What is an MCP Server?

An MCP (Model Context Protocol) Server is a server that provides a standardized interface for LLMs to securely access and interact with external data sources and tools, such as BigQuery.

How does the BigQuery MCP Server work?

The BigQuery MCP Server acts as a bridge between LLMs and BigQuery datasets, translating natural language queries into SQL queries, retrieving data, and presenting it in a user-friendly format.

What are the benefits of using the BigQuery MCP Server?

The benefits include natural language querying, secure read-only access to data, support for tables and materialized views, dataset schema exploration, and integration with the UBOS platform.

What are the prerequisites for using the BigQuery MCP Server?

The prerequisites include Node.js 14 or higher, a Google Cloud project with BigQuery enabled, Google Cloud CLI or a service account key file, and Claude Desktop.

How do I install the BigQuery MCP Server?

You can install it either via Smithery using the command npx @smithery/cli install @ergut/mcp-bigquery-server --client claude or through manual configuration.

What permissions are needed to use the BigQuery MCP Server?

You need either roles/bigquery.user or both roles/bigquery.dataViewer and roles/bigquery.jobUser.

What are the current limitations of the BigQuery MCP Server?

The limitations include MCP support currently only available in Claude Desktop, connections limited to local MCP servers, read-only queries with a 1GB processing limit, and potential limitations with complex view types.

How does the BigQuery MCP Server integrate with the UBOS platform?

The BigQuery MCP Server seamlessly integrates with the UBOS platform to build powerful AI Agents that leverage BigQuery data, orchestrate AI Agents, connect to enterprise data, build custom AI Agents, and create Multi-Agent Systems.

What is the query limit?

The query limit is 1GB by default.

Is the data access read-only?

Yes, the data access is read-only, ensuring data integrity and compliance.