dbt Language Server

Thin VS Code client for dbt-ls. The extension only launches the Python language server and connects to it over stdio; all features (ref/source completion, go-to-definition, database-aware column info) live in the server.

Install

The server is a Python package, so installation is two parts:

1. Install the server (requires Python 3.10+)

uv tool install dbt-ls
# or: uv tool install "dbt-ls[postgres]"

Pick the extra that matches your warehouse: duckdb, postgres, mysql, sqlserver, pyspark, databricks, aws, or all.

Verify it is on your PATH:

dbt-ls --help

2. Install the extension

• From the Marketplace (once published): search "dbt Language Server" publshed by Datanuggets.

If dbt-ls is not on your PATH (e.g. it lives in a project venv), set dbtLs.serverCommand to its absolute path in your VS Code settings:

{
  "dbtLs.serverCommand": "/path/to/venv/bin/dbt-ls"
}

Multiple dbt projects in one workspace

A workspace may contain any number of dbt projects (project/dbt_project1, project/dbt_project2, …). The extension starts one server per dbt_project.yml, lazily, the first time you open a SQL file belonging to it — so each file gets completions from its own project. Each server logs to its own dbt-ls (<project>) output channel.

SQL files that are not inside a dbt project do not start a server.

Settings

Develop / package

cd editors/vscode
npm install
npm install -g @vscode/vsce

# Try it: open this folder in VS Code and press F5 (Extension Development Host).

vsce package        # -> dbt-ls-<version>.vsix
vsce publish        # -> VS Code Marketplace (needs publisher + PAT)
# ovsx publish      # -> Open VSX (VSCodium / Cursor)