feat: Add MCP Server integration guide for AI agents

[om-sarvam]

Overview

This PR adds a comprehensive integration guide for the Pyth Network MCP Server, enabling developers to easily integrate Pyth price feeds with AI agents, LLMs, and autonomous systems through the Model Context Protocol (MCP).

What is MCP?

The Model Context Protocol (MCP) is an open protocol developed by Anthropic that allows AI applications to securely connect to external data sources and tools. This integration makes Pyth's 1,930+ price feeds accessible to AI systems through a standardized interface.

What's Added

📄 New Page:

Located at: api-instances-and-providers/mcp-server.mdx

Contents:

• Introduction to MCP: What it is and why it matters for AI applications
• Pyth MCP Server Features: Access to all price feeds, real-time & historical data, TWAP calculations
• Installation Guide: Step-by-step setup instructions
• Integration Methods:
  • Claude Desktop (natural language queries)
  • Programmatic Python integration
  • LangChain tool integration
• Common Use Cases:
  • AI trading agents
  • Portfolio management bots
  • DeFi liquidation monitoring
• Data Format Guide: How to parse and use price data
• Best Practices: Rate limiting, price freshness checks, TWAP usage
• Production Deployment: Self-hosting and node provider recommendations

📝 Updated Navigation

Updated _meta.json to include "MCP Server (AI Agents)" in the API Instances and Providers section.

Benefits for the Pyth Ecosystem

✅ Expands Developer Reach: Makes Pyth accessible to AI/ML developers
✅ Natural Language Interface: Users can query prices conversationally through Claude Desktop
✅ Autonomous Systems: Enables AI agents to use Pyth data for decision-making
✅ Open Source: Community-maintained MCP server available to all
✅ Standardized Protocol: Follows MCP specification for wide compatibility

Target Audience

This documentation is valuable for:

• AI researchers building autonomous agents
• Developers integrating LLMs with financial data
• Trading bot developers using AI for decision-making
• Anyone using Claude Desktop, LangChain, or MCP-compatible tools

MCP Server Repository

GitHub: https://github.com/itsomsarraf/pyth-network-mcp
Author: @itsOmSarraf
License: MIT

The server is actively maintained and provides clean, AI-optimized access to all Pyth price feeds through the Hermes API.

Testing

• ✅ All code examples validated
• ✅ Links verified and working
• ✅ Follows existing documentation style and format
• ✅ Properly integrated into navigation structure

Preview

The new documentation will appear under:
Price Feeds → API Instances and Providers → MCP Server (AI Agents)

---

This integration significantly expands Pyth Network's accessibility to the rapidly growing AI/ML developer community and autonomous systems ecosystem. The MCP server has already been successfully tested with Claude Desktop and provides a seamless experience for AI-powered applications.

[vercel]

@itsOmSarraf is attempting to deploy a commit to the Pyth Network Team on Vercel.

A member of the Team first needs to authorize it.