Skip to main content
Glama

CSV MCP Server

CSV MCP Server

A Model Context Protocol (MCP) server for comprehensive CSV file management using stdio transport exclusively. This server provides tools for creating, editing, analyzing, and managing CSV files using the MCP protocol over standard input/output.

Features

  • File Management: Create, read, update, and delete CSV files
  • Absolute Path Support: Work with CSV files anywhere in the filesystem using absolute paths
  • Data Analysis: Basic statistical analysis and data exploration
  • Data Transformation: Filter, sort, group, and transform data
  • Data Validation: Check data integrity and format validation
  • Import/Export: Support for various CSV formats and encodings
  • Stdio Transport: Uses JSON-RPC 2.0 over standard input/output for communication

Installation

uv add csv-mcp-server

Usage

Running the Server

# Using stdio transport (default and only option) uv run csv-mcp-server # With custom log level uv run csv-mcp-server --log-level DEBUG # Development mode uv run mcp dev csv_mcp_server/server.py

Available Tools

  • create_csv: Create a new CSV file with headers and initial data
  • create_csv_at_path: Create a CSV file at a specific absolute or relative path
  • read_csv: Read and display CSV file contents
  • update_csv: Update specific cells or rows in a CSV file
  • delete_csv: Delete a CSV file
  • add_row: Add new rows to an existing CSV file
  • remove_row: Remove specific rows from a CSV file
  • get_info: Get basic information about a CSV file
  • get_statistics: Get statistical summary of numeric columns
  • filter_data: Filter CSV data based on conditions
  • sort_data: Sort CSV data by specified columns
  • group_data: Group and aggregate CSV data
  • validate_data: Validate CSV data integrity and format
  • get_path_info: Get detailed information about a file path (supports absolute paths)

Available Resources

  • csv://{filename}: Access CSV file contents as a resource
  • csv-info://{filename}: Get metadata about a CSV file

Available Prompts

  • analyze_csv: Generate analysis prompts for CSV data
  • transform_csv: Generate transformation suggestions

Configuration

The server can be configured with environment variables:

  • CSV_STORAGE_PATH: Base path for CSV file storage (default: current directory)
  • CSV_MAX_FILE_SIZE: Maximum file size in MB (default: 50)
  • CSV_BACKUP_ENABLED: Enable automatic backups (default: true)
  • CSV_SUPPORT_ABSOLUTE_PATHS: Enable absolute path support (default: true)

Absolute Path Support

The CSV MCP server now supports working with CSV files anywhere in the filesystem using absolute paths. This feature allows you to:

  • Create CSV files in any accessible directory
  • Read and modify existing CSV files from anywhere on the system
  • Work with files outside the default storage directory
  • Maintain backward compatibility with relative paths

Security Features

  • Path Validation: Automatically validates absolute paths for safety
  • System Directory Protection: Prevents access to critical system directories
  • Permission Checking: Verifies directory and file access permissions
  • Symlink Resolution: Safely resolves symbolic links to prevent path traversal attacks

Usage Examples

# Create a CSV file at an absolute path create_csv_at_path( filepath="/path/to/your/data/sales.csv", headers=["Date", "Product", "Sales"], data=[["2024-01-01", "Laptop", 1200]] ) # Get information about any file path get_path_info(filepath="/path/to/your/file.csv") # All existing tools work with absolute paths read_csv("/path/to/your/data/analysis.csv") update_csv("/path/to/your/data/analysis.csv", row_index=0, column="Sales", value=1500)

Transport

This server exclusively uses stdio transport with JSON-RPC 2.0 protocol, making it ideal for:

  • Integration with MCP clients that support stdio transport
  • Command-line tools and scripts
  • Development and testing environments
  • Containerized deployments

Examples

See the examples/ directory for usage examples with various MCP clients:

  • demo_client.py: Basic MCP client demonstration
  • sales_analysis.py: Sales data analysis example
  • absolute_path_demo.py: Demonstration of absolute path functionality
-
security - not tested
F
license - not found
-
quality - not tested

local-only server

The server can only run on the client's local machine because it depends on local resources.

Enables comprehensive CSV file management including creating, editing, analyzing, and transforming CSV data anywhere in the filesystem. Provides statistical analysis, data validation, filtering, and grouping capabilities through MCP protocol over stdio transport.

  1. Features
    1. Installation
      1. Usage
        1. Running the Server
        2. Available Tools
        3. Available Resources
        4. Available Prompts
      2. Configuration
        1. Absolute Path Support
          1. Security Features
          2. Usage Examples
        2. Transport
          1. Examples

            Related MCP Servers

            • A
              security
              F
              license
              A
              quality
              An MCP server that provides comprehensive Excel file operations, data analysis, and visualization capabilities for working with various spreadsheet formats like XLSX, CSV, and JSON.
              Last updated -
              8
              68
            • -
              security
              F
              license
              -
              quality
              An MCP server that allows LLMs to read, analyze, and interact with Excel files through file operations, data discovery, and comprehensive analysis tools.
              Last updated -
              1
            • A
              security
              A
              license
              A
              quality
              Comprehensive CSV processing MCP server with 40+ operations for data manipulation, analysis, and validation. Features auto-save, undo/redo, and handles GB+ files
              Last updated -
              39
              5
              MIT License
              • Apple

            View all related MCP servers

            MCP directory API

            We provide all the information about MCP servers via our MCP API.

            curl -X GET 'https://glama.ai/api/mcp/v1/servers/NovaAI-innovation/csv-mcp-server'

            If you have feedback or need assistance with the MCP directory API, please join our Discord server