Skip to content
Ready Mark

CHECKLIST, CONDITION 4 OF 10, read 13 Nov 2026

Quality tests pass: 17 of 17 tools check it by rule, 0 leave it to a person

The checklistCoverage matrix

The test
Quality or health tests exist for the asset and are passing now.
Why it matters
An agent has no instinct for bad data. If nulls, duplicates or broken joins get through, the agent reports them as fact.
How to check it
Check the latest run of the dbt tests or monitors on the asset. Look for coverage on keys, nulls and row counts, and confirm the last run passed.
What failing looks like
A duplicate-key test has failed for three days, and the agent still sums the table and double counts orders.

What each use case needs

Use caseNeedWhy
Metric agentrequiredNumbers the agent reports are only as sound as the data under them.
Coding agentrecommendedTests help a coding agent check its own change, but the work does not depend on current data being clean.
Write-back agentrequiredWriting bad values into a system of record spreads the error and is hard to undo.

17 tools check quality tests pass by rule

0 tools leave quality tests pass to a person

  • None on 13 Nov 2026.

0 tools show none found on 13 Nov 2026

  • None on 13 Nov 2026.

What the sources say

Other conditions

Sources

  1. Gartner: AI-ready data: what it is and how to get there (Rita Sallam; archived copy), read 13 Nov 2026
  2. Monte Carlo: Redefining AI-ready data, read 13 Nov 2026
  3. Qlik: The Essentials of AI-Ready Data, read 13 Nov 2026
  4. McKinsey: A data leader's technical guide to scaling gen AI (archived copy; mckinsey.com did not load from this run), read 13 Nov 2026

Last reviewed 13 Nov 2026.