YAPS Influence Score Server wraps the Kaito YAPS API to provide tokenized attention scores and credibility metrics for X/Twitter accounts. It enables LLMs to query real-time social media influence data, assess account trustworthiness, and identify key opinion leaders in the crypto space. Key features: - Query YAPS credibility scores for any X/Twitter account with 10-minute Redis caching - Compare two influencers side-by-side with score deltas and natural language summaries - Access daily top-10 crypto influencer leaderboards with 24-hour cache refresh cycles - Graceful handling of the YAPS API's 100 calls per 5-minute rate limit

YAPS MCP Server

An MCP (Model Context Protocol) server that wraps Kaito's YAPS API to provide tokenized attention scores for X/Twitter accounts. This server enables LLMs to query credibility and influence metrics for any X account, with built-in caching and rate limiting.

Features

Getting Started

Prerequisites

Installation

1. Clone the repository

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

2. Install dependencies

npm install

3. Create a .env file based on the following template:

# YAPS API (no API key needed - using public API with rate limits)
YAPS_API_ENDPOINT=https://api.kaito.ai/api/v1/yaps
Redis cache
REDIS_URI=redis://localhost:6379
Server
PORT=3000
NODE_ENV=development
YAPS cache TTL in seconds
YAPS_CACHE_TTL=600  # 10 minutes
Leaderboard settings
LEADERBOARD_CACHE_TTL=86400  # 24 hours

4. Build the project

npm run build

5. Start the server

npm start

For integration with LLM providers, use the stdio transport:

npm start -- --stdio

MCP Resources and Tools

Resources

Tools

Using with OpenAI GPT

Example of using the YAPS MCP server with OpenAI:

import OpenAI from 'openai';
import { spawn } from 'child_process';
const openai = new OpenAI({ apiKey: 'your-api-key' });
// Start the MCP server as a child process
const mcpProcess = spawn('node', ['dist/index.js', '--stdio'], {
  stdio: ['pipe', 'pipe', process.stderr]
});
// Example function to call GPT with tools
async function askGptWithTools(question) {
  try {
    const response = await openai.chat.completions.create({
      model: 'gpt-4-turbo',
      messages: [{ role: 'user', content: question }],
      tools: [
        {
          type: 'function',
          function: {
            name: 'get_yaps_score',
            description: 'Get YAPS tokenized attention score for an X/Twitter account',
            parameters: {
              type: 'object',
              properties: {
                username: {
                  type: 'string',
                  description: 'Twitter username (with or without @)'
                }
              },
              required: ['username']
            }
          }
        },
        // Additional tools would be defined here
      ],
      tool_choice: 'auto'
    });
    console.log(response.choices[0].message);
    
    // Process tool calls if any
    if (response.choices[0].message.tool_calls) {
      // Implementation for handling tool calls would go here
    }
    return response;
  } catch (error) {
    console.error('Error calling GPT:', error);
    throw error;
  }
}
// Example usage
askGptWithTools('What is the YAPS score for @VitalikButerin?');

Performance and Availability

API Rate Limits

The YAPS API is available as an open source API with the following limitations:

Technical Details

License

MIT