2026-10-23 - ADI Proves a Rule Works Before Offering to Activate It¶
Release¶
AgileData.io - Tested Before Trusted: ADI Checks Its Own Rules
When ADI builds a change rule for you, it now proves the rule works before suggesting you activate it. It runs the rule over a recent sample of your data and checks five things: the rule runs, the row count makes sense, the grain holds, the columns are right and the numbers reconcile.
ADI also keeps one draft per rule. If you ask it to try again on a table that already has a draft, it updates that draft instead of saving a second rule for the same target. If the target already belongs to someone else’s draft or a live rule, ADI stops and asks you first.
What’s New:
ADI previews and checks each rule against a recent sample of your data
A second check of the saved draft before activation is offered
One draft per rule, with no duplicate rules for the same target
ADI stops and asks when a target already has another rule
What this means for you:
Fewer broken rules reaching activation
No clutter of duplicate drafts from repeated attempts
The same level of checking you’d expect from an experienced data engineer
Last Refreshed¶
Doc Refreshed: 2026-11-13