Hn Companion MCP

🚀 Hacker News Companion MCP Project Introduction

A Model Context Protocol (MCP) for summarizing Hacker News discussions using Claude.

🚀 Quick Start

This MCP fetches and processes discussions from Hacker News, formatting them into a format that Claude can use to generate high - quality summaries. It handles the hierarchical structure of comments and their metadata (scores, downvotes, etc.), helping Claude understand the relative importance and relationships of different comments.

✨ Features

• Handle Hacker News links or post IDs.
• Download and analyze the comment structure on Hacker News.
• Score comments based on community engagement.
• Prepare data in a format suitable for Claude's summary generation.

📦 Installation

Install via Smithery

To automatically install the Hacker News Companion for the Claude desktop version via Smithery:

npx -y @smithery/cli install @georgeck/hn-companion-mcp --client claude

Manual Installation

1. Clone the repository:

   git clone https://github.com/yourusername/hn-companion-mcp.git
   cd hn-companion-mcp
   

2. Install dependencies:

   npm install
   

💻 Usage Examples

CLI Command Line

node index.js <post-id-or-url>

Example:

node index.js 43448075
# Or
node index.js https://news.ycombinator.com/item?id=43448075

API Server

Start the server:

npm start

Send a request:

curl -X POST http://localhost:3000/api/summarize \
  -H "Content-Type: application/json" \
  -d '{"input": "https://news.ycombinator.com/item?id=43448075"}'

📚 Documentation

POST /api/summarize Interface

Request body:

{
  "input": "https://news.ycombinator.com/item?id=43448075"
}

Response:

{
  "status": "success",
  "data": {
    "systemPrompt": "...",
    "userPrompt": "...",
    "commentPathIdMapping": { ... },
    "postTitle": "...",
    "postId": "...",
    "commentCount": 123
  }
}

🔧 Technical Details

This MCP is designed to prepare data for Claude for summarization. When a user asks Claude to summarize a Hacker News discussion, Claude can call this MCP to obtain formatted data and then generate a summary based on the provided system and user prompts.

"hn-companion": {
      "command": "node",
      "args": ["<full path to src>/hn-companion-mcp/server.js"]
    }
  }

📄 License

MIT