Quickstart

Install the CLI

The lmk CLI manages your project, validates YAML, and syncs definitions to the Loomkindle cloud.

pip install loomkindle-cli
lmk --version

Initialize your project

Run lmk init in a new directory. This creates the project scaffold with sample YAML configs.

mkdir my-data-project && cd my-data-project
lmk init --warehouse snowflake

Connect your data warehouse

Edit lmk.yml with your warehouse credentials. Supported: Snowflake, BigQuery, Redshift, DuckDB, Postgres.

connection:
  type: snowflake
  account: your-account.us-east-1
  warehouse: COMPUTE_WH
  database: ANALYTICS
  schema: PUBLIC

Define your first semantic model

Create a YAML file in models/ to define a metric. The agent will validate schema, infer entities, and register the model.

version: 1
semantic_model:
  name: revenue
  source: orders
  metrics:
    - name: total_revenue
      type: sum
      expr: order_amount_usd
      label: Total Revenue (USD)

Validate and deploy

Run validation locally then push to Loomkindle cloud. The agent will confirm your model is consistent with the live schema.

lmk validate
lmk deploy --env production

You should see: Deployed 1 semantic model. 0 schema conflicts detected.