mcp-excel-reader

This MCP server provides tools for reading data from Excel (.xlsx) files and returning the results in JSON format. It exposes three primary tools. read_excel reads content from all sheets in a workbook, returning each sheet as a separate JSON array under the sheet name. read_excel_by_sheet_name reads a specific sheet by its name (or the first sheet if no name is provided). read_excel_by_sheet_index reads a sheet by its index (or the first sheet if no index is provided). All returned values are converted to strings, and empty cells are represented as empty strings. Use these tools to quickly extract structured data from Excel workbooks for downstream processing, validation, or integration with other MCP-enabled services.

Prerequisites:

• Python 3.10 or higher
• Internet access to install dependencies

Installation steps (recommended):

1. Install the server package using uv (recommended):

uvx excel-reader-server

2. Alternatively, install with pip and run directly (not using uv):

pip install excel-reader-server

Usage notes:
- The package provides an MCP-compatible server exposing read_excel, read_excel_by_sheet_name, and read_excel_by_sheet_index tools.
- Ensure you have a valid Excel file accessible by the path you provide to the tools.
- The server depends on openpyxl >= 3.1.5 for Excel reading.

Tips and common issues:

• If a sheet name is not found when using read_excel_by_sheet_name, the server will return a descriptive error indicating the invalid sheet name. Consider listing available sheet names from the workbook first.
• When using read_excel_by_sheet_index, indices are zero-based. If an index is out of range, you’ll receive a clear error.
• All data is returned as strings; you may need to cast types on the client side.
• Ensure the Excel file is not password-protected or otherwise restricted, as password-protected files may not be supported.
• If you encounter performance issues with very large workbooks, consider reading only the required sheets by name or index to reduce memory usage.
• The server is designed to work with xlsx files; other formats (xls) may not be supported.