Pyth Network MCP

🚀 Pyth Network MCP Server

An MCP server offering real - time access to Pyth Network's decentralized oracle price feeds via the Hermes API, optimized for seamless integration into AI agents and autonomous systems.

Author: @itsomsarraf

🚀 Quick Start

The Pyth Network MCP Server provides real - time access to decentralized oracle price feeds. To get started, first, clone the repository and install the necessary dependencies. Then, configure the MCP client and run the server. You can test the server using the included client.

✨ Features

🔧 Tools (5 Available)

• Search Price Feeds: Search and filter through over 1,930 price feeds by symbol or asset type across more than 107 blockchains.
• Get Latest Prices: Fetch real - time price updates for specific price feed IDs with parsed data (no binary clutter).
• Historical Price Data: Query price updates at specific timestamps for backtesting and analysis.
• Publisher Stake Caps: Access the latest publisher stake caps data from the Pyth network.
• Time - Weighted Average Prices (TWAP): Calculate TWAP with custom time windows (1 - 600 seconds) for more stable pricing.

💬 Prompts (4 Available)

• Analyze Price Feed: Get detailed analysis of any asset with current price, TWAP, and volatility metrics.
• Compare Prices: Side - by - side comparison of multiple assets with volatility analysis.
• Market Overview: Comprehensive market report for any asset type (crypto, equity, fx, metal, rates).
• Price Alert Setup: Step - by - step guidance for setting up price monitoring systems.

📚 Resources (3 Available)

• Pyth Network Information: Overview of the network, data providers, and capabilities.
• Popular Price Feeds: Quick reference of commonly used price feed IDs for major assets.
• API Documentation Reference: Quick reference guide for Hermes API endpoints.

🌐 Universal Features

• Universal Price Feed IDs: Unlike chain - specific oracles, Pyth uses universal price feed IDs that work across all supported blockchains.
• Full MCP Compliance: Implements tools, prompts, and resources according to MCP specification.

📦 Installation

Clone the Repository

git clone https://github.com/itsomsarraf/pyth-network-mcp.git
cd pyth-network-mcp

Install Dependencies

Using uv (recommended):

uv sync

Using pip:

pip install mcp httpx

The project requires:

• mcp>=1.0.0 - Official MCP SDK
• httpx>=0.27.0 - For HTTP requests to Pyth Hermes API

Configure MCP Client

To use this server with an MCP client like Claude Desktop, add the following to your MCP settings configuration file:

For Claude Desktop (~/Library/Application Support/Claude/claude_desktop_config.json on macOS, %APPDATA%\Claude\claude_desktop_config.json on Windows):

{
  "mcpServers": {
    "pyth-network": {
      "command": "python3",
      "args": ["pyth_mcp_server.py"],
      "cwd": "/absolute/path/to/pyth-network-mcp"
    }
  }
}

For Cursor/Windsurf (.cursorrules or MCP settings):

{
  "mcpServers": {
    "pyth-network": {
      "command": "python3",
      "args": ["pyth_mcp_server.py"],
      "cwd": "/absolute/path/to/pyth-network-mcp"
    }
  }
}

Using uv:

{
  "mcpServers": {
    "pyth-network": {
      "command": "uv",
      "args": ["run", "pyth_mcp_server.py"],
      "cwd": "/absolute/path/to/pyth-network-mcp"
    }
  }
}

💻 Usage Examples

Basic Usage

The server exposes 5 tools, 4 prompts, and 3 resources via the MCP protocol, accessible through Claude Desktop, Cursor/Windsurf, or programmatically.

Tools

Tool: get_price_feeds

Description: Search and filter price feeds from the Pyth network by symbol or asset type. Parameters:

• query (string, optional): Filter results to price feeds whose symbol contains this string (case insensitive).
• asset_type (string, optional): Filter by asset type (case insensitive).

Returns: List of price feed metadata with IDs and attributes.

Natural Language Example (Claude Desktop):

> **User**: "What Bitcoin price feeds are available?"
> 
> **Assistant**: "I found several Bitcoin price feeds including:
> - BTC/USD (ID: e62df6c8b4a85fe1a67db44dc12de5db330f7ac66b72dc658afedf0f4a415b43)
> - BTC/ETH
> - BTC/EUR
> 
> These feeds are available across all 107+ supported blockchains."

Example JSON Response:

{
  "price_feeds": [
    {
      "id": "e62df6c8b4a85fe1a67db44dc12de5db330f7ac66b72dc658afedf0f4a415b43",
      "attributes": {
        "symbol": "Crypto.BTC/USD",
        "asset_type": "Crypto",
        "base": "BTC",
        "quote_currency": "USD",
        "description": "Bitcoin vs US Dollar"
      }
    }
  ]
}

Prompts

Prompt: analyze_price_feed

Description: Get a detailed analysis of any asset including current price, TWAP, confidence intervals, and volatility metrics. Arguments:

• asset_symbol (required): Symbol of the asset to analyze.

Example Usage in Claude Desktop:

> **User**: "Use the analyze_price_feed prompt for ETH"
> 
> **Claude**: *Executes multi - step analysis including:*
> 1. Searches for ETH price feeds
> 2. Gets latest price
> 3. Calculates 5 - minute TWAP
> 4. Provides comprehensive analysis with trends

Resources

Resource: pyth://network/info

Content: General information about Pyth Network

• Network overview
• Over 125 data providers
• Coverage statistics (over 1,930 feeds, over 107 chains)
• Key features and capabilities

Access in Claude Desktop: "Show me the Pyth Network information resource"

📚 Documentation

Supported Asset Types and Coverage

Pyth Network aggregates data from over 125 first - party financial institutions, providing:

• Over 1,930 price feeds across over 107 blockchains
• Asset Types:
  • crypto - Cryptocurrencies (BTC/USD, ETH/USD, SOL/USD, etc.)
  • equity - Traditional equities and stocks
  • fx - Foreign exchange pairs
  • metal - Precious metals (Gold, Silver, etc.)
  • rates - Interest rates and rate products

Pyth operates on a pull - based oracle model, delivering high - frequency price updates (refreshed every ~400ms) directly from data providers to smart contracts across multiple chains.

Prerequisites

• Python: Version 3.10 or higher
• Package Manager: pip or uv (recommended)
• Claude Desktop (optional): For integration with Anthropic's Claude interface
• Cursor/Windsurf (optional): For IDE integration

Running the Server

The server uses stdio transport (standard input/output) for MCP communication:

python3 pyth_mcp_server.py

To test the server with the included client:

python3 pyth_mcp_client.py

Testing the Server

The repository includes a demonstration client (pyth_mcp_client.py) that shows how to use all five tools with the official MCP protocol. Run it with:

python3 pyth_mcp_client.py

Finding Price Feed IDs

To find the correct price feed ID for your desired asset:

1. Use the get_price_feeds tool to search by symbol.
2. Or visit the Pyth Network Price Feeds page to browse all available feeds.
3. Common price feed IDs:
   • BTC/USD: e62df6c8b4a85fe1a67db44dc12de5db330f7ac66b72dc658afedf0f4a415b43
   • ETH/USD: ff61491a931112ddf1bd8147cd1b641375f79f5825126d665480874634fd0ace
   • SOL/USD: ef0d8b6fda2ceba41da15d4095d1da392a0d2f8ed0c6c7bc0f4cfac8c280b56d

Project Structure

pyth-network-mcp/
├── pyth_tools.py                # Pyth Network API client
├── pyth_mcp_server.py           # Official MCP server
├── pyth_mcp_client.py           # Example client
├── pyproject.toml               # Dependencies
├── README.md                    # This file
├── STRUCTURE.md                 # Structure guide
├── .gitignore
└── fastmcp-implementation/      # Alternative FastMCP implementation

Architecture

This official MCP server uses:

• Official MCP SDK: Standards - compliant MCP protocol implementation
• stdio transport: For efficient communication with MCP clients
• Pyth Hermes API: Pyth's REST API for accessing price feed data (https://hermes.pyth.network)
• httpx: Async HTTP client for making API requests

The server operates statelessly, fetching fresh data from Pyth's Hermes API on each request. Binary data is disabled by default for clean, human - readable responses.

When to Use This vs FastMCP?

This repository contains two implementations of the Pyth Network MCP server:

Official MCP (This Root Directory) ✅ Recommended

When to use:

• ✅ Production deployments - Standards - compliant and battle - tested
• ✅ Claude Desktop integration - Official protocol support
• ✅ Cursor/Windsurf integration - Best compatibility with AI IDEs
• ✅ Better performance - Efficient stdio transport
• ✅ Future - proof - Follows MCP specification exactly
• ✅ Stable & reliable - Official SDK with proper error handling

Transport: stdio (standard input/output)
Protocol: Official MCP SDK
Best for: Production, AI agents, real applications

FastMCP Implementation (Subdirectory)

When to use:

• 🧪 Learning & experimentation - Simpler to understand
• 🧪 HTTP/SSE preferred - If you need HTTP - based transport
• 🧪 Quick prototyping - Fast to set up and test
• 🧪 Educational purposes - Great for learning MCP concepts

Transport: HTTP with Server - Sent Events (SSE)
Protocol: FastMCP framework
Best for: Learning, experimentation, HTTP - based workflows

Quick Comparison

📝 Note: Both implementations provide the same five tools and access to the same Pyth Network data. The difference is in the underlying transport and protocol implementation.

See fastmcp-implementation/README.md for FastMCP documentation.

🔧 Technical Details

The server uses the official MCP SDK to implement the MCP protocol. It communicates with MCP clients via stdio transport. For accessing price feed data, it uses the Pyth Hermes API through the httpx async HTTP client. The server operates statelessly, ensuring that each request fetches fresh data from the Hermes API.

📄 License

This project is licensed under the MIT License.

External Resources

• Pyth Network: https://pyth.network
• Pyth Documentation: https://docs.pyth.network
• Hermes API Docs: https://hermes.pyth.network/docs
• Price Feeds: https://pyth.network/price-feeds
• MCP Protocol: https://modelcontextprotocol.io
• GitHub: https://github.com/itsomsarraf/pyth-network-mcp

Contributing

Contributions are welcome! Please feel free to submit a Pull Request.

Support

For issues related to:

• This MCP Server: Open an issue on GitHub
• Pyth Network: Visit Pyth's Discord or documentation
• MCP Protocol: See Model Context Protocol documentation