datalineage-summary

Summarizes Google Cloud Data Lineage graphs into Markdown reports for BigQuery and GCS assets.

19.1k|1.5k|Updated Mar 31, 2026
One-click install
npx skills add https://github.com/google/skills --skill datalineage-summary
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: datalineage-summary
Source: https://github.com/google/skills/tree/main/skills/cloud/datalineage-summary
Command: npx skills add https://github.com/google/skills --skill datalineage-summary

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes references (resource) components.

What problem does it solve?

Understanding how data flows into and out of a BigQuery table or GCS asset requires navigating complex lineage graphs of nodes and links, which is difficult to interpret when debugging data quality issues or verifying data provenance.

Core Features & Use Cases

  • Bidirectional Lineage Retrieval: Queries the Google Cloud Data Lineage MCP server in both upstream and downstream directions with configurable depth, location, and result limits.
  • Table and Column-Level Lineage: Supports Table-Level Lineage and Column-Level Lineage via wildcard or specific field configuration in search_lineage calls.
  • Plain-English Markdown Summaries: Produces a left-to-right narrative covering overall flow type, systems involved, upstream sources, downstream consumers, and analysis metadata.
  • Use Case: A data engineer investigating a corrupted BigQuery reporting table uses this Skill to trace which upstream Dataproc jobs and source tables feed it, and which downstream dashboards consume it, receiving a readable summary instead of raw graph data.

Quick Start

Summarize the upstream and downstream data lineage for the BigQuery table my-project.my_dataset.my_table and explain where its data comes from and goes.

Frequently Asked Questions about datalineage-summary

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I summarize data lineage for a BigQuery table?

Use the Data Lineage MCP server's search_lineage tool with the table's fully qualified name (bigquery:project.dataset.table), making one call with direction UPSTREAM and another with DOWNSTREAM. The results are then summarized into a Markdown narrative of sources and consumers.

How to get column-level lineage in Google Cloud?

Configure the search_lineage rootCriteria with a field array containing the specific column name, or use the "*" wildcard to retrieve all column-level lineage links alongside table-level links for the asset.

Does Google Cloud data lineage support cross-region assets?

Yes, but you must populate the locations array in the search_lineage call with all supported physical regions. The current list of regions should be fetched dynamically from the Knowledge Catalog locations documentation to avoid missing cross-regional lineage.

When should I not use lineage summary for BigQuery?

Do not use it for generic BigQuery queries, editing lineage relationships, or downstream blast-radius impact analysis. For impact analysis of deprecating an asset, use the datalineage-bigquery-asset-impact-analysis skill instead.

Why is my lineage search returning incomplete results?

Incomplete results usually stem from a restricted locations array or shallow depth limits. Use maxDepth of 10, maxResults of 5000, and include all supported GCP regions in the locations array to capture the full graph.