Docker is optional. Use it when local Python or Node setup is blocked.
git clone https://github.com/duckcode-ai/DataLex.git
cd DataLex
docker build -t datalex:local .docker run --rm -p 3030:3001 datalex:localOpen:
http://localhost:3030
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:localIn the UI, choose:
/workspace
as the dbt project path.
- 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.