deepdiff-mcp

abnerjacobsen/deepdiff-mcp

3.2

If you are the rightful owner of deepdiff-mcp and would like to certify it and/or have it hosted online, please leave a comment on the right or send an email to henry@mcphub.com.

The DeepDiff MCP Server allows users to leverage the DeepDiff library's functionality through the Model Context Protocol, facilitating seamless interaction with DeepDiff from any MCP client.

Tools

  1. compare

    Compare two objects and return their differences.

  2. get_deep_distance

    Calculate the deep distance between two objects.

  3. search

    Search for an item within an object.

  4. grep

    Search for an item within an object using grep-like behavior.

  5. hash_object

    Hash an object based on its contents.

  6. create_delta

    Create a delta that can transform one object into another.

  7. apply_delta

    Apply a delta to transform an object.

  8. extract_path

    Extract a value from an object using a path.

  9. compare_files

    Compare two CSV, Excel, or JSON files directly.

DeepDiff MCP Server

An MCP (Model Context Protocol) server for the DeepDiff library. This server allows you to use DeepDiff functionality through the Model Context Protocol, enabling interaction with DeepDiff from any MCP client.

Installation

Using pip

pip install deepdiff-mcp

Using uv (recommended)

uv pip install deepdiff-mcp

Usage

Running the server

You can run the DeepDiff MCP server in several ways:

As a command-line tool
# Using stdio transport (default)
deepdiff-mcp

# Using HTTP transport
deepdiff-mcp --transport http --host 127.0.0.1 --port 8000

# Using SSE transport
deepdiff-mcp --transport sse --host 127.0.0.1 --port 8000
As a Python module
from deepdiff_mcp import create_server

# Create a DeepDiff MCP server
server = create_server("My DeepDiff Server")

# Run the server with stdio transport (default)
server.run()

# Or with HTTP transport
server.run(transport="http", host="127.0.0.1", port=8000, path="/mcp")

# Or with SSE transport
server.run(transport="sse", host="127.0.0.1", port=8000)

Connecting to the server

You can connect to the server using any MCP client. Here's an example using FastMCP:

import asyncio
from fastmcp import Client

async def main():
    # Connect to a DeepDiff MCP server running as a subprocess
    async with Client("deepdiff-mcp") as client:
        # List available tools
        tools = await client.list_tools()
        print(f"Available tools: {[tool.name for tool in tools]}")
        
        # Compare two dictionaries
        t1 = {"a": 1, "b": 2}
        t2 = {"a": 1, "b": 3, "c": 4}
        
        result = await client.call_tool("compare", {"t1": t1, "t2": t2})
        print(f"Differences: {result.text}")

if __name__ == "__main__":
    asyncio.run(main())

Available Tools

  • compare_files - Compare two CSV, Excel, or JSON files directly

The DeepDiff MCP server provides the following tools:

  • compare - Compare two objects and return their differences
  • get_deep_distance - Calculate the deep distance between two objects
  • search - Search for an item within an object
  • grep - Search for an item within an object using grep-like behavior
  • hash_object - Hash an object based on its contents
  • create_delta - Create a delta that can transform one object into another
  • apply_delta - Apply a delta to transform an object
  • extract_path - Extract a value from an object using a path

Documentation

For more information about DeepDiff, see the DeepDiff documentation.

For more information about MCP and FastMCP, see the Model Context Protocol and FastMCP documentation.

Comparando Arquivos CSV e Excel

O DeepDiff MCP suporta comparação direta de arquivos CSV, Excel e JSON:

import asyncio
from fastmcp import Client

async def main():
    async with Client("deepdiff-mcp") as client:
        # Compare dois arquivos CSV
        result = await client.call_tool(
            "compare_files", 
            {
                "file1_path": "caminho/para/arquivo1.csv", 
                "file2_path": "caminho/para/arquivo2.csv",
                "ignore_order": True  # Ignorar a ordem das linhas
            }
        )
        
        print(f"Diferenças: {result.text}")
        
        # Compare arquivos Excel
        excel_result = await client.call_tool(
            "compare_files", 
            {
                "file1_path": "caminho/para/arquivo1.xlsx", 
                "file2_path": "caminho/para/arquivo2.xlsx"
            }
        )
        
        print(f"Diferenças: {excel_result.text}")

if __name__ == "__main__":
    asyncio.run(main())

Esta funcionalidade é especialmente útil para análise de dados e verificação de integridade de dados em pipelines.