SELF-CHECK, read 13 Nov 2026
Twelve questions tell you how ready one asset is and what to fix first
Pick one asset: a table, a metric or a dashboard that you plan to give to an agent. Answer each question for that asset only, then read the level and the first fixes.
The page is a plain form. Printed, it is a worksheet. With JavaScript on, a small script adds the answers in your browser and names the level; it does not store the answers and does not send them anywhere.
The levels
| Level | Score | What it means |
|---|---|---|
| Not ready | 0 to 6 | Most basics are missing, so keep agents away from this data for now. |
| Early | 7 to 13 | Some basics are in place, so limit the agent to read-only help on a few assets. |
| Nearly ready | 14 to 19 | Most conditions hold, so fix the remaining gaps before you allow wider use or writes. |
| Ready | 20 to 24 | The asset meets nearly every condition and is fit for the use you checked. |
What the questions map to
| Question | Conditions |
|---|---|
| 1. Does this asset have a named person or team recorded as its owner who still answers messages? | Named owner |
| 2. Does the asset and each column the agent will use have a plain business description? | Documented |
| 3. Is the asset marked certified, endorsed or verified by someone with authority? | Certified |
| 4. Do quality tests run on the asset and are they passing today? | Quality tests pass |
| 5. Is the asset updated within the freshness window your use expects, and is that window written down? | Fresh |
| 6. Is sensitive data classified, with PII columns flagged or the asset confirmed PII-free? | Sensitivity classified |
| 7. Can you trace the asset back to its sources, down to the key columns? | Lineage known |
| 8. Is the asset in active use and not marked deprecated? | Active use |
| 9. Does each metric the agent will report have one governed definition? | Metric defined |
| 10. Does an access policy say this agent may use the asset for this purpose? | Access policy allows AI use |
| 11. If the agent writes back, is each write target approved, limited and reversible? | Access policy allows AI use, Named owner, Quality tests pass |
| 12. If the agent edits models, can it see what sits downstream and who reviews its changes? | Lineage known, Named owner, Documented |
Sources
- Atlan: How to assign owners and stewards for AI-ready data, read 13 Nov 2026
- Qlik: The Essentials of AI-Ready Data, read 13 Nov 2026
- Atlan: AI-ready data for talk-to-data, read 13 Nov 2026
- dbt Labs: 5 dbt MCP server patterns that work in production, read 13 Nov 2026
- Gartner: Lack of AI-Ready Data Puts AI Projects at Risk (Q&A with Roxane Edjlali, Feb 26, 2025; archived copy), read 13 Nov 2026
- 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.