# AI data readiness worksheet
Use fictional or approved information. Readiness is specific to a task, not a universal dataset score.

## Task boundary
- Supported questions, products and languages:
- Excluded uses and actions:
- Owner and review date:

## Source register
| Source / version | Data use: training, retrieval or evaluation | Permission evidence | Effective date | Extraction check | Coverage gap | Owner / next action |
|---|---|---|---|---|---|---|
| | | | | | | |

## Coverage record
- Planned question types and selection method:
- Types with verified evidence / total planned:
- Unsupported types and fallback:
- Why this is not an accuracy estimate:

## Rehearsal
| Case | Expected evidence | Expected boundary | Actual result | Next action |
|---|---|---|---|---|
| Standard question | | | | |
| Missing exception | | | | |
| Archived policy | | | | |
| Unresolved permission | | | | |

## Bounded decision
- Ready for which use:
- Blocked sources and unresolved evidence:
- Source / index / cache change owner:
- Stop condition and retest trigger:

Source guide: https://learn.aicompetence.org/knowledge/assess-ai-data-readiness/