dbt

by dbt Labs

v1.22.1

Apache License Version 2.

Repository Docs

Databases & Data

datadbtanalyticssemantic-layer

Summary

dbt Labs' official MCP server — explore project metadata, run dbt commands, and query the Semantic Layer.

Features

Explore models, sources, and lineage in a dbt project
Run dbt commands and read the results
Query the Semantic Layer for governed metrics

Installation

Set up this MCP server in your favorite AI agent — copy a ready-made configuration below.

Any MCP-compatible agent

Most agents (Claude, Cursor, Windsurf, VS Code, and more) read a standard mcpServers configuration.

Open your agent's MCP configuration file.
Merge the snippet below into it.
Restart the agent — the "dbt" tools will be available.

{
  "mcpServers": {
    "dbt": {
      "command": "uvx",
      "args": [
        "dbt-mcp"
      ]
    }
  }
}

Claude Code

Add the server with a single CLI command.

claude mcp add dbt -- uvx dbt-mcp

Claude Desktop

Add the server to claude_desktop_config.json (Settings → Developer → Edit Config).

{
  "mcpServers": {
    "dbt": {
      "command": "uvx",
      "args": [
        "dbt-mcp"
      ]
    }
  }
}

Cursor

{
  "command": "uvx",
  "args": [
    "dbt-mcp"
  ]
}

Windsurf

{
  "command": "uvx",
  "args": [
    "dbt-mcp"
  ]
}

VS Code

{
  "command": "uvx",
  "args": [
    "dbt-mcp"
  ]
}

Description

dbt Labs' official MCP server connects an assistant to a dbt project and, if you use dbt Cloud, to its metadata and Semantic Layer.

Three areas of capability

Project discovery. List models, sources, and tests; read model definitions; and trace lineage upstream and downstream. This is the fastest way to answer "what feeds this table and what breaks if I change it".
Command execution. Run dbt build, run, test, and compile and read the output, so an assistant can iterate on a model until its tests pass.
Semantic Layer queries. Ask for a metric by name and get governed, consistent numbers rather than a hand-written SQL approximation that quietly disagrees with the dashboard.

Why the Semantic Layer part matters

The recurring problem with AI querying a warehouse is that it invents its own definition of a metric. Routing through the Semantic Layer means "revenue" means what your organisation has agreed it means.

Installed with uvx dbt-mcp, configured with your project directory and, optionally, dbt Cloud credentials.

Related MCP Servers

Featured

DBHub

by Bytebase

A universal database MCP server — one integration for PostgreSQL, MySQL, SQL Server, SQLite, and MariaDB.

Databases & Data

databasesqlmulti-database

Featured

MongoDB

by MongoDB

MongoDB's official MCP server — query databases, explore schemas, and manage Atlas clusters from an AI client.

Databases & Data

databasenosqlmongodb

Upstash Redis

by Upstash

Run Redis commands against Upstash's serverless Redis over HTTP, with documentation search built in.

Databases & Data

redisserverlesscache

Milvus

by Zilliz

Search and manage Milvus vector collections from an AI client — vector, full-text, and hybrid queries.

Databases & Data

vector-databasemilvussearch

Browse all MCP servers →