home / mcp / paws-on-mcp mcp server

Paws-on-MCP MCP Server

A comprehensive Model Context Protocol (MCP) server implementing the latest specification.

Installation

Add the following to your MCP client configuration file.

Configuration

{
  "mcpServers": {
    "hemanth-paws-on-mcp": {
      "url": "http://127.0.0.1:8000/mcp/",
      "headers": {
        "PYTHONPATH": "<PYTHONPATH>"
      }
    }
  }
}

You can run a production-ready MCP 2025-03-26 server that exposes tools, resources, prompts, roots, and enhanced sampling through a simple Python process. This server supports production-grade MCP features, including integration points for data sources and AI-driven analysis, and serves requests locally for development and testing.

How to use

You interact with the MCP server using a client that talks to the server over standard protocols. Start the server from your development environment, then use the client to search data sources, fetch resources, and perform enhanced sampling with model preferences. The server provides endpoints for HackerNews and GitHub data, AI-powered analysis prompts, and server-side sampling with roots.

How to install

Prerequisites you need before starting are Python 3.x and access to a terminal. Install the required Python dependencies, then run the server.

# Install dependencies
pip install -r requirements.txt

# Run the MCP server from the src directory
cd src
python mcp_server.py

Configuration and notes

The MCP server runs locally and serves on the default endpoint http://127.0.0.1:8000/mcp/. It indicates available features such as HackerNews integration, GitHub discovery, server-side sampling with roots, and prompts for tech trends analysis. Use a client to connect to this endpoint and begin making requests. Enhanced sampling supports model preferences and context-aware selections, and all core tools, resources, and prompts are ready for use.

If you need to stop the server, use Ctrl+C in the terminal where the server is running. For testing, you can run the comprehensive test suite to verify Tools, Resources, Prompts, Roots, and Enhanced Sampling under load conditions.

Usage examples and quick actions

- Start the server and confirm it is listening at the default URL.

- Use a client to perform a HackerNews search, fetch GitHub repo info, and request a sampling analysis with model preferences.

Available tools

search_hackernews

Search HackerNews stories and return results based on a query.

get_github_repo_info

Retrieve details about a specified GitHub repository.

get_server_roots

List available sampling roots on the server.

get_server_prompts

List prompt templates available for use.

create_sampling_request

Create an MCP sampling request with model preferences and context data.

analyze_hackernews_trends_with_ai

Perform AI-powered analysis of HackerNews trends.

code_review_with_ai

Provide AI-assisted code review guidance.

request_client_roots

Request access to client file system roots for sampling.