see the configuration snippet in this README).

If you prefer a local Python-based setup (for development):

1. Install prerequisites:
   • Python 3.8 or higher
   • pip
   • Docker (optional for containerized deployment)
2. Clone the repository:

   git clone https://github.com/hostinger/selfhosted-mcp-server-template.git
   cd selfhosted-mcp-server-template
   

3. Create and activate a virtual environment (recommended):

   python -m venv venv
   # macOS/Linux
   source venv/bin/activate
   # Windows
   venv\Scripts\activate
   

4. Install dependencies:

   pip install -r requirements.txt
   

5. Run the development/local server according to your setup (e.g., via a docker-compose file or a direct Python entrypoint as documented in the repo). Ensure the server is reachable at your chosen URL/endpoint (default port 8080 if following the template).

Note: The template emphasizes a remote MCP deployment pattern (via Hostinger or other hosting platforms) where Docker handles the runtime, port mapping, and dependency installation automatically.

Tips and considerations:

• The server is designed as a remote MCP server; configure the appropriate URL in your MCP client to point to /mcp on your deployed host (e.g., https://your-app.example.com/mcp).
• The available tools (analyze_seo, seo_quick_check, seo_meta_tags_check) have distinct usage prompts; use those prompts in your MCP client to trigger the corresponding analyses.
• If you deploy locally, ensure your firewall allows traffic to port 8080 if that is your configured port.
• Environment variables can include PORT, ENV, and any custom flags your deployment requires; tailor them to your hosting environment.
• When using the Hostinger template, the deployment process typically handles Docker compose setup, dependency installation, and port configuration automatically; you mainly need to register the endpoint in your MCP client configuration.