MCP Server (Claude / AI Agents)¶

fin-ratios ships an MCP (Model Context Protocol) server that exposes all scoring models as tools for AI agents, including Claude Desktop.

Installation¶

pip install "financial-ratios[mcp,fetchers]"

Start the Server¶

fin-ratios serve

Claude Desktop Configuration¶

Add to ~/Library/Application Support/Claude/claude_desktop_config.json:

{
  "mcpServers": {
    "fin-ratios": {
      "command": "fin-ratios",
      "args": ["serve"]
    }
  }
}

Restart Claude Desktop. You can now ask:

"What is Apple's investment score based on the last 7 years of EDGAR data?"

"Compare the moat scores of AAPL, MSFT, and GOOGL."

"What is the fair value range for Microsoft at a 9% WACC?"

Available MCP Tools¶

Tool	Description
`investment_score`	Full investment score from ticker + source
`quality_score`	Quality Factor Score
`moat_score`	Economic Moat Score
`valuation_score`	Valuation Attractiveness Score
`compute_ratios`	Compute individual ratios from data
`fetch_annual_data`	Fetch annual financials from any supported source

Programmatic Usage¶

from fin_ratios.mcp_server import create_mcp_server

server = create_mcp_server()
# Use with any MCP-compatible client