Skip to content

MeganSteenkamp/yipyap

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

10 Commits
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

YipYap - AI Tech News MCP Server

An MCP (Model Context Protocol) server that fetches AI tech news from Reddit and Hacker News, with AI-powered article summarization.

Features

Four tools to help you stay up-to-date with AI news:

  • summarise_weekly - Get top AI news from the past week with AI-generated TLDRs
  • get_news - Search for news about specific AI topics with TLDRs
  • get_trending - Find posts with high engagement (lots of discussion) with TLDRs
  • get_drama - Discover controversial AI discussions with TLDRs

Each tool now includes AI-powered article summarization using Groq's Llama model, providing concise TLDRs for linked articles. Photo-only posts are automatically filtered out to focus on substantive content.

Installation

uv sync

Setup

  1. Get a free Groq API key from https://console.groq.com

  2. Create a .env file in the project root:

GROQ_API_KEY=your_groq_api_key_here

Usage

Testing with MCP Inspector

Test the server locally:

uv run mcp dev yipyap.py

Using with Claude Desktop

Add to your Claude Desktop config file:

macOS: ~/Library/Application Support/Claude/claude_desktop_config.json

Windows: %APPDATA%/Claude/claude_desktop_config.json

{
  "mcpServers": {
    "yipyap": {
      "command": "python",
      "args": ["/path/to/yipyap/yipyap.py"]
    }
  }
}

Then restart Claude Desktop.

Tools

summarise_weekly()

Get the top 10 AI news posts from the past week across Reddit and Hacker News, sorted by score. Each post includes an AI-generated TLDR of the linked article.

Example:

summarise_weekly()

get_news(keyword, days=7)

Search for news about a specific AI topic with AI-generated TLDRs.

Parameters:

  • keyword (required) - Search term (e.g., "GPT-5", "Claude", "Gemini")
  • days (optional) - Number of days to look back (default: 7)

Example:

get_news(keyword="Claude", days=3)

get_trending(days=7)

Find posts with high engagement - lots of comments relative to upvotes. Includes AI-generated TLDRs.

Parameters:

  • days (optional) - Number of days to look back (default: 7)

Example:

get_trending(days=7)

get_drama(days=7)

Discover controversial AI discussions on Reddit with AI-generated TLDRs.

Parameters:

  • days (optional) - Number of days to look back (default: 7)

Example:

get_drama(days=3)

Sources

Reddit subreddits:

  • MachineLearning
  • artificial
  • LocalLLaMA
  • OpenAI
  • ClaudeAI
  • PromptEngineering
  • learnmachinelearning
  • neuralnetworks
  • singularity
  • programming
  • LangChain
  • StableDiffusion
  • ArtificialInteligence
  • Futurology
  • ControlProblem

Hacker News:

  • All AI-related stories via Algolia API

How It Works

  1. Fetches posts from Reddit and Hacker News
  2. Filters out photo-only posts
  3. Fetches article content from URLs
  4. Generates concise TLDRs using Groq's Llama 3.3 70B model
  5. Returns formatted results with metadata and summaries

License

MIT

About

Tech news MCP

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

 
 
 

Contributors

Languages