mcp-registry/dbt-mcp

    ==================
      
       /// MCP ///
      /// DBT ///
        
    ==================
        
    [server:online]
    [protocol:ready]

dbt-mcp

by dbt-labs

An MCP (Model Context Protocol) server that exposes dbt Core/Cloud functionality – CLI commands, Semantic Layer, model discovery and SQL execution – to any MCP-capable client.

325
38
Open Source
GitHub

Installation

1. Prerequisites:
• Python 3.9 or newer
• A working dbt profile (normally located in ~/.dbt/)
2. Install from PyPI (preferred – if the package is published):
pip install dbt-mcp
OR install from source:
# clone the repo git clone https://github.com/dbt-labs/dbt-mcp.git cd dbt-mcp # create isolated environment (recommended) python -m venv .venv source .venv/bin/activate # install dependencies and package in editable mode pip install -U pip pip install -e .
3. Start the MCP server (typical pattern – confirm with repo docs):
dbt-mcp serve --port 8000 --profiles-dir ~/.dbt
The server will listen on http://localhost:8000 by default.
4. Configuration (commonly required):
• Ensure the environment variables your dbt project needs (e.g., DATABASE credentials) are exported.
• If you keep multiple dbt projects, pass --project-dir <path>.
5. Test that the server is running:
curl http://localhost:8000/health

Documentation

License: Apache License 2.0
Updated 7/30/2025

Table of Contents