Provides tools for Great Expectations data quality validation through MCP, enabling AI agents to load datasets, define validation rules, and programmatically execute data quality checks.

Installation

Docker (Recommended)

docker run -d -p 8000:8000 --name gx-mcp-server davidf9999/gx-mcp-server:latest

From Source Code

git clone https://github.com/davidf9999/gx-mcp-server && cd gx-mcp-server
just install

PyPI

uv pip install gx-mcp-server

With Snowflake Support

uv pip install -e .[snowflake]

With BigQuery Support

uv pip install -e .[bigquery]

Configuration

STDIO Mode

{
  "mcpServers": {
    "gx-mcp-server": {
      "type": "stdio",
      "command": "uv",
      "args": ["run", "python", "-m", "gx_mcp_server"]
    }
  }
}

HTTP Mode with Authentication

{
  "mcpServers": {
    "gx-mcp-server": {
      "type": "http",
      "url": "https://your-server.com:8000/mcp/",
      "headers": {
        "Authorization": "Basic dXNlcjpwYXNz"
      }
    }
  }
}

Features

• Load CSV data from file, URL, or inline (up to 1 GB, configurable)
• Load tables from Snowflake or BigQuery using URI prefixes
• Define and modify ExpectationSuites
• Validate data and retrieve detailed results (synchronously or asynchronously)
• Choose in-memory storage (default) or SQLite for datasets and results
• Optional Basic authentication or Bearer token for HTTP clients
• Configure HTTP request rate limiting per minute
• Restrict sources via allowed-origins configuration
• Prometheus metrics on configurable metrics port
• OpenTelemetry tracing via OTLP exporter

Environment Variables

Optional

• MCP_SERVER_USER - Username for basic authentication
• MCP_SERVER_PASSWORD - Password for basic authentication
• MCP_CSV_SIZE_LIMIT_MB - CSV file size limit in MB (1-1024, default 50)
• GX_ANALYTICS_ENABLED - Disable anonymous Great Expectations usage data transmission (set to false)
• MCP_SERVER_URL - Server URL for custom clients
• MCP_AUTH_TOKEN - Authentication token for custom clients

Usage Examples

Load CSV data id,age\n1,25\n2,19\n3,45 and validate ages 21-65, show failed records

Notes

Supports both STDIO and HTTP transport modes. HTTP mode includes authentication options (Basic and Bearer JWT), rate limiting, CORS control, and health endpoints. Can run with Docker for easy deployment. Includes connectors for Snowflake and BigQuery data warehouses with special URI prefixes. Provides Prometheus metrics and OpenTelemetry tracing for production monitoring.