dbt MCP Server

What is dbt MCP Server?

The dbt MCP Server is a Model Context Protocol server that provides tools for AI agents to interact with dbt projects, including dbt Core, dbt Fusion, and dbt Platform. It enables agents to query metadata, generate SQL, run dbt commands, and manage dbt resources.

How to use dbt MCP Server?

Connect the dbt MCP server to an agent product like Claude or Cursor. An experimental MCP Bundle (dbt-mcp.mcpb) is available for MCPB-aware clients. For building custom agents, see the examples directory in the repository.

Key features of dbt MCP Server

• Execute and generate SQL on dbt Platform infrastructure
• Query dbt Semantic Layer metrics, dimensions, and entities
• Discover project metadata via the dbt Discovery API
• Run dbt CLI commands (build, compile, run, test, etc.)
• Manage dbt Platform jobs and runs via the Admin API
• Generate boilerplate code with dbt Codegen tools
• Trace column-level lineage using dbt LSP tools
• Search and fetch dbt product documentation

Use cases of dbt MCP Server

• An AI agent that answers questions about your dbt project's models, sources, and lineage
• A natural-language interface for generating SQL and running dbt commands
• A chatbot that monitors job runs, fetches artifacts, and triggers retries
• An assistant that helps write dbt documentation or generate model YAML files
• A tool that searches dbt documentation and returns relevant pages for troubleshooting

FAQ from dbt MCP Server

What tools does the dbt MCP Server provide?

Tools are organized into categories: SQL, Semantic Layer, Discovery, dbt CLI, Admin API, dbt Codegen, dbt LSP, Product Docs, and MCP Server Metadata. Each category provides several specific tools for interacting with dbt.

How do I install or set up the dbt MCP Server?

An experimental MCP Bundle (dbt-mcp.mcpb) is published with each release. Use Anthropic's mcpb CLI to install or inspect it. Alternatively, follow the examples directory to build a custom agent.

What dependencies does the server have?

Dependencies are pinned to specific versions and are not updated automatically. Only security-related dependency updates are submitted via automated pull requests.

Where can I get help or provide feedback?

Create a GitHub Issue or join the dbt community Slack in the #tools-dbt-mcp channel.

What data does the server access?

The server connects to dbt Platform infrastructure (dbt Cloud), dbt Core, and local project files. It can execute SQL, run dbt commands, and query dbt APIs (Semantic Layer, Discovery, Admin).