Skip to content

Latest commit

 

History

History
48 lines (34 loc) · 849 Bytes

File metadata and controls

48 lines (34 loc) · 849 Bytes

6. Run DataLex with Docker

Docker is optional. Use it when local Python or Node setup is blocked.

Build the image

git clone https://github.com/duckcode-ai/DataLex.git
cd DataLex
docker build -t datalex:local .

Run the app

docker run --rm -p 3030:3001 datalex:local

Open:

http://localhost:3030

Run against an existing dbt repo

cd ~/path/to/your-dbt-project
docker run --rm -p 3030:3001 \
  -v "$PWD":/workspace \
  -e REPO_ROOT=/workspace \
  -e DM_CLI=/app/datalex \
  datalex:local

In the UI, choose:

/workspace

as the dbt project path.

Notes

  • AI provider keys should still be configured through the UI or environment.
  • Ollama must be reachable from the container if you use a local model.
  • Generated DataLex files are written back to the mounted dbt repo.