mcp

richard-senior/mcp

3.2

If you are the rightful owner of mcp and would like to certify it and/or have it hosted online, please leave a comment on the right or send an email to henry@mcphub.com.

A Model Context Protocol (MCP) server implementation for Amazon Q / Claude, providing a single cross-platform executable with multiple tools and resources.

Tools
  1. Google Search

    Fetches top 'n' links from Google search for a given term.

  2. Html to Markdown

    Converts HTML content to Markdown format for LLM processing.

  3. Image Finder

    Retrieves images from Wikipedia based on search terms.

github.com/richard-senior/mcp

A Model Context Protocol (MCP) server implementation for Amazon Q / Claude etc. A single cross platform executable containing multiple tools/resources/prompts etc.

Project Structure

This project follows the standard Go project layout:

  • /cmd - Main application entry point
  • /internal - Private application and library code
  • /pkg - Library code that's ok to use by external applications
  • /configs - Configuration files
  • /test - Test applications and test data

Getting Started

  1. Build the project:
    ./build.sh
    
  2. Configure Amazon Q Chat
    1. Create directory : ~/.aws/amazonq
    2. Create file : ~/.aws/amazonq/mcp.json
      {
      "mcpServers": {
         "mcp": {
            "name": "mcp",
            "command": "$$ The absolute path to compiled MCP binary $$",
            "description": "Custom MCP server",
            "timeout": 2000
         }
      }
      }
      
    3. Stop and restart Q chat
    4. In the Q chat prompt use /tools This should show the list of tools published by MCP
    5. Ask Q chat to 'get an image of Elvis Presley into the local directory'

Current tools

Google Search

Using a private API key (which you should change in config) The tool can get the description, link and name of the top 'n' links found by google search for the search term. For example ask Q Chat 'Please use google to find information about Elvis Presley'

Html to Markdown

LLM's prefer markdown as a format, so we need a tool to convert html to markdown This allows the LLM to 'precis' a web page. For example ask Q Chat 'please precis the information in https://en.wikipedia.org/wiki/Elvis_Presley' or 'Use the web to get information about Elvis Presley'

Image Finder

Uses Wikipedia to get binary images (photo's etc) by search term for example ask Q Chat to 'get an image of Elvis Presley into the local directory'

Development

This project is in the initial setup phase.