mcp-server-spreadsheet and mcp-excel
About mcp-server-spreadsheet
marekrost/mcp-server-spreadsheet
Data-first MCP server for reading and writing spreadsheet files (.xlsx, .csv, .ods) - cell-level ops + DuckDB SQL engine
Implements MCP (Model Context Protocol) as stdio transport to integrate with Claude Desktop and Claude Code, enabling AI assistants to perform spreadsheet operations. Combines cell-level workbook manipulation (insert/delete rows, copy ranges across sheets) with in-memory DuckDB SQL queries that support JOINs and mutations written back to disk. Uses stateless, explicit file/sheet parameters with atomic temp-file writes and automatic type coercion, treating CSV as single-sheet workbooks while preserving native typing across Excel and OpenDocument formats.
About mcp-excel
jwadow/mcp-excel
MCP server for AI agents to analyze Excel spreadsheets through atomic operations. Like SQL for Excel. Fast, accurate, and efficient. No context overflow.
Implements atomic query operations (filter, aggregate, group_by, rank) that return precise results instead of raw data, preventing context overflow when analyzing large Excel files. Built on Python with openpyxl for .xls/.xlsx support, it integrates with MCP-compatible AI agents (Claude, Cursor, Cline) via stdio transport, sending only computed results and metadata to the model while keeping bulk data local. Supports 12 filter operators, 8 aggregation functions, statistical analysis, and generates Excel formulas for reproducible results.
Related comparisons
Scores updated daily from GitHub, PyPI, and npm data. How scores work