How the roll-up gate works
This tool grades every change in a schema / ontology release against what it breaks downstream, then rolls the release up to its worst change. It reads 7 changes and rolls them into one verdict — the worst, never the average — then a gate that can only make it worse.
The 7 changes
A change that reads ADDITIVE on its three answers but whose consumers were never listed is escalated to SILENT BREAK — 'additive' is a claim about everyone who reads the field, and you can't make it about consumers you can't name. The release then takes its worst change: one silent break reads SILENT BREAKS PRESENT however high the additive share, and the gate can only ever escalate a quiet change, never soften a loud one.
Why: Two of seven changes are additive and two are ordinary breaking changes — loud, expensive, and safe. But three are silent breaks: C-01 and C-05 move a field's meaning while everything still validates and runs, and C-03 reads ADDITIVE on its answers yet nobody could name its consumers, so the gate escalates it. One silent break overrides the 29% additive share, so the release reads SILENT BREAKS PRESENT.
No single change is the crisis — the clustering is. The roll-up takes the worst, the gate escalates when trouble compounds, and it names the one thing to fix first. Grades the changes you describe, never people. A working aid, not legal or compliance advice, and not an audit or a certification. ADDITIVE means your four answers said so, not that nothing downstream will notice.