xiyan-table-mcp-server 0.6.0

A Model Context Protocol (MCP) server that allows AI assistants to display table contents and perform natural language queries.

Xiyan Table MCP Server

A Model Control Protocol (MCP) server for table data querying that allows users to configure local table data and query it using natural language.

Features

• Local CSV file configuration and reading
• Table data preview functionality (first 20 rows)
• Natural language querying support
• Intelligent responses powered by Large Language Models
• Flexible configuration using YAML
• Easy-to-use MCP tools and resources

Installation

1. Clone the repository:

git clone https://github.com/yourusername/xiyan-table-mcp-server.git
cd xiyan-table-mcp-server

2. Install dependencies:

pip install -r requirements.txt

Configuration

1. Copy the example configuration file:

cp config.yml.example config.yml

2. Edit config.yml with your settings:

table:
  path: "data/example.csv"  # Path to your CSV file
  encoding: "utf-8"         # File encoding
  preview_rows: 20          # Number of preview rows

model:
  type: "openai"           # Model type
  api_key: ""             # Your API key
  model_name: "gpt-4"     # Model name
  temperature: 0.1        # Temperature parameter

server:
  name: "xiyan-table"     # Server name
  version: "0.1.0"        # Version number

Usage

1. Start the server:

python server.py

2. Connect using an MCP client:

from mcp.client import MCPClient

# Connect to the server
client = MCPClient()

# Get table preview
preview = client.get_table_preview()
print(f"Table columns: {preview['columns']}")
print(f"Preview rows: {preview['preview_rows']}")

# Query data
result = client.query_table("What is the total number of rows in the table?")
print(f"Query result: {result}")

API Reference

Resources

get_table_preview

Returns a preview of the table data.

Response format:

{
    "columns": List[str],        # Column names
    "data": List[List[Any]],     # Table data
    "total_rows": int,           # Total number of rows
    "preview_rows": int          # Number of preview rows
}

Tools

query_table

Query the table using natural language.

Parameters:

• question (str): The natural language query

Returns:

• String response containing the answer to the query

Development

Requirements

• Python 3.8+
• pandas
• pyyaml
• openai
• mcp-core

Project Structure

xiyan-table-mcp-server/
├── config.yml          # Configuration file
├── server.py          # Main server implementation
├── requirements.txt   # Project dependencies
└── README.md         # Documentation

License

MIT License

Contributing

1. Fork the repository
2. Create your feature branch
3. Commit your changes
4. Push to the branch
5. Create a new Pull Request