Excel MCP Server

An MCP server that brings Excel/CSV data analysis capabilities to Claude, with a custom-built formula engine and AI-powered features.

What makes this different: The world's first conversational data analysis MCP server powered by HyperFormula (395 Excel functions) + multi-provider AI. Ask questions in natural language and get instant insights from your Excel/CSV files. Built for analysts who want AI-powered conversations with their data, not just spreadsheet manipulation.

How It Works

This server acts as a bridge between Claude and your data files, providing:

Formula Engine: A complete Excel-compatible formula parser and evaluator built from scratch, supporting 82 functions with proper operator precedence and dependency tracking.
Data Operations: Direct read/write access to CSV and Excel files with intelligent parsing and type detection.
AI Integration: Multi-provider AI support that intelligently routes requests to the best available AI service for natural language processing and formula generation.
Bulk Processing: Parallel operations across multiple files for improved performance on large datasets.

Features

395 Excel functions via HyperFormula (battle-tested library)
AI-powered natural language data queries
Multi-provider AI support (Anthropic, OpenAI, DeepSeek, Gemini)
Large file support with intelligent chunking for files >100MB
Bulk operations across multiple files (3x faster)
Smart data validation and consistency checking
Advanced statistical analysis and data profiling
Formula evaluation with dependency tracking
Pivot tables and aggregation
Cross-file referential integrity validation
Excel styling and formatting with auto-fitting columns and cell formatting

🏢 CFO-Level Financial Analysis

DCF Valuation Models - Complete discounted cash flow analysis with NPV, IRR, terminal value
Financial Ratio Analysis - Comprehensive ratios with industry benchmarks (liquidity, profitability, leverage)
Budget Variance Analysis - Automated budget vs actual comparisons with variance reporting
Scenario Modeling - What-if analysis with multiple assumption scenarios
Financial Functions - NPV, IRR, PMT, FV, PV, RATE, depreciation schedules, and more
Industry Benchmarking - Automatic comparison against industry standards

Installation

git clone https://github.com/ishayoyo/excel-mcp.git
cd excel-mcp
npm install
npm run build

Claude Code Setup

Option 1: Direct MCP Configuration

# Add to Claude Code
claude mcp add excel-csv "node" "C:/path/to/excel-mcp/dist/index.js"

Option 2: Global Install

npm install -g .
# Then configure in Claude Code with:
# Command: npx
# Args: -y excel-csv-mcp

Cursor Claude Desktop Setup

Option 1: Direct MCP Configuration

Edit your Cursor MCP configuration file (%APPDATA%\Cursor\mcp.json on Windows, ~/Library/Application Support/Cursor/mcp.json on macOS):

{
  "mcpServers": {
    "excel-csv": {
      "command": "node",
      "args": ["C:/path/to/excel-mcp/dist/index.js"]
    }
  }
}

Replace C:/path/to/excel-mcp/dist/index.js with the actual path to your built dist/index.js file.

Option 2: Using npx (Recommended)

{
  "mcpServers": {
    "excel-csv": {
      "command": "npx",
      "args": ["-y", "excel-csv-mcp"]
    }
  }
}

This requires a global npm install:

npm install -g .

AI Providers

Create a .env file with your API keys:

cp .env.example .env

ANTHROPIC_API_KEY=your-key-here
OPENAI_API_KEY=your-key-here
DEEPSEEK_API_KEY=your-key-here
GEMINI_API_KEY=your-key-here

Supported providers: Anthropic, OpenAI, DeepSeek, Gemini, Local fallback

Usage Examples

Ask questions in natural language:

"What's the total revenue in sales_data.csv?"
"Sum all sales where region is 'North'"
"Find duplicates in customer_data.xlsx"
"Clean messy customer data - fix phone numbers and names"
"Remove duplicate rows from my contact list"
"Help me VLOOKUP customer details between two files"
"Generate a formula to calculate average of last 30 days"
"Validate data consistency between sales.xlsx and branches.xlsx"
"Create a pivot table by department and export to report.xlsx"
"Analyze this 500MB dataset in chunks to avoid memory issues"
"Get file info and optimal chunk size for large_dataset.csv"

Available Tools

Core Operations

read_file - Read CSV/Excel files (supports offset/limit for large files)
read_file_chunked - Stream large files in manageable chunks to avoid memory limits
get_file_info - Analyze file size and get chunking recommendations for large datasets
get_cell / get_range - Access specific cells or ranges
search / filter_rows - Find and filter data
aggregate - SUM, AVG, COUNT, MIN, MAX operations

Excel Workflow Tools (NEW)

find_duplicates - Smart duplicate detection and removal with multiple strategies
data_cleaner - Batch data cleaning operations (trim whitespace, fix phones, standardize names, etc.)
vlookup_helper - Intelligent VLOOKUP setup with fuzzy matching and error handling

Bulk Operations

bulk_aggregate_multi_files - Aggregate across multiple files (3x faster)
bulk_filter_multi_files - Filter multiple files simultaneously

Formula Engine

evaluate_formula - Execute Excel formulas
parse_natural_language - Convert text to formulas
explain_formula - Explain formula functionality

Analytics

statistical_analysis - Comprehensive statistics
correlation_analysis - Find relationships between columns
data_profile - Complete data profiling
pivot_table - Group and aggregate data

AI Features

smart_data_analysis - AI-powered analysis suggestions
ai_provider_status - Check AI provider availability

Styling & Formatting

format_cells - Apply fonts, colors, borders, alignment to Excel cells
auto_fit_columns - Automatically adjust column widths to fit content
write_multi_sheet - Create complex Excel files with formulas and formatting

Validation & Export

validate_data_consistency - Cross-file data integrity checks
write_file / export_analysis - Create and export files

Key Features

🤖 AI-Powered Intelligence

Ask "What's the correlation between sales and marketing spend?" → Gets instant statistical analysis
Say "Find all customers who spent over $1000 last quarter" → AI generates complex formulas automatically
Request "Explain what this VLOOKUP formula does" → Gets plain English explanation

⚡ Bulk Data Processing

"Sum total revenue across 10 quarterly files" → Processes in parallel (3x faster than manual)
"Find all employees with 'Manager' in their title across 5 HR spreadsheets" → Multi-file search
"Aggregate sales data from Q1, Q2, Q3, Q4 files into one report" → Automated consolidation

🧮 Programmatic Formula Engine

Evaluates =SUMIFS(C:C, A:A, ">=1/1/2024") without Excel installed
Computes complex nested formulas with proper precedence
Handles circular references and dependency tracking
82 functions implemented programmatically

📊 Advanced Analytics

Comprehensive statistical profiling of every column
Correlation analysis between any numeric columns
Smart data quality validation across multiple files
Automated outlier detection and data consistency checks

🔍 Smart Cross-File Operations

"Validate that all customer IDs in sales.xlsx exist in customers.xlsx"
"Check for duplicate emails across 3 contact lists"
"Find missing data between related spreadsheets"
Referential integrity validation (like database foreign keys but for spreadsheets)

🎯 Natural Language Data Conversations

"Show me the top 10 products by revenue" → AI understands intent and executes
"Calculate average order value for customers from California" → Generates optimal query
"Create a summary of sales trends by month" → Suggests best analysis approach

📈 Large File Processing (NEW)

Intelligent Chunking: Automatically handles files >100MB without memory errors
Smart Recommendations: get_file_info analyzes files and suggests optimal chunk sizes
Seamless Navigation: Read any chunk with proper headers and metadata
Memory Efficient: Constant memory usage regardless of file size (tested up to 1M+ rows)
Progress Tracking: Chunk navigation with hasNext/hasPrevious indicators

Why This Approach Matters

Traditional Excel tools focus on file manipulation - reading, writing, formatting. They treat spreadsheets as static documents.

This server treats your data as a dynamic knowledge base you can have conversations with. It's not about pretty formatting - it's about CFO-level financial modeling, strategic analysis, and executive decision-making powered by AI.

Built for CFOs, financial analysts, and executives who need: 💰 DCF valuations and investment analysis 📊 Financial ratio analysis with benchmarks 🎯 Scenario planning and risk modeling 💼 Budget variance tracking and reporting 🤖 AI-powered financial insights at conversational speed

License

MIT

Excel MCP Server

An MCP server that brings Excel/CSV data analysis capabilities to Claude, with a custom-built formula engine and AI-powered features.

How It Works

This server acts as a bridge between Claude and your data files, providing:

Formula Engine: A complete Excel-compatible formula parser and evaluator built from scratch, supporting 82 functions with proper operator precedence and dependency tracking.
Data Operations: Direct read/write access to CSV and Excel files with intelligent parsing and type detection.
AI Integration: Multi-provider AI support that intelligently routes requests to the best available AI service for natural language processing and formula generation.
Bulk Processing: Parallel operations across multiple files for improved performance on large datasets.

Features

395 Excel functions via HyperFormula (battle-tested library)
AI-powered natural language data queries
Multi-provider AI support (Anthropic, OpenAI, DeepSeek, Gemini)
Large file support with intelligent chunking for files >100MB
Bulk operations across multiple files (3x faster)
Smart data validation and consistency checking
Advanced statistical analysis and data profiling
Formula evaluation with dependency tracking
Pivot tables and aggregation
Cross-file referential integrity validation
Excel styling and formatting with auto-fitting columns and cell formatting

🏢 CFO-Level Financial Analysis

DCF Valuation Models - Complete discounted cash flow analysis with NPV, IRR, terminal value
Financial Ratio Analysis - Comprehensive ratios with industry benchmarks (liquidity, profitability, leverage)
Budget Variance Analysis - Automated budget vs actual comparisons with variance reporting
Scenario Modeling - What-if analysis with multiple assumption scenarios
Financial Functions - NPV, IRR, PMT, FV, PV, RATE, depreciation schedules, and more
Industry Benchmarking - Automatic comparison against industry standards

Installation

git clone https://github.com/ishayoyo/excel-mcp.git
cd excel-mcp
npm install
npm run build

Claude Code Setup

Option 1: Direct MCP Configuration

# Add to Claude Code
claude mcp add excel-csv "node" "C:/path/to/excel-mcp/dist/index.js"

Option 2: Global Install

npm install -g .
# Then configure in Claude Code with:
# Command: npx
# Args: -y excel-csv-mcp

Cursor Claude Desktop Setup

Option 1: Direct MCP Configuration

Edit your Cursor MCP configuration file (%APPDATA%\Cursor\mcp.json on Windows, ~/Library/Application Support/Cursor/mcp.json on macOS):

{
  "mcpServers": {
    "excel-csv": {
      "command": "node",
      "args": ["C:/path/to/excel-mcp/dist/index.js"]
    }
  }
}

Replace C:/path/to/excel-mcp/dist/index.js with the actual path to your built dist/index.js file.

Option 2: Using npx (Recommended)

{
  "mcpServers": {
    "excel-csv": {
      "command": "npx",
      "args": ["-y", "excel-csv-mcp"]
    }
  }
}

This requires a global npm install:

npm install -g .

AI Providers

Create a .env file with your API keys:

cp .env.example .env

ANTHROPIC_API_KEY=your-key-here
OPENAI_API_KEY=your-key-here
DEEPSEEK_API_KEY=your-key-here
GEMINI_API_KEY=your-key-here

Supported providers: Anthropic, OpenAI, DeepSeek, Gemini, Local fallback

Usage Examples

Ask questions in natural language:

"What's the total revenue in sales_data.csv?"
"Sum all sales where region is 'North'"
"Find duplicates in customer_data.xlsx"
"Clean messy customer data - fix phone numbers and names"
"Remove duplicate rows from my contact list"
"Help me VLOOKUP customer details between two files"
"Generate a formula to calculate average of last 30 days"
"Validate data consistency between sales.xlsx and branches.xlsx"
"Create a pivot table by department and export to report.xlsx"
"Analyze this 500MB dataset in chunks to avoid memory issues"
"Get file info and optimal chunk size for large_dataset.csv"

Available Tools

Core Operations

read_file - Read CSV/Excel files (supports offset/limit for large files)
read_file_chunked - Stream large files in manageable chunks to avoid memory limits
get_file_info - Analyze file size and get chunking recommendations for large datasets
get_cell / get_range - Access specific cells or ranges
search / filter_rows - Find and filter data
aggregate - SUM, AVG, COUNT, MIN, MAX operations

Excel Workflow Tools (NEW)

find_duplicates - Smart duplicate detection and removal with multiple strategies
data_cleaner - Batch data cleaning operations (trim whitespace, fix phones, standardize names, etc.)
vlookup_helper - Intelligent VLOOKUP setup with fuzzy matching and error handling

Bulk Operations

bulk_aggregate_multi_files - Aggregate across multiple files (3x faster)
bulk_filter_multi_files - Filter multiple files simultaneously

Formula Engine

evaluate_formula - Execute Excel formulas
parse_natural_language - Convert text to formulas
explain_formula - Explain formula functionality

Analytics

statistical_analysis - Comprehensive statistics
correlation_analysis - Find relationships between columns
data_profile - Complete data profiling
pivot_table - Group and aggregate data

AI Features

smart_data_analysis - AI-powered analysis suggestions
ai_provider_status - Check AI provider availability

Styling & Formatting

format_cells - Apply fonts, colors, borders, alignment to Excel cells
auto_fit_columns - Automatically adjust column widths to fit content
write_multi_sheet - Create complex Excel files with formulas and formatting

Validation & Export

validate_data_consistency - Cross-file data integrity checks
write_file / export_analysis - Create and export files

Key Features

🤖 AI-Powered Intelligence

Ask "What's the correlation between sales and marketing spend?" → Gets instant statistical analysis
Say "Find all customers who spent over $1000 last quarter" → AI generates complex formulas automatically
Request "Explain what this VLOOKUP formula does" → Gets plain English explanation

⚡ Bulk Data Processing

"Sum total revenue across 10 quarterly files" → Processes in parallel (3x faster than manual)
"Find all employees with 'Manager' in their title across 5 HR spreadsheets" → Multi-file search
"Aggregate sales data from Q1, Q2, Q3, Q4 files into one report" → Automated consolidation

🧮 Programmatic Formula Engine

Evaluates =SUMIFS(C:C, A:A, ">=1/1/2024") without Excel installed
Computes complex nested formulas with proper precedence
Handles circular references and dependency tracking
82 functions implemented programmatically

📊 Advanced Analytics

Comprehensive statistical profiling of every column
Correlation analysis between any numeric columns
Smart data quality validation across multiple files
Automated outlier detection and data consistency checks

🔍 Smart Cross-File Operations

"Validate that all customer IDs in sales.xlsx exist in customers.xlsx"
"Check for duplicate emails across 3 contact lists"
"Find missing data between related spreadsheets"
Referential integrity validation (like database foreign keys but for spreadsheets)

🎯 Natural Language Data Conversations

"Show me the top 10 products by revenue" → AI understands intent and executes
"Calculate average order value for customers from California" → Generates optimal query
"Create a summary of sales trends by month" → Suggests best analysis approach

📈 Large File Processing (NEW)

Intelligent Chunking: Automatically handles files >100MB without memory errors
Smart Recommendations: get_file_info analyzes files and suggests optimal chunk sizes
Seamless Navigation: Read any chunk with proper headers and metadata
Memory Efficient: Constant memory usage regardless of file size (tested up to 1M+ rows)
Progress Tracking: Chunk navigation with hasNext/hasPrevious indicators

Why This Approach Matters

Traditional Excel tools focus on file manipulation - reading, writing, formatting. They treat spreadsheets as static documents.

License

MIT