tableau

This MCP server provides an integration layer between Tableau and GenAI applications using the MCP (Model Context Protocol). It enables Tableau API coverage, supports multiple authentication methods, and handles errors and retries with async capabilities for improved performance. You can deploy it as a Python-based MCP server and interact with Tableau through the server’s MCP endpoints or by using the provided Python client abstraction. Once running, you can initialize the server in your application and start processing MCP requests that involve Tableau data sources, visualizations, and API calls. The included environment variables allow you to configure Tableau API access securely, so you can connect to Tableau Cloud or Tableau Server and perform operations such as data source retrieval, workbook management, and embedding scenarios within GenAI-driven workflows.

Prerequisites:

• Python 3.8+ installed on the host
• Git installed (optional, for cloning)
• Access credentials for Tableau API (API key, OAuth credentials as needed)

Install from PyPI:

pip install tableau-mcp-server

Or install from source:

git clone https://github.com/LokiMCPUniverse/tableau-mcp-server.git
cd tableau-mcp-server
pip install -e .

Configure environment variables (examples):

export TABLEAU_API_BASE_URL="https://api.tableau.com"
export TABLEAU_API_KEY="<your-api-key>"
export TABLEAU_CLIENT_ID="<your-client-id>"
export TABLEAU_CLIENT_SECRET="<your-client-secret>"

Run the MCP server (as a module):

python -m tableau_mcp

Or, if you expose a different entry point, adjust accordingly to your deployment setup.

Tips and common considerations:

• Store sensitive Tableau credentials as environment variables or in a secure secret manager; avoid hard-coding keys.
• Ensure the appropriate Tableau API permissions are granted for the actions your MCP server will perform.
• The server is designed to run asynchronously; consider configuring appropriate concurrency settings and timeouts to balance throughput with rate limits.
• If you encounter authentication errors, verify that the base URL, API key, and OAuth credentials (if used) are correct and not expired.
• For production deployments, consider running behind a process manager (e.g., systemd, PM2) and using a reverse proxy for TLS termination.
• Review MCP request/response payloads for compatibility with your GenAI workflows and ensure proper error handling and retries are in place.