Teradata

You can use the Teradata MCP Server to interact with Teradata databases through modular tools and prompts. It enables AI agents and you to query, analyze, and manage data efficiently by organizing capabilities into focused tool groups that align with common data workflows.

How to use

Connect to your Teradata MCP Server from an MCP-compatible client to start querying, analyzing, and managing data. Use the provided stdio configuration to run a local MCP server process that communicates with your Teradata instance. When you run the server through the configured command, you’ll access a collection of tool groups such as Search, Query, Table, Data Quality, DBA, Data Scientist, and BAR to perform end-to-end data operations.

How to install

Follow these steps to install and run the MCP server and connect to your Teradata platform.

Step 1. Prepare your Teradata access
- Have your Teradata database credentials ready.
- If you do not have a Teradata system, you can start with a sandbox or a test environment.

Step 2. Install prerequisites
- Install Claude Desktop to manage the MCP server quickly (no permanent installation required for quick tests).
- Install uv so you can run the MCP server through the management tool.
  - macOS: `brew install uv`
  - Windows: `pip install uv` (alternative to the installer)

Step 3. Configure the MCP server in Claude Desktop
- Create or edit `claude_desktop_config.json` and add the MCP server configuration shown below, updating the database credentials and host URL as needed.

{
  "mcpServers": {
    "teradata": {
      "command": "uvx",
      "args": ["teradata-mcp-server"],
      "env": {
        "DATABASE_URI": "teradata://<USERNAME>:<PASSWORD>@<HOST_URL>:1025/<USERNAME>"
      }
    }
  }
}

Additional notes

Prerequisites include Teradata credentials or access to a sandbox, Claude Desktop, and uv. The MCP server is started via a local process, using the uvx command with the teradata-mcp-server argument and the DATABASE_URI environment variable that points to your Teradata instance.

Troubleshooting and tips

- Ensure your Teradata credentials are correct and the host URL is reachable from your environment.
- If the MCP server fails to start, verify that `uvx` and `teradata-mcp-server` are correctly installed and that the environment variable `DATABASE_URI` is properly formatted.
- When testing, you can use a sandbox environment to validate tool behavior before connecting to production data.

Notes on usage scope

The Teradata MCP Server provides grouped tools and prompts for searching vector stores, querying Teradata data, working with feature stores and semantic definitions, performing data quality checks, administering the platform, building AI-driven data workflows, and handling backup and restore tasks through BAR integrations.

Available tools

Search tools

Tools, prompts, and resources to search and manage vector stores, enabling rapid retrieval and organization of data across Teradata platforms.

Query tools

Base querying tools to navigate Teradata databases and extract structured information.

Table tools

Access to structured data models, including features for the Teradata Enterprise Feature Store and semantic layer definitions.

Data Quality tools

Prompts and resources to accelerate exploratory data analysis and quality checks on data sets.

DBA tools

Prompts and resources to facilitate platform administration tasks and security management.

Data Scientist tools

Tools and prompts to build AI agents and data-driven workflows, including Teradata Vector Store Tools and Teradataml Functions Tools.

BAR tools

Backup and restore operations integrated with Teradata DSA across multiple storage solutions.