dbt-mcp

Ask your dbt project what's wrong, in plain English. An MCP server that exposes a dbt project's run state as tools, so an AI assistant can compose its own answers to questions like "is the warehouse healthy?" or "what broke and why?" — no orchestration written by hand.

Built on dbt-sentinel, which does the artifact parsing and row sampling.

Tools

Related MCP server: @us-all/dbt-mcp

Quickstart

uv sync
export DBT_TARGET_DIR=/path/to/dbt/target
export DBT_DUCKDB_PATH=/path/to/warehouse.duckdb   # or BQ_PROJECT=my-project
uv run dbt-mcp

Inspect it interactively:

npx @modelcontextprotocol/inspector \
  -e DBT_TARGET_DIR=$DBT_TARGET_DIR \
  -e DBT_DUCKDB_PATH=$DBT_DUCKDB_PATH \
  uv run dbt-mcp

Configuration

Design decisions

Why MCP rather than a CLI. A CLI answers the question you anticipated. MCP tools let an agent compose answers to questions you didn't — it decides which tools to call and in what order.

Thin tools, not one god-tool. Each tool does one legible thing so the model can reason about when to use it. The docstrings are the interface: they become the tool descriptions the model reads.

Read-only by contract. The warehouse is opened read-only; this inspects, never mutates.

Errors are messages, not stack traces. A missing config returns "DBT_TARGET_DIR is not set; point it at a dbt target/ directory" — something an agent can act on.

Status

M1 complete: server, four tools, verified against a real dbt project via MCP Inspector. Next: explain_failure (grounded root-cause analysis), model_lineage, test_history.

Development

uv sync --group dev
uv run ruff check .
uv run pytest -v