README.md

metadata

title: EigenLedger MCP Service
emoji: 🔧
colorFrom: blue
colorTo: indigo
sdk: docker
pinned: false
license: mit

EigenLedger MCP Service

This deployment package exposes core EigenLedger and empyrical analytics through MCP tools.

Available Tools

• health_check
• list_modules
• list_symbols
• compute_simple_returns
• compute_cagr
• compute_max_drawdown
• compute_sharpe_ratio
• create_engine

Local stdio Usage

Run local stdio mode for Claude Desktop / CLI:

cd EigenLedger/mcp_output/mcp_plugin
python main.py

or:

cd EigenLedger/mcp_output
python start_mcp.py

Default transport is stdio.

HTTP Usage

Run HTTP mode locally:

cd EigenLedger/mcp_output
MCP_TRANSPORT=http MCP_PORT=8000 python start_mcp.py

MCP endpoint:

• http://localhost:8000/mcp

Docker / HuggingFace Spaces

This package is configured for HuggingFace Spaces Docker runtime:

• Port: 7860
• Transport: http
• Entry point: python EigenLedger/mcp_output/start_mcp.py

After deployment, clients should connect to:

• https://<your-space-host>/mcp