Tableau MCP

Tableau MCP

Tableau MCP is a developer toolkit that allows AI applications to seamlessly access Tableau published data sources and metadata via standardized, secure protocol endpoints. It enables LLM-powered tools to query, explore, and interact with Tableau data using a unified interface, facilitating integration with platforms like Claude, Cursor, and VSCode.

Author: tableau


View Protocol

What is Tableau MCP?

Tableau MCP is a suite of developer primitives—including tools, resources, and prompts—that enables AI-powered applications to integrate with Tableau by exposing published data sources and metadata through the Model Context Protocol. It standardizes access to Tableau data for large language models and AI assistants, supporting both Tableau Server and Tableau Cloud deployments.

How to Configure Tableau MCP

To configure Tableau MCP:

  • Ensure you have Node.js installed, or prepare to use Docker for deployment.
  • Clone the Tableau MCP repository and build the project using npm install and npm run build.
  • Set up required environment variables such as SERVER, SITE_NAME, PAT_NAME, and PAT_VALUE either directly in your config file (when running locally) or in a Docker env.list file.
  • For local use, add your MCP server details to the AI tool’s configuration, referencing the executable path and the appropriate environment variables.
  • For Docker, specify the Docker command, environment file path, and ensure your AI tool references the correct container configuration.
  • Optional environment variables (e.g., DATASOURCE_CREDENTIALS, DEFAULT_LOG_LEVEL, INCLUDE_TOOLS) can refine server behavior.
  • Use the MCP Inspector tool to verify your configuration is correct before integrating with client applications.

How to Use Tableau MCP

  • After configuration, launch Tableau MCP using Node.js or Docker.
  • Connect your MCP-compatible AI tool (like Claude Desktop, Cursor, or VSCode) to the running Tableau MCP instance by specifying it in the tool's configuration file (e.g., .cursor/mcp.json, .vscode/mcp.json).
  • AI tools will automatically list and discover available Tableau MCP tools via the MCP protocol.
  • Invoke supported operations, such as listing data sources, fetching field metadata, or running queries, by selecting the desired action within your AI tool’s UI or by using MCP Inspector.
  • Ensure that the authenticated Tableau user (via PAT) has the necessary permissions for the data sources you want to access.

Key Features

  • Seamless integration with published Tableau data sources on Tableau Server or Cloud.
  • Secure authentication using Tableau Personal Access Tokens.
  • Access to field-level metadata via the Tableau Metadata API.
  • Ability to run queries and retrieve data through VizQL Data Service (VDS) API.
  • Modular MCP tools for listing data sources, inspecting fields, running table queries, and more.
  • Multiple deployment modes: local Node.js or Docker container support.
  • Integration-ready with LLM applications and AI platforms that support MCP Tools.

Use Cases

  • Enable AI assistants to answer natural language questions using live Tableau data.
  • Automate dashboard or analytics workflows by programmatically retrieving data source metadata and running queries.
  • Allow LLMs to generate, validate, and explain Tableau queries based on user prompts.
  • Fast prototyping of data-driven features in chatbots, IDE plugins, or business productivity tools.

FAQ

1. What types of Tableau data can I access with Tableau MCP?

Tableau MCP only supports published data sources on Tableau Server or Tableau Cloud. It does not support workbooks or unpublished data.

2. How does authentication work?

You must provide a Tableau Personal Access Token (PAT) for authentication. The token must have access permissions for the desired data sources.

3. Which AI tools support Tableau MCP?

Any tool that implements the Model Context Protocol (MCP) can work with Tableau MCP. Popular examples include Claude Desktop, Cursor, and VSCode with MCP integration.

4. Can I run Tableau MCP with Docker?

Yes, Tableau MCP fully supports Docker deployment. Use the provided Dockerfile and set your environment variables in an env.list file.

5. Does Tableau MCP support access control and credential masking?

Yes, sensitive data is masked in logs by default, and credentials are securely managed via environment variables or Docker configuration.