Basedash

What the Basedash MCP server does

The Basedash MCP server exposes a remote MCP interface to a Basedash workspace. It lets connected AI clients work with live company data while keeping question answering governed through Basedash. The server supports both data discovery and management of dashboards and charts.

The ask_question tool accepts a question about live workspace data and returns a governed answer. get_data_sources shows which data sources are available in the workspace, helping an agent understand what information it can query.

For reporting workflows, the Basedash MCP server can create and update dashboards and charts. Dashboard operations return a durable Basedash URL. Chart creation and editing return a Basedash URL and may also return a screenshot link when one is available.

How it works

The server is hosted at:

https://charts.basedash.com/api/public/mcp

It communicates over Streamable HTTP, so an MCP client connects to the remote endpoint rather than running a local server process. Authentication uses a browser-based Basedash OAuth flow. No API key is required according to the provided connection details.

After the client connects and the OAuth prompt is completed, the available tools can query the workspace, inspect existing visualization objects, or create and modify them. The service is published in the official MCP Registry under com.basedash/mcp.

Setup and configuration

Configure an MCP client with the Basedash endpoint and complete the browser OAuth prompt when requested. The README names Claude, Cursor, ChatGPT, and other MCP clients as connection targets, but it does not provide client-specific configuration files or commands.

No environment variables or API keys are documented for this server. The required connection details are the remote endpoint and the OAuth authorization step.

Tools and capabilities

The Basedash MCP server exposes these tools:

• ask_question: Ask about live workspace data and receive a governed answer.
• get_data_sources: List the workspace's available data sources.
• create_dashboard: Create a dashboard and return its durable Basedash URL.
• edit_dashboard: Edit a dashboard and return its durable Basedash URL.
• create_chart: Create a chart, optionally attaching it to a dashboard; returns a Basedash URL and a screenshot link when available.
• edit_chart: Edit a chart; returns a Basedash URL and a screenshot link when available.
• list_dashboards: List dashboards in the workspace.
• get_dashboard: Retrieve a dashboard together with its charts.
• list_charts: List charts in the workspace.
• get_chart: Retrieve a chart and its details.

Limitations and notes

The supplied material describes access to live data in a Basedash workspace, not a general-purpose database connector. The available data sources depend on the connected workspace. Chart screenshot links are conditional and may not be returned for every operation. The documentation also does not specify client-specific setup instructions, permissions behavior, query syntax, or local deployment options.