xl-addin-mcp

MCP server that exposes XLQuote, DataFrame, Technical Indicator, and Vega Chart Excel add-in functions to AI assistants. Ask your AI to build Excel workbooks with real-time market data, technical indicators, data pipelines, and charts — all using the add-in formulas you already know.

Quick start

npx xl-addin-mcp

No install needed. Requires Node.js 18+.

Setup by AI client

Claude Code

claude mcp add xl-addin -- npx xl-addin-mcp

Claude Desktop

Add to your claude_desktop_config.json:

{
  "mcpServers": {
    "xl-addin": {
      "command": "npx",
      "args": ["xl-addin-mcp"]
    }
  }
}

Config file location:

• Windows: %APPDATA%\Claude\claude_desktop_config.json
• macOS: ~/Library/Application Support/Claude/claude_desktop_config.json

Cursor / Windsurf

Add to your MCP settings (.cursor/mcp.json or equivalent):

{
  "mcpServers": {
    "xl-addin": {
      "command": "npx",
      "args": ["xl-addin-mcp"]
    }
  }
}

SSE transport (web-based clients)

npm run start:sse

Or, from npm without a global install:

npx -p xl-addin-mcp xl-addin-mcp-sse

xl-addin-mcp-sse is a bin exposed by the xl-addin-mcp package, not a standalone package name. This starts an Express server on port 3100. Connect your client to http://localhost:3100/sse. Set PORT env var to change the port.

Available tools

| Tool | Description | |------|-------------| | list_functions | List add-in functions, optionally filtered by category | | search_functions | Search functions by keyword across all four add-ins | | get_function_details | Get complete details for a specific function | | get_formula | Build an exact Excel formula string | | get_symbol_format | Get correct symbol format for a data source | | generate_workbook | Generate an .xlsx workbook with add-in formulas | | list_templates | List predefined workbook templates |

What you can ask your AI

• "Create a portfolio dashboard for AAPL, MSFT, GOOGL with price, P/E, and dividend yield"
• "Pull SPY OHLCV data and add SMA 50/200 indicators"
• "What's the correct symbol format for TradingView?"
• "Build a workbook with a Vega line chart of SMA crossovers for the last 90 days"
• "Show me all available technical indicator functions"

Workbook generation

The generate_workbook tool creates .xlsx files with:

• Ticker sheets — per-ticker rows with formulas (portfolio dashboards)
• Array sheets — dynamic array functions (OHLCV, SMA, etc.) that spill automatically
• DataFrame pipelines — df → df.select → df.head → df.view chains
• Vega Charts — VEGACHART.LINE, VEGACHART.BAR, etc.

Predefined templates: xlquote_portfolio, xlquote_technical, xlquote_earnings, xlquote_risk.

Dynamic array support

Generated workbooks include XLDAPR metadata so dynamic array formulas spill correctly in Excel 365 — no @ implicit intersection operator.

Data sources & symbol formats

Symbol format varies by source. Use the get_symbol_format tool or ask your AI:

| Source | Example (Apple) | |--------|----------------| | Yahoo | AAPL | | TradingView | AMERICA/NASDAQ:AAPL | | Interactive Brokers | AAPL/SMART | | EOD Historical | AAPL.US | | Coinbase | BTC-USD |

18 sources supported — run get_symbol_format for the full list.

Requirements

• Node.js 18 or later
• Excel with XLQuote / DataFrame / Technicals / Vega Chart add-ins installed (formulas show #NAME? without them)

License

MIT