haiiibin/data-profiler-mcp

Overview

An MCP server that profiles tabular data files such as CSV, Parquet, Excel, and JSON for use by LLM agents like Claude. It extracts structural information and schema details from local data files so agents can understand datasets before working with them.

Best for

Best for
Developers building MCP-connected LLM tools that need lightweight dataset introspection.

Use cases

• Provide an LLM agent with a quick schema summary of a CSV file before code generation
• Inspect Parquet or Excel file structure in an MCP-compatible assistant
• Prepare data context for automated data cleaning or transformation tasks

How to use

Install

pip install data-profiler-mcp

Tools exposed

• profile_dataset
• preview_data
• column_stats
• detect_quality_issues
• suggest_dtypes
• compare_datasets
• correlation_matrix

Tested with

Claude Desktop, Claude Code, Cursor

Notes

An MCP server that profiles tabular data files such as CSV, Parquet, Excel, and JSON for use by LLM agents like Claude. It extracts structural information and schema details from local data files so agents can understand datasets before working with them.

2 stars on GitHub. Last updated 2026-08-22. Licensed MIT.

Use cases

• Provide an LLM agent with a quick schema summary of a CSV file before code generation
• Inspect Parquet or Excel file structure in an MCP-compatible assistant
• Prepare data context for automated data cleaning or transformation tasks

Pros

• Supports multiple common tabular formats
• Integrates with MCP clients for agent workflows
• Simple Python implementation with a narrow focus

Cons

• Very low project traction (2 stars), suggesting limited adoption or maintenance
• No evidence of advanced profiling metrics beyond basic structure
• Scope is limited to profiling, not data transformation or analysis

Indexed from awesome-mcp-servers-punkpeye and enriched against its public facts.