---

name: data-metadata-and-sqllab description: "Route CubeStudio dataset, metadata, dimension, SQLLab, and ETL data workflows." metadata: disco-role: operating disable-model-invocation: true license: NOASSERTION

Data, Metadata, and SQLLab

Use this sub-skill for dataset catalog records, metadata and dimension tables, SQLLab request lifecycle, and ETL data movement.

Use this sub-skill when

• you need to inspect or explain dataset fields, storage paths, downloads, previews, or version history
• you need to validate metadata tables, metric catalogs, or dimension-table schemas and permissions
• you need to check SQLLab engine names, URI template shape, query status, or result download behavior
• you need to understand ETL pipeline records, task syncing, or the bundled data-import/export templates

Consult first

• references/data-workflows.md
• references/sqllab-and-engines.md
• references/etl-pipelines.md
• references/troubleshooting.md
• scripts/validate_sqllab_request.py

Route away from here

• General Argo workflow mechanics → sibling ../pipelines-and-job-templates/
• Notebook and image catalog work → sibling ../compute-notebooks-and-images/
• Cluster install, registry, or deployment operations → sibling ../deploy-and-operate/
• Serving, AIHub, or inference-service deployment → sibling ../serving-aihub-and-llm/

Safe operating rules

• Prefer static inspection and bundled validation only.
• Do not start databases, schedulers, containers, or long-running services.
• Keep source evidence distilled into the bundled references and script.