col-lineage

name	col-lineage
description	Generate column-level lineage for a specific task

Column Lineage Skill

Generates column-level lineage for a specific transform task, showing how each output column traces back to its source columns through SQL transformations. This provides fine-grained data provenance information.

Usage

starlake col-lineage [options]

Options

• --task <value>: Task name in the form domain.task (required)
• --output <value>: Output JSON file path (default: console output)
• --accessToken <value>: Access token for authentication (e.g. GCP)
• --reportFormat <value>: Report output format: console, json, or html

How It Works

Analyzes the SQL query for a task and traces each output column back to its source columns. For example, given:

-- metadata/transform/kpi/revenue_summary.sql
SELECT
    o.order_id,
    o.timestamp AS order_date,
    SUM(ol.quantity * ol.sale_price) AS total_revenue
FROM
    starbake.orders o
    JOIN starbake.order_lines ol ON o.order_id = ol.order_id
GROUP BY
    o.order_id, o.timestamp

The column lineage output would show:

• order_id ← starbake.orders.order_id
• order_date ← starbake.orders.timestamp
• total_revenue ← starbake.order_lines.quantity, starbake.order_lines.sale_price

Examples

Generate Column Lineage for a Task

starlake col-lineage --task kpi.revenue_summary

Save Column Lineage to File

starlake col-lineage --task kpi.order_summary --output col_lineage.json

Column Lineage with Access Token

starlake col-lineage --task kpi.revenue_summary --accessToken $GCP_TOKEN

Related Skills

• lineage - Task-level dependency graph
• table-dependencies - Table relationship graph
• transform - Run transformation tasks