eigenlayer-mcp-server

eigenlayer-mcp-server

1

The EigenLayer MCP Server is designed to interface EigenLayer documentation with AI assistants through the Model Context Protocol. It can operate as both a local server and a Vercel serverless function, aimed at testing rather than production usage.

EigenLayer MCP Server built on Next.js

Features

  • Provides EigenLayer documentation to Claude or other AI assistants via MCP
  • Runs as a standalone server locally or as a serverless function on Vercel

Inspired by initial testing here. Cloned from Vercel Next.js MCP template

Test the public endpoint with Claude:

  1. Add the live URL to Claude via the following command (in any folder):
    claude mcp add --transport sse eigenlayer-mcp-server https://eigenlayer-mcp-server-sand.vercel.app/sse
  2. Install Claude Code. Run command claude to enter Claude Code and confirm the mcp server is added
  3. Test with a query like "Can you explain how EigenLayer's restaking mechanism works?"

Build, Run, Testing Locally

In terminal window 1

rm -rf .next node_modules .vercel
pnpm store prune
pnpm install
pnpm build
pnpm dev

node scripts/test-client.mjs https://localhost:3000

In terminal window 2 Test with MCP Inspector

npx @modelcontextprotocol/inspector node public/index.js

Additional notes here.

Add to Claude Code

claude mcp add

(then follow on screen instructions)

Sample Client

script/test-client.mjs contains a sample client to try invocations.

node scripts/test-client.mjs http://localhost:3000

Security Bugs

Please report security vulnerabilities to . Do NOT report security bugs via Github Issues.

Disclaimer

🚧 EigenLayer MCP Server is under active development and has not been audited. EigenLayer MCP Server is rapidly being upgraded, features may be added, removed or otherwise improved or modified and interfaces will have breaking changes. EigenLayer MCP Server should be used only for testing purposes and not in production. EigenLayer MCP Server is provided "as is" and Eigen Labs, Inc. does not guarantee its functionality or provide support for its use in production. 🚧